Embedded Generation Project - Final Report v1

Embedded Generation Project –
Final Report
Energy Networks Association

November 2015
Disclaimer
This document is intended for public distribution.
Marchment Hill Consulting, its partners, employees and agents neither owe nor accept any duty
of care or responsibility to such persons, and shall not be liable in respect of any loss, damage or
expense of any nature which is caused by any use they may choose to make of this report. The
information outlined herein is proprietary and its expression in this document is copyrighted,
with all rights reserved to Marchment Hill Consulting. Any form of reproduction, dissemination,
copying, disclosure, modification, distribution and/or publication of this document without
express written permission from Marchment Hill Consulting is strictly prohibited.
Table of Contents
Executive Summary .........................................................1
Abbreviations used in this report ................................................... 12
1 Background ........................................................ 15
2 Scope and approach.............................................. 17

2.1 Scope ......................................................................... 17
2.2 Approach .................................................................... 17
2.3 Document maps............................................................. 18
3 Technical impact assessment ................................... 20

Note on authorship ................................................................... 20
3.1 Introduction ................................................................. 20
3.2 Technical constraints and network impacts ............................ 21
3.3 Options for mitigating network impacts and risks ..................... 61
3.4 Options for maximising network benefits ............................... 97
3.5 Australian Standards gap analysis ...................................... 146
4 Regulatory impact assessment ................................ 160

Note on authorship ................................................................. 160
4.1 Introduction ............................................................... 160
4.2 The National Electricity Customer Framework....................... 160
4.3 Non-National Electricity Customer Framework states .............. 162
4.4 Cost of connection ....................................................... 165
5 Commercial impact assessment ............................... 168

5.1 Introduction ............................................................... 168
5.2 Commercial impacts ..................................................... 169
5.3 Valuation framework .................................................... 181
5.4 Commercial opportunities .............................................. 194
5.5 Implications for the network service provider business model .... 210
6 Policy and regulatory options ................................. 213

6.1 Introduction ............................................................... 213
6.2 Key technical findings ................................................... 213
6.3 Key regulatory and commercial findings .............................. 218
7 Recommendations ............................................... 227

7.1 Introduction ............................................................... 227
7.2 Enabling technologies and pricing ..................................... 228
7.3 Enabling business models ............................................... 233
7.4 Enabling partnerships .................................................... 235
Appendix A – Survey responses ......................................... 237
Appendix B – Analysis of embedded generation/distributed energy resources

uptake and network impact forecasting techniques ....... 250
Introduction ......................................................................... 250
Review of methodologies .......................................................... 250
References ........................................................................... 257
Appendix C – Maximising network benefits: overview of relevant literature

..................................................................... 262
Photovoltaics and storage ......................................................... 262
Diesel, gas, steam and hybrid generation systems............................. 265
Embedded Generation Project - Final Report
Executive Summary
One of the most significant technical, regulatory and commercial challenges faced by the
electricity industry to date is the continued uptake of embedded generation (EG)1. This has
involved integrating EG into Australian electricity distribution networks, while maintaining the
required levels of quality, safety, reliability and security of supply.
The Energy Networks Association (ENA) has launched the Electricity Network Transformation
Roadmap (NTR) project to identify the preferred transition which the electricity network
industry must make in the next decade, to be ready to support better customer outcomes under
a diverse range of long-term energy scenarios. This includes investigating and assessing the
impacts of continued EG growth on Australian electricity network service providers (NSPs) and
determining how EG integration can be undertaken in a way that is beneficial for both customers
and NSPs.
This report is a foundational piece for the NTR project, and is scoped to provide an overview of
the key technical, commercial and regulatory impacts of EG. It looks at the opportunities EG
presents for NSPs, how regulatory and policy options can be pursued to address the impacts, and
how customers can receive greater value as either consumers, producers or prosumers2 of
electricity.
The report takes a proactive approach to new technologies and services. It has been prepared
with the overarching principle that networks should actively develop industry-wide solutions,
including integrating new technologies, to serve the long-term interest of their customers.
Audience
This report is intended to be released as a public document. It has been prepared on behalf of
the ENA and its members in a process that has included extensive stakeholder consultation,
input and review. Its public release is intended to help advance the customer’s ability to benefit
from efficient and effective network operation through economic integration of EG, and
seamless service offerings in EG related markets.
Objectives
The objectives of this report are to:
 investigate and assess the technical, regulatory and commercial impacts of continued
growth in EG on Australian distribution and transmission networks
1An embedded generating unit is defined as ‘a unit that generates electricity at a customer’s premises and is
connected to a distribution system’. It includes, but is not limited to, micro-embedded generation units, electrical
energy storage and electric vehicles.
2 A prosumer is defined as a customer who both produces electricity (via an embedded generator, which can be
supplied to the grid) and consumes electricity from the grid. A consumer is a customer who only consumes electricity
from the grid.
1
 identify potential policy and regulatory options that the ENA and its members should
consider in response to these impacts, to ensure the safe, reliable, affordable and
efficient operation of the electricity network in the long-term interests of customers
 develop a high-level commercial framework to value EG and assess the commercial
opportunities for NSPs.
Approach
In preparing this report, Marchment Hill Consulting (MHC) and the CSIRO undertook a desktop
review of relevant literature, a survey of ENA members to gather views regarding the impact of
EG on their network, and in-depth interviews to gather further insights. Workshops were held
with the ENA and its members to test initial hypotheses, generate additional ideas, and review
and refine the report’s findings and recommendations. Draft sections of the report were also
distributed to the ENA and its members for review and feedback prior to completion of the final
report.
Authorship
The authorship of this report was shared between CSIRO and MHC. CSIRO focused on the
technical impacts of EG, while MHC focused on the commercial and regulatory impacts. All
policy and regulatory options and recommendations have been developed by MHC and represent
their views only.
Limitations
The report has the following key limitations:
 It is necessarily broad in scope. Its intention was to set out the broad landscape as a
background to further work by the ENA and its members, and it therefore addresses a
range of topics and highlights areas for further work for which there were neither time
nor resources to investigate further.
 ENA member views were gathered through a survey and a selection of interviews,
however, the results do not necessarily represent the views of ENA or all relevant NSP
stakeholders.
 The impact of EG often varies considerably between NSPs, which could be due to factors
such as different EG penetration levels, network asset conditions and peak demand
levels. Wherever possible, we have attempted to highlight the variability of key findings
between networks, but note that not all findings and recommendations will be of similar
relevance to all NSPs.
Key Findings
This report reviews the technical, commercial and regulatory impacts of EG on the Australian
electricity network and builds on previous work such as the Impacts and Benefits of Embedded
2
Generation (IBEG) report3. We have found that increasing levels of EG on the network presents
both opportunities and challenges for NSPs and customers – and that given appropriate
conditions and incentives, EG can deliver significant benefits to NSPs and their customers alike.
The challenge for industry is to ensure that the policy, regulatory and commercial environment,
as well as relevant safety and technical standards, support the efficient deployment and use of
EG in the long-term interests of all customers.
The sections below present the key technical, regulatory and commercial findings identified in
this report. For highly complex findings, additional points of clarification or context are also
provided.
Key Technical Findings
The findings in this section address the technical issues and options relating to increasing
penetration of EG. Technical Constraints and Network Impacts will describe the key challenges
associated with the network impacts of EG. Following on from this, Options for Mitigating
Network Impacts and Risks will itemise ways in which networks can address these issues. Options
for Maximising Network Benefits will demonstrate several in which EG can act to provide
significant benefit to NSPs, industry stakeholders and customers (both directly and indirectly
through reduced pricing). Finally the Australian Standards Gap Analysis itemises a number of key
gaps in existing standards relating to EG.
Technical Constraints and Network Impacts

The findings identified in this section represent the ways in which EG can have negative impacts
on the function, operation and maintenance of networks, or where EG can apply a new network
constraint (or impinge upon an existing network constraint). These findings relate to the current
state of networks as well as future business-as-usual operation; they may be manageable
through deployment of techniques such as those included in Options for Mitigating Network
Impacts and Risks below. Within the context of traditional network operation, we conclude that
EG has benefits, albeit with an increase in the complexity of network operation.
1. Increasing penetration of EG, including greater uptake of rooftop solar and other EG, is
changing load profiles on Australian electricity networks and making network load
prediction more challenging. Appropriate planning and management of network load
demand across different timeframes is essential to ensure both adequate power quality
and a reliable supply of electricity of the network.
o High penetration of EG increases the gap between the maximum and minimum
network load. In addition to peak demand forecasts, accurate minimum demand
and energy consumption predictions are required. These will assist in network
planning relating to managing voltage and potential reverse power flow issues.
o Accurate prediction of network load demand is essential for effective utilisation
and management of EG sources on the network, and to establish sustainable load
management systems for the smart grid.
2. Reverse power flow has been observed at the medium-voltage and high-voltage network
levels. In radial networks designed for unidirectional power flow, this can affect the
3 Energy Networks Association, The Impacts and Benefits of Embedded Generation, 2011
3
operation of protection and voltage regulation devices. In islanded networks, such as

some remote communities, it can also have a detrimental effect on generator control
and stability.
3. EG has the potential to contribute current to any fault in the vicinity. Protection system
design and settings must be considered in rural and off-grid systems to ensure faults are
cleared quickly. In urban areas, the added fault level (the maximum current that can
flow in the network segment as a result of a fault) may restrict the amount of EG that
can be connected due to system design fault levels.
Options for Mitigating Network Impacts and Risks

This section highlights the options available to networks for minimising — and if possible,
eliminating — the negative impacts and constraints of EG identified in Section 3.2, to realise the
benefits of EG while reducing the associated risks. The options discussed focus primarily on
technical considerations; cost of implementation is discussed under the commercial impacts in
Section 5.
4. Grid-scale storage has the potential to provide a range of network benefits relating to
EG, including voltage management, energy balancing, and improving network stability.
These benefits can increase the flexibility, reliability and efficiency of power delivery to
consumers.
o Currently, grid-scale storage is limited due to a combination of system cost and
market maturity, and as such is generally only deployed in specific applications
and locations as noted in Section 3.3.2.1. However, considering the market for
energy storage devices has been estimated to reach 3000 MW by 2030, as well as
expected improvements in related technologies, the deployment of grid-scale
energy storage is predicted to see increasing uptake in the coming years.
5. Distributed storage can provide numerous benefits to networks, including improved
management of voltage and power flows, peak load and generation management, and
reactive power support. Customer-owned storage has the capacity to provide these
benefits alongside direct benefits to the customer, despite not being network-owned.
o For customer-owned storage, appropriate dispatch management is key. Unlocking
its network benefits requires appropriate incentives to ensure that it is
dispatched at times that are beneficial to networks. If incentives are not well-
designed, customers may utilise storage in ways that are detrimental to the
network. If direct control of non-network storage is desired, the deployment of
operating systems (e.g. through demand response systems such as AS/NZS 4755
Part 3.5) and relevant communications platforms will enable controlled dispatch
via consistent protocols.
6. Power electronics solutions, such as STATCOMs and smart inverters, have the capability
to mitigate power quality issues relating to EG, including managing voltage ramp-rates
and excursions. In some cases, they can also reduce harmonic content.
o Smart inverters in particular can use a pre-existing generation resource to self-
manage their own generation. As this functionality is not currently mandated, the
penetration of these devices in networks is low. Retrofitting existing fleets of
inverters would require a large capital investment.
4
o The current revision of AS/NZS 4777 part 2 includes some smart inverter functions
to self-manage voltage issues where possible. However, this functionality is not
currently mandatory within the standard, which may continue to limit market
penetration even once the updated standard is published.
7. Improved prediction of renewable generation, informed through dedicated metering, can
improve EG integration by increasing the utilisation of assets such as storage and solar
photovoltaic (PV)-controlled air conditioning. It can also inform decisions on power
system unit commitment and network planning. These benefits can be broadened beyond
EG to medium and large-scale generation.
8. Improved network load prediction techniques that incorporate increased penetration

levels of EG are necessary to retain the current benefits of short, medium and long-term
prediction of net load in a high-penetration EG environment. The prediction techniques
also need to incorporate any other factors that may be contributing to changing
utilisation patterns of electricity networks.
Options for Maximising Network Benefits

The findings described below summarise the key benefits of EG for networks and network
operators: reduced operating and maintenance costs, and increased lifespan of network assets.
9. Utilising EG for power and voltage profile levelling reduces the operation of expensive
peaking plants. It can also ease voltage management requirements, improve power
system reliability and provide frequency support.
10. Insecure4, non-dispatchable energy source types of EG alone (predominantly PV and
wind) provide only limited power and voltage levelling and frequency support. Adding
storage to these offsets the insecure, intermittent nature of PV and wind, and allows
such EG to provide these services effectively.
11. EG can contribute to peak load reduction. This alleviates congestion, reduces line losses
and results in deferral of equipment capacity upgrades where capacity is constrained.
12. Evidence from the literature and distribution network service providers suggests that the
reduction in cumulative net load due to EG can extend the life of both substation and
distribution transformers.
Australian Standards Gap Analysis

An analysis of the state of Australian Standards relating to EG identified the following key gaps:
13. Australian standards do not well cover the functionality and performance requirements
of protection relays used to prevent islanding and reverse power flow relating to
embedded generators. This reduces the consistency with which DNSPs can implement
these devices, with significant flow-on effects for the market and other industry
stakeholders. These effects include increased cost of development, implementation and
4 ‘Insecure’ relates to the operational security of the generation unit: i.e., a diesel generator can be considered
secure because it is generally dispatchable, while a solar PV system without storage is uncontrolled, and is hence
insecure.
5
commissioning; increased cost and reduced consistency of type-testing; and potentially

reduced performance and reliability in-situ.
14. There is a significant lack of standards relating to the integration and operation of
electric vehicles (EVs), including vehicle-to-grid-enabled EVs. This may increase the risks
and reduce the potential benefits associated with EVs by limiting the uptake of EV
demand response and vehicle-to-grid systems, as well as reducing consistency of
approach and increasing implementation costs.
15. Grid-connected stationary energy storage systems utilising emerging technologies
including lithium-chemistry and zinc-bromine flow batteries, are not well standardised in
Australia particularly in the areas of safe installation, operation and disposal. This will
likely have an impact on consistency of approach between jurisdictions, with significant
implications for the market and industry; it may also increase the likelihood of incidents
relating to the improper use of these systems.
Key Regulatory and Commercial Findings

This section reviews the current regulatory framework as it relates to connecting EG to the
network, the commercial impacts of EG, and the development of a commercial framework to
value these impacts. It also assesses the commercial opportunities of EG available to NSPs.
The key findings are categorised below as either regulatory or commercial.
Key Regulatory Findings

The below findings include impacts that need addressing through a regulatory focus.
16. Although recent reforms to Chapter 5 and 5A of the National Energy Rules (NER) have
improved the connection process related to EG, the lack of a consistent national
framework means that these have not been adopted uniformly across all jurisdictions.
17. Where augmentation costs may occur because of high levels of small scale solar PV
penetration on a network (i.e. basic connections), the NER does not allow for these costs
to be recovered directly from the EG owners in a cost-reflective approach. Instead,
these costs are recovered from the entire customer base which introduces an element of
cross-subsidisation between customers.
18. Within the Demand Management Incentive Scheme (DMIS), Demand Management
Innovation Allowance (DMIA), RIT-D and RIT-T processes networks have opportunities to
consider and implement non-network solutions including EG to resolve network
constraints in the most cost effective manner, however there are areas for improvement
and potential for change as part of current regulatory reform initiatives.
19. Currently, there is no universally applicable and agreed regulatory model to value the
impact of EG on networks. Networks require an appropriate valuation approach if they
are to identify the costs and benefits of EG and to signal the efficient sizing, location
and operation of EG. In addition, there is a gap in the current understanding of the
impact of EG on network reliability, safety and quality of supply and an appropriate
valuation approach is also needed to allow NSPs to identify these impacts.
20. Ring-fencing requirements for NSPs currently differ by jurisdiction and are generally
seen as inadequate for the purposes of emerging markets and technologies. In particular
6
it is unclear how storage technologies are to be treated under the current ring-fencing
requirements and how NSPs might support the efficient creation of microgrids5 or
standalone systems as an option for the most efficient approach to energy supply.
Key Commercial Findings
The below findings include impacts that need addressing through a commercial focus.
21. Current volume-based network pricing structures includes a cross-subsidy from non-EG
owners to EG owners. The recent changes to the National Electricity Rules require
networks to set prices that reflect the costs of providing electricity to consumers with
different patterns of consumption. These tariffs are cost reflective and are not specific
to the technology choices of customers (i.e. are ‘technology neutral’). There is
opportunity within these new rules to support the efficient uptake and use of EG that
reduces future costs to customers to below the level they might reach under volume-
based pricing structures.
22. Several of the potential commercial opportunities for networks to offer new products
and services based on EG depend on the networks having cost-effective access to
customer data and/or a direct channel to the customer. (e.g. in home EG and demand
optimisation, residential storage). The proposed new metering rules present
opportunities for NSP related entities to participate in new customer–oriented markets
for metering and related services but they also create risks associated with potential
new costs for data access and the loss of a traditional DNSP service.
23. Active involvement in the EG and energy services markets on the part of NSPs or their
related entities as either a market participant or to simply incentivise the efficient use
of EG for the benefit of all customers, will require developing partnerships with other
stakeholders including technology service providers and retailers.
24. The NSP business model needs to evolve to facilitate the integration of multiple
distributed energy resources, including EG. This could provide greater network capacity
and energy diversity to optimise grid performance for both supply and demand.
Recommendations
These recommendations have been prepared for the consideration of the ENA and its members
to act upon, particularly in future stages of the Network Transformation Roadmap initiative.
They have also been prepared with the overarching principle of the National Electricity
Objective in mind. That is, that investment in and operation of the network should maximise
the efficiency of the electricity supply chain, allowing consumers to enjoy the quality, safety,
reliability and security of supply that reflects their long-term interests, at the lowest cost. This
should result in the efficient deployment and use of EG at times, locations and scales that
reflect its value to the customer, community and network.
5A microgrid is defined as ‘a group of interconnected loads and distributed energy resources within clearly defined
electrical boundaries that acts as a single controllable entity with respect to the grid and that connects and
disconnects from such grid to enable it to operate in both grid-connected or “island” mode.’ For the purposes of this
report, this definition is extended to also include permanent island networks.
7
EG may benefit customers by delivering new energy services and supporting more efficient use
of existing network infrastructure. However, increasing levels of EG on the network presents
challenges for NSPs which would ultimately have cost or risk consequences for network
customers if not managed appropriately. The challenge for networks, retailers, regulators,
technology proponents and customer representatives alike is how to work together to develop
business models and pricing structures that fairly apportion value and costs.
The recommendations presented here recognise the often diverse responsibilities and interests
of stakeholder groups, and aim to align their actions for the benefit of customers.
The recommendations have been grouped into three key main categories:
 Enabling Technologies and Pricing — what are the key technologies likely to deliver
future value, and how could they be incentivised?
 Enabling Business Models — how could the network of the future operate, and what
regulatory changes should be implemented to promote this transition?
 Enabling Partnerships — what partnerships are likely to be required for network service
providers to actively participate in the EG market?
Policy and regulatory options that support the recommendations are also provided below, as
appropriate. Policy options relate to approaches that involve a change to existing government
policy or to the general operational policies or approaches adopted by NSPs.
Enabling Technologies and Pricing
An appropriate valuation approach to capture all the costs and benefits of EG to networks
should be developed and implemented in close consultation with stakeholders to enable
effective signalling of the network value of EG to customers.
Policy and regulatory options that could be considered in future stages of the Network
Transformation Roadmap to support this recommendation include NSPs:
 Working closely with a range of industry stakeholders (including EG proponents, the

Australian Energy Market Operator and the Australian Energy Regulator) to develop an
appropriate valuation framework that can be leveraged to develop pricing arrangements
to promote efficient uptake of EG.
 Developing appropriate price signals to encourage efficient uptake of EG, reducing future
network costs to below the level they might reach in the absence of EG uptake and
thereby benefiting all customers. This could include:
o discounting connection charges, network charges or providing direct payment,

reflecting the costs and benefits available to the NSP, or
o implementing price signals or incentives for customers to undertake efficient

investment in storage, retrofit distributed inverters and to be appropriately
rewarded for demand management.
 Using the valuation framework to manage the voltage regulation, power quality,
reliability and safety impacts of EG on the network. This could, for example, support the
8
strategic deployment of medium-scale power electronics solutions, such as STATCOMS, at

locations where they may be the least-cost solution to help manage power quality issues.
Support the implementation of cost-reflective network price signals that promote uptake
of EG in ways that are beneficial to all customers.
 Supporting a national approach to cost-reflective pricing that removes inconsistent

jurisdictional obligations on tariff structures and assignment.
 Supporting the implementation of a balanced framework for the uptake of smart meters
for the fastest economic roll out to benefit all customers.
 Encouraging adoption of technologies that support efficient pricing and customers’ ability
to access data and manage their demand (e.g. in-home displays, apps, data portals).
 Working closely with retailers, technology proponents and metering service providers to
bundle products and tariffs to optimise value for customers and simplify product offerings
while supporting efficient network outcomes.
Support incentives and policies that promote innovation and commercialisation in relation
to energy storage technologies in ways that enhance the efficient use of the network for
the benefit of all customers.
 Identifying and supporting opportunities to restructure current incentives and subsidies so

that they can support technologies and services that result in more efficient use of the
network for the benefit of all customers (for example, restructuring premium feed-in-
tariffs to enable customer to choose to use these subsidies to install storage solutions
which support lower peak demand and reduced network expenditure).
 Working with the industry and regulators to build on the principles of the Demand
Management Innovation Allowance to enhance the industry’s ability to support research
and development activities that support innovation and integration of new technologies,
and ultimately create a more efficient network for the benefit of customers.
 Investigating opportunities for the development of cost-effective policies and incentives

(such as model availability requirements and fuel standards) supporting EV uptake, where
it promotes efficient use of the network for the benefit of customers.
Support the development of appropriate standards that facilitate the safe integration of
EG on the network.
 Supporting the development of a new Australian Standard to manage the functionality of

protection relays for inverter energy systems.
9
 Revisiting the development of standards such as AS/NZS 4755 Part 3.4, or facilitating the
local acceptance of a similar international standard for managing EV charging and
discharging.
 Supporting the development of standards relating to the safety of small-scale and

large/grid-scale energy storage systems. Focus should be directed to those addressing
safe installation/operation and fire protection. In addition, lithium chemistry may
require particular attention, because its standards are immature.
Enabling Business Models
There should be no inherent advantage or disadvantage for network service providers in

competing to provide services and pursuing commercial opportunities related to EG.
 Supporting changes to current state-based ring-fencing rules to create a national

framework with the overarching principle to ensure no market participant is
unnecessarily constrained in a competitive environment with regard to providing services
to customers.
 Supporting the removal of state-based rules barring or limiting network providers’ ability
to own and operate EG (for the purpose of network support).
 Ensuring that they are free to enter into direct commercially based trading relationships
with customers and that NSP’s related entities should not be restricted from providing
energy selling services to customers, where they meet appropriate ring-fencing
requirements.
 Supporting a review of relevant rules to ensure no barriers exist to the creation,

operation and maintenance of microgrids or standalone systems by networks as an option
for the most efficient approach to energy supply.
There should be no inherent advantage or disadvantage for network service providers to

utilise EG for network purposes compared with network upgrades.
 Working with the industry and regulators to identify where there may still be a
disadvantage for network providers to utilise EG for network purposes compared with
network upgrades. NSPs could then build on the recent reforms to the Demand
Management Incentive Scheme to ensure that any disadvantages for NSPs using EG in this
way is addressed.
Support the evolution of the network provider business model to include that of the grid
integrator — efficiently and effectively integrating multiple distributed energy resources,
including EG.
10
 Propose that in future market arrangements, the NSP can coordinate the day-to-day
operation of multiple sources and types of EG and other distributed energy resources,
where this provides the most cost-effective option for the customer.
Enabling Partnerships
Network providers should work towards developing partnerships with relevant energy
service providers to promote sustainable and efficient uptake and operation of EG.
 Removing knowledge and organisational barriers to cooperation between NSPs,
technology providers and retailers to offer bundled products.
 Reviewing the current regulatory framework to determine its suitability for a more
customer-oriented electricity market. This should include assessing its flexibility to adapt
to an increasingly complex market as a result of the adoption of new technologies and
the changing priorities of customers.
Critical Further Work

This report has proposed elements of a high-level framework to value the impact of EG on the
network. Integrating the valuation framework into the regulatory framework would require the
Australian Energy Market Commission to initiate a rule change proposal, and further work should
include initiating or supporting this process. Such work would include improving and solidifying
the valuation framework put forward in this report, with the ultimate goal of capturing all the
costs and benefits relating to EG.
Once a robust valuation framework is implemented, it could be used to develop appropriate
price signals to encourage the efficient uptake of EG and minimise future network costs for
customers. It could also be used to determine where current subsidies and incentives could be
restructured and redirected towards technologies and services that support efficient use of the
network. A single solution is unlikely to be suitable for all jurisdictions. Further work should
therefore focus on research and modelling to understand appropriate approaches for each
jurisdiction’s circumstances.
We have also identified a lack of relevant Australian Standards for certain types of EG. Given the
potential benefits to customers and NSPs alike from small-scale energy storage, further work
should support the development of standards for the safe installation, operation and control of
such EG systems.
11
Abbreviations used in this report

AC alternating current
ABS Australian Bureau of Statistics
AEMC Australian Energy Market Commission
AEMO Australian Energy Market Operator
AER Australian Energy Regulator
AGC automatic generation control
AEP American Electric Power
BES battery energy storage
BESS battery energy storage system
BEV battery electric vehicle
CAES compressed air energy storage
CAIDI Customer Average Interruption Duration Index
CBD central business district
CEC Clean Energy Council
CES community energy storage
CHP combined heat and power units
CSO community service obligation
CSP concentrating solar power
CST concentrating solar thermal
DC direct current
DER distributed energy resources
DFIG doubly fed induction generator
DG distributed generation
Demand Management and Embedded Generation
DMEGCIS Connection Incentive Scheme
DMIA Demand Management Innovation Allowance
DMIS Demand Management Incentive Scheme
DNSP distribution network service provider
DR demand response
DRED demand response enabling device
DRM demand response mode
DSTATCOM distributed static synchronous compensator
DTx distribution transformer
DUOS distribution use of system
EG embedded generation
EHV extra high voltage
ENA Energy Networks Association
The European Network of Transmission System Operators
ENSTO-E for Electricity
EPRI Electric Power Research Institute
ERCOT Electric Reliability Council of Texas
EV electric vehicle
FFT fast Fourier transform
12
GW gigawatt
HESS hybrid energy storage system
HV high voltage
IEC International Electrotechnical Commission
IEEE Institute of Electrical and Electronics Engineers
kVA kilo-volt ampere
LOL loss-of-life
LSCPV large-scale centralised photovoltaic
LV low voltage
MAIFI Momentary Average Interruption Frequency Index
MHC Marchment Hill Consulting
MT microturbine
MV medium voltage
MVA mega-volt ampere
MW megawatt
NECF National Energy Customer Framework
NEM National Electricity Market
NER National Electricity Rules
NPV net present value
NSP network service provider
OAF overambient flicker
OLTC on-load tap changer
OpenADR Open Automated Demand Response
PCC point of common coupling
PEV plug-in electric vehicle
PHEV plug-in hybrid electric vehicle
PHS pumped hydroelectric storage
pu per unit
PV photovoltaic
RAB regulatory asset base
RIIO Regulation = Incentives + Innovation + Outputs
RIT-D Regulatory Investment Test for Distribution
SAIDI System Average Interruption Duration Index
SAIFI System Average Interruption Frequency Index
SCB supercapacitor bank
SMES superconducting magnetic energy storage
STATCOM static synchronous compensator
SVC static VAR compensator
SVR step voltage regulator
SWER single-wire earth return
SWIS South West Interconnected System
THD total harmonic distortion
UPS uninterruptible power supplies
US United States
V2G vehicle-to-grid
VAR volt-amps reactive
VCR value of customer reliability
13
VRB vanadium redox battery

XP-DPR Xtreme Power-Dynamic Power Resource
X/R ratio ratio of reactance to resistance
14
1 Background
Over the last five years, solar photovoltaic (PV) systems have emerged as a viable alternative
source for household energy supply. This has been driven by government incentives and a sharp
increase in manufacturing capacity to reduce costs.
The historical uptake of solar PV in Australia is presented in Figure 1.
4,500
4,000
Installed Capacity (MW)
3,500
3,000
2,500
2,000
1,500
1,000
500
-
Figure 1 – Australia’s historical solar photovoltaic uptake, April 2001 to January 2015
Source: Australian PV Institute
The relatively generous, state-based feed-in tariffs (FiTs) that were offered in the past have now
been reduced to below $0.10/kWh in all states. Nevertheless, solar PV uptake has remained
strong and is generally forecast to grow at a steady pace. This is at least partially driven by the
emergence of new solar PV business models that focus on making solar PV more accessible to
consumers, such as solar financing with no upfront costs or residential solar power purchase
agreements via alternate energy sellers.
The increasing penetration of solar PV is starting to pose several challenges for Australian
network service providers (NSPs). One is maintaining a reliable and safe electricity supply while
managing high levels of embedded generation (EG) penetration on their network. Another is an
erosion of NSP customer grid-supplied electricity consumption, and subsequent cross-
subsidisation from non-solar PV customers to solar PV owners. Recent reforms to introduce cost-
reflective network pricing could partly address the impacts on cross-subsidisation, but may take
some time to implement effectively.
EG connection and operation is regulated under the National Energy Rules, as well as different
state-specific legislation and guidelines. The cost and duration of the connection process affects
the investment decisions of EG proponents and their expected return on investment. It also
poses challenges for NSPs in providing timely information and connection services while
15
managing their networks in a safe, secure and reliable manner. Recent regulatory reforms6,7
have been introduced to make the connection process more transparent and streamlined.
Despite these challenges, EG provides NSPs with commercial opportunities. For example, certain
EG technologies, if managed properly, could be used to reduce stress on the network during
peak times, which would defer augmentation investments and improve network reliability and
quality of supply.
This report will review the current technical, regulatory and commercial impacts of EG to
provide an overview of the commercial opportunities available to NSPs. It also proposes a
valuation framework that can be used to gain a better understanding of the true impacts of EG
on the network. This will enable the design of appropriate pricing and incentive mechanisms to
promote future, efficient EG uptake.
6Australia Energy Market Commission, Rule Determination [ERC0147] – National Electricity Amendment (Connecting
Embedded Generators) Rule 2014, 17 April 2014
7Australian Energy Market Commission, Rule Determination [ERC0158] - National Electricity Amendment (Connecting
Embedded Generators Under Chapter 5A) Rule 2014, 13 November 2014
16
2 Scope and approach

2.1 Scope
The scope of this report includes:
 Investigating and assessing the impacts of the continued growth in embedded generation
(EG) for Australian distribution and transmission8 networks. This includes the
o technical impacts of connection, including planning, design, operation and
maintenance of networks
o technical issues for customers to establish and manage a connection to the
distribution network
o regulatory impacts of connection, including the robustness of current regulatory
instruments in meeting distribution network and customer needs in facilitating
safe connection of EG
o cost of facilitating connection, and how these costs should be recovered
o commercial impacts of connection, including the development of a high-level
commercial framework to value EG and an assessment of commercial
opportunities for network service providers (NSPs) arising from EG.
 Identifying policy and regulatory options that the Energy Networks Association (ENA)
should consider in response to these impacts, to ensure the safe, reliable and efficient
operation of the electricity network in the long-term interests of customers.
2.2 Approach
In preparing this report, Marchment Hill Consulting and CSIRO have:
 reviewed the relevant literature
 surveyed ENA member organisations
 interviewed key stakeholders from ENA member organisations
 held workshops with ENA members, and incorporated their feedback in document
reviews.
For the technical section, the literature review targeted the existing state of play of EG in
Australian networks, and was also informed by the international context. This was then
expanded to look at the likely development of the EG space over the next 5, 10 and 20-year
timeframes, referencing relevant literature as well as real-world trials, and reflecting ongoing
modelling and technology development.
8 As the definition of an embedded generating unit is ‘a unit that generates electricity at a customer’s premises and is
connected to a distribution system’, we understand that impact assessment will primarily focus on impacts relating to
low-voltage and medium-voltage distribution networks. The exception to this is in relation to the upstream impacts on
the transmission network caused by EG connected to the distribution network.
17
For the regulatory and commercial sections, the literature review was designed to inform the
picture of the current regulatory framework in Australia relating to EG connection. Further
research formed the international and Australian context for commercial frameworks to value
the impact of EG and the commercial opportunities for NSPs.
The literature review further guided the development of the survey distributed to ENA members.
The survey gathered views of ENA members about their experience of the technical, regulatory
and commercial impacts of EG on the network. Details of the survey respondents and
interviewees are in Appendix A – Survey responses.Following the survey, five interviews were
undertaken with ENA members to gather more in-depth insights into the impact of EG.
A workshop was also held with ENA members to test initial hypotheses and project findings and
to generate additional ideas for further work. All sections of the report were distributed to the
ENA and its member organisations for review and feedback. A second workshop was held with
ENA members to discuss the findings and recommendations of the draft report and to collect
final feedback before completing the final report.
2.3 Document maps

Given the breadth and complexity of the report’s findings, each section includes a concise
document map to highlight their significance and relevance to different network types. A brief
introduction to the icons used throughout is provided on the following page.
18
Document map key
The impact, constraint, mitigation option or opportunity applies at the customer

premises.
The impact, constraint, mitigation option or opportunity is relevant for the low-
voltage distribution network.
The impact, constraint, mitigation option or opportunity is relevant for the

medium-voltage distribution network.
The impact, constraint, mitigation option or opportunity is relevant for the high-
voltage transmission network.
The impact, constraint, mitigation option or opportunity is relevant for networks

with typically high impedance and limited interconnection. These networks are
most often rural.
The impact, constraint, mitigation option or opportunity is relevant for networks

with typically low impedance and greater interconnection. These networks are
most often urban.
The impact, constraint, mitigation option or opportunity is relevant for microgrids,

where there is no connection to the wider grid or there is capacity for sustained,
independent, islanded operation.
The impact, constraint, mitigation option or opportunity is of low significance given

its probable effect on network performance or business operation.
The impact, constraint, mitigation option or opportunity is of moderate significance

given its probable effect on network performance or business operation.
The impact, constraint, mitigation option or opportunity is of high significance

given its probable effect on network performance or business operation.
The impact, constraint, mitigation option or opportunity is likely to be highly

relevant to networks in the near term, probably within the next few years.
The impact, constraint, mitigation option or opportunity is likely to be highly

relevant to networks in the medium term, probably within the next 5—10 years.
The impact, constraint, mitigation option or opportunity is unlikely to be relevant

to networks in the near or medium term.
19
3 Technical impact assessment

Note on authorship
This section was written by CSIRO and reviewed by Marchment Hill Consulting to highlight links
to the commercial and regulatory implications that relate to the findings of this chapter.
3.1 Introduction
The role of embedded generation (EG) in the Australian electricity sector has grown dramatically
over the last decade, driven by falling technology costs and rising electricity prices. With
aggressive storage price projections, and Australia’s probable move towards more cost-reflective
pricing in the future, the value proposition for EG looks set to grow further in coming years.
This presents both an opportunity and a challenge for Australia’s network providers. The system
is moving ever more rapidly from a predominantly one-way bulk energy transport system to one
of distributed generation and bidirectional power flows, including intermittent renewable
generation. The demand-side also represents an increasingly strong participant.
This chapter explores both sides of this potentially dramatic shift. It catalogues the key
technical constraints and issues that may dampen enthusiasm for the change, but also captures
the very real benefits that may be delivered if suitable mitigation strategies are embraced and
appropriate standards deployed. The work draws extensively on more than 200 references to
unify and summarise findings from academic studies, industrial trials and Australian standards.
The result may not always deliver clean answers for a still-emerging field, but it does provide
the rich context and contemporary insight necessary for considered decision making.
The chapter has four main sections, each summarising key technical findings drawn from the
literature:
 The Technical Constraints and Network Impacts section highlights the prevailing and
impending issues that may limit the value proposition of EG for Australian networks.
 The Options for Mitigating Network Impacts and Risks section explores the relevant
strategies for addressing such issues, drawing on theoretical and practical studies to
assess the early strides that have been made.
 The Options for Maximising Network Benefits section underlines the potential for EG to
deliver meaningful value in the Australian context, illustrating that effectively managed
systems should be viewed as a genuine opportunity.
 The Australian Standards Gap Analysis highlights the rich suite of standards relevant to EG
and identifies where further refinement, modification or addition may aid in the delivery
of optimal EG performance.
While we consider the implementation cost of EG options in broad terms, this is not a focus of
this chapter. The commercial impacts of EG (both positive and negative) are covered in greater
detail in Chapter 5.
20
3.2 Technical constraints and network impacts
3.2.1 Introduction
Though EG presents an exciting opportunity for network and consumers alike, it also promises a
new set of constraints, challenges and issues that must be better understood and proactively
addressed.
This chapter reviews academic literature and trial findings to highlight the hurdles that may
inhibit the potential of EG for the Australian energy sector, with a focus on power quality,
protection, network stability and power regulation impacts.
The document map presented across the following two pages provides a précis of the key issues
identified within this chapter and summarises where, when and how significantly the impact may
be felt. Taken together, the impact and timing indicators provide a preliminary assessment of
risk. For readers interested in only those constraints and impacts that apply for particular
network types, or only those issues that are considered most pressing, the map will guide you to
the most relevant sections of this chapter.
21
Document map
Technical
Chapter
constraint or Network level Network type Impact Timing
section
issue
Power regulation
3.2.2.1
(energy balance)
Power regulation
3.2.2.2
(ramp rates)
Power regulation
3.2.2.3
(thermal loading)
Power regulation
3.2.2.4
(losses)
Power regulation
3.2.2.5
(load prediction)
Power regulation
(bidirectional power 3.2.2.6
flow)
Protection
(network fault 3.2.3.1
currents)
Protection
3.2.3.1
(network fault level)
Protection
(fault current limits 3.2.3.1
for microgrids)
Protection
(fault detection in
3.2.3.1
self-islanding
microgrids)
Protection 3.2.3.2
(bidirectional power
flow)
Protection
(unintentional 3.2.3.3
islanding)
Protection
(impedance relay 3.2.3.4
operation)
22
Technical
Chapter
constraint or Network level Network type Impact Timing
section
issue
Network stability
(power station 3.2.4.1
voltage stability)
Network stability
(power station
3.2.4.1
frequency and rotor
angle stability)
Network stability
(circulating reactive 3.2.4.2
power)
Power quality
(voltage
3.2.5.1
intermittency and
voltage rise)
Power quality
3.2.5.2
(voltage imbalance)
Power quality
3.2.5.3
(voltage flicker)
Power quality
3.2.5.4
(harmonics)
Power quality
3.2.5.5
(voltage regulation)
23
3.2.2 Power regulation
What is it?
Power system regulation is a collection of services that adapt in real or near real time to
unanticipated system changes. Such services include automatic generation control, frequency
droop and voltage regulation.
What role does embedded generation have?
The need for generation to occur at the time of use — coupled with constant and often
unpredictable changes in demand — makes electricity supply systems large, dynamic and
complex, and their regulation a challenging task. Historically, distribution networks have been
regarded as a passive termination of the transmission network, having the goal of supplying end
users reliably and efficiently. The mix of a greater penetration of EG in electricity networks will
likely result in a gradual, but inevitable, change of distribution networks towards a new kind of
active network. The requirement for power regulation is increasing, due to a larger proportion
of the electricity supply consisting of variable and unpredictable EG.
Why does it matter to Australian network operators?
Australia has seen a significant increase in the capacity of EG in the last few years. This is a
continuing trend, driven by factors such as government incentives, increased electricity costs
and reduced manufacturing costs of EG systems. An example of this can be seen for the
increased rate of solar photovoltaic (PV) capacity in Australia, as shown earlier in Figure 1.
When assessing the impacts of increased EG penetration, the key aspects requiring consideration
include voltage profiles, regulation, electrical losses, power factor, capacity planning, power
quality, system operations and protection. A core operational and market function of electricity
network operators is to forecast system load, and then to ensure that sufficient generation and
non-generation resources are committed such that intra-hourly deviations can be accommodated
by those resources. These deviations can take place in the upward or downward direction, and
are currently largely caused by changes in load. With increased penetration levels of variable EG
sources, the net load-following requirement could increase substantially in certain hours, due to
the variability and forecast uncertainty of wind and solar production.
The remainder of this section provides background information about the impacts of increased
EG on regulation aspects of energy balance, ramp rates, thermal loading, losses, load prediction
and bidirectional power flow: all of which are becoming more dynamic, instead of being limited
to a single, peak-demand-driven constraint.
Technical impact assessment | 24

3.2.2.1 Energy balance
Network level Network type Impact Timing
In any power system, there has to be a reliable continuous balance between production and
consumption, and enough resources must be available to counterbalance changes in production
and consumption (i.e. load demand). This balance of energy production and consumption is
known as the energy balance.
Historically, conventional generation sources, such as coal or gas-fired power stations, have
been controllable, dispatchable and large-scale, with a dynamic wholesale market driving least-
cost and reliable supply of electricity. Each generation unit is dispatched and scheduled
according to its heat rate, fuel cost and availability, associated transmission losses and output
ramp rate, with the goal of reliably satisfying electricity demand at the lowest possible cost [21]
.
The availability and quantity of electricity produced by conventional generation sources can be
controlled by system operators, together with the generators’ governors and electricity system-
wide automatic generator control (AGC) system or similar. This stands in sharp contrast to solar
and wind outputs, in which weather variations may lead to highly intermittent generation. The
inclusion of such technologies means that conventional generators must follow not only the usual
demand variations, but also account for the output variations caused by intermittent generation.
Five normal functions of generation operations that could be impacted are: load-frequency
control, load following, ramping rate, unloadable generation and operating reserve.
Load-frequency control
The amount of load and generation at any given time needs to be matched to maintain the
system frequency at the desired level (50 Hz in Australia). When load exceeds generation, the
system frequency will drop. When a change in the system frequency is detected, power system
operators and AGC/governor interactions will increase or decrease the output of conventional
generators to match the load.
Intermittent renewable generation technologies generally cannot participate in these system
frequency regulations, because their output is already maximised based on available renewable
resource. Output control is therefore typically limited to curtailment. Thus, as the penetration
of renewable EG systems increase, the capacity to efficiently manage system frequency
decreases. The current draft of AS/NZS 4777 Part 2 supports this in compliant inverters.
Load following
If an increase in solar or wind power is not coincident with the system load increase, other
generating units in the system will have to be offloaded to absorb all the solar or wind power.
When solar or wind power production falls, the output from other units will have to increase to
take up the generation slack. Utilities normally use intermediate plants to follow the load. The
integration of high-penetration, intermittent EG may therefore increase load-following duties for
the conventional generators assigned for system regulation.

Ramping rate
Ramping rate represents the generator’s ability to change its output over time. The ramping rate
of online generators may need to be increased to follow the sharper system load changes
brought about by increased intermittent generation. Some conventional generators may not be
used if they cannot operate at higher ramp rates; for those can, the increased ramp-rate
requirement may reduce equipment life or decrease operation efficiency. The increased ramp
rate requirement may necessitate more online peaking plants, which could raise wholesale
electricity prices.
Unloadable generation
The down-ramping rate of a generator may differ from its up-ramping rate, both of which are
important to meet the normal system load-following requirement. The amount of generation
that can be offloaded (down-ramped) is called unloadable generation. To accommodate the
maximum output from intermittent generating technologies, system operators have to make
certain that online conventional generators can be backed down quickly enough: particularly
when simultaneously facing a sudden increase of intermittent generation output and a system
load increase. Such an accommodation to absorb energy from intermittent generation cannot be
made by tripping off a unit, because it may be needed again shortly after being taken offline
[22] . Note that this can be readily achieved by curtailing the intermittent generator.
Operating reserve
The impact on the electric system operating reserve is also related to the intermittency of solar
and wind generation technologies. Utilities carry operating reserve to guard against sudden loss
of generation and unexpected load fluctuations. Any load and generation variations that cannot
be forecast have to be considered when determining the amount of operating reserve. If utilities
cannot predict the short-term fluctuations of intermittent renewable generation sources, more
operating reserves must be scheduled to adequately regulate the system. This requirement will
increase the cost of integrating intermittent EG sources such as solar and wind [22] .
Together, the above five impacts imply that more units may need to be brought online or put on
regulating duty, which may increase system operating costs. Conventional generators may be
forced to be more flexible, with their output resulting in a higher per unit cost. Adequate system
flexibility is a key requirement for managing increased levels of EG on the electricity network.
Several studies have looked at the likely impacts of increased levels of intermittent, renewables-
based EG. A study that modelled net load (total load minus EG output) for different penetration
levels of solar PV in the California Independent System Operator network showed significant
impact on minimum net load [23] . The load duration curve shown in Figure 2 shows original
(without solar PV) net load and the predicted impact of incorporating solar PV at 10, 30 and 50%
penetration levels. Some reduction in peak demand is seen (far left), and the 30 and 50%
penetration scenarios have a large impact on minimum net load. This may affect the generation
portfolio by either driving down generation loading or reducing generation run-times. Both of
these affect plant efficiencies and equipment lifetimes, and ultimately, economic viability.
Potential mothballing of systems here would fundamentally affect the capacity and flexibility of
the generation mix.
As an example, the standard United States fuel mix has a large proportion of coal-fired
generation. The rate of power change of coal-fired generation is largely limited by thermal
inertia and unit-specific, fuel-system limitations. This makes them less flexible than gas-fired

generation, for example. If the 30% solar PV penetration load duration curve is used for the
United States fuel mix, an incursion into coal-fired generation occurs as shown in Figure 3. This
then raises questions of generation flexibility, and how coal-fired generation will manage
requirements to reduce output and allow further integration of EG. One potential response is to
replace cheaper, less flexible plants with expensive, more flexible plants [23] . The Australian
fuel mix for electricity generation in Figure 4 [25] shows that about 70% of all electricity
generated comes from coal-fired generation. Therefore, the same requirements for the
United States fuel mix scenario may apply in Australia. Note that the Australian fuel mix for
electricity generation in Figure 4 does not include EG or non-grid private generation, but
includes generation from semi-scheduled and large, non-scheduled intermittent generators.
Figure 2 – Predicted load duration curve, California Independent System Operator July 2007, with 10, 30 and 50% solar
photovoltaic penetration [23]
Figure 3 – Dispatch order for United States fuel mix with 30% solar photovoltaic penetration [23]

Figure 4 – Principal electricity generation in Australia by fuel type, 2013—2014 [25] (Reproduced courtesy of the
Energy Supply Association of Australia)
3.2.2.2 Ramp-rates compared with conventional generation
Ramp rates of conventional generation, which are a potential impact caused by increased levels
of EG, were briefly discussed in Section 3.2.2.1. They are covered in more detail in this section,
including discussion of increased variability in load when intermittent, renewable-based EG is
integrated.
At a high penetration level, the ramp rate of intermittent EG output can cause adverse impacts
on an electricity network. One of the main challenges to the power system is correlated to the
instantaneous penetration of intermittent EG: that is, the fraction of total system load provided
by the EG source at a given instant in time [22] . Several studies have looked into net load with
varying penetration levels of intermittent EG [22] and investigated net load variability due to
solar generation. Solar generation is viewed as negative load. When it is combined with the
system load, it yields a net load corresponding to the power that must be supplied by other
sources in the system. When load demand and solar power are both increasing or decreasing
coincidentally, the need for other generation sources to vary their output will decrease.
However, when load and solar power move in opposite directions at the same time (e.g. when
load is decreasing, while solar power output is increasing), the large variation in output level

required may significantly affect the operation and performance of traditional generation
plants.
The existing electricity system already incorporates significant variability in load demand. This is
managed through generator dispatch and ancillary service mechanisms. However, as the
penetration levels of intermittent EG increase, further measures may be required (e.g.
additional ancillary services). A study on the net load variability in the Californian grid was
reported in [24] . In this study, hourly change in total load (delta) for load-only and net load
(i.e. load minus renewable generation) was analysed for a model of the Californian grid assuming
33% penetration of renewable energy: in this case, wind and solar. Figure 5 illustrates the hourly
change for different loading conditions, with the data split into deciles. The first decile is a
measurement of delta when load is between 90 and 100% of peak load (top 10% of peak-load
hours), while the 10th decile is a measurement of delta for loads up to 10% of peak load. The
thin lines above and below the bars represent the standard deviation of the positive and
negative deltas, respectively. Incorporation of renewables increases the hourly net load
variability across the majority of deciles. For example, the standard deviation of net load
variability in the 10th decile increases by 47% with the integration of 33% renewable energy.
Figure 5 – Predicted hourly net load variability in the Californian grid with 33% renewable penetration. (L-W-S = load
minus wind minus solar) [24] (Reproduced courtesy of the California Energy Commission)
This analysis was taken further to examine the net load variability for each hour of the day, the
distribution of which is shown in Figure 6. At 6:00 am, the maximum and minimum deltas are
similar, but the average hourly variation increases from 2000 to 2500 MW when incorporating
33% penetration of renewables. During the afternoon peak hours, the averages are similar, but
the maximum and minimum deltas are significantly larger for net load with solar and wind
energy integrated. The standard deviation is also noticeably larger for net load, increasing from
approximately 1300 to 1600 MW (about 23%).

Figure 6 – Hourly net load variability by hour of day. (L-W-S = load minus wind minus solar) [24] (Reproduced courtesy
of the California Energy Commission)
A recent study [26] investigated the impacts of solar radiation variability on PV power output in
the existing Alice Springs electricity network. The aim was to obtain an estimate for the
maximum penetration of grid-connected solar power generation that could be integrated
without energy storage. The study took into account existing power stations, both conventional
and renewables-based, and aggregate load on the Alice Springs network.
The load demand in Alice Springs varies substantially across the year, with daytime maxima
ranging from 27 MW (winter peak) to 51.5 MW (summer peak), and the minima ranging from
18 MW (winter low point) to 23.2 MW (summer low point). Alice Springs has three main
centralised power stations, incorporating 19 dual-fuel gas/diesel generators with a combined
generation capacity of approximately 100 MW. Existing installed solar PV capacity is around
4 MW, and consists of residential and commercial systems as well as a 1-MW PV plant. The net
load variability for the whole Alice Springs network was modelled for three different scenarios:
• existing load demand with 10-MW single PV array
• existing load demand with three sets of 3.3-MW PV arrays
• existing load demand with nine units of 1.1-MW PV arrays.
The net load variability was the highest for the single 10-MW PV array scenario, which was
significantly reduced for the nine units of 1.1-MW PV arrays scenario, due to the geographical
dispersion of the PV arrays. The study found that the level of variance for the single 10-MW plant
was more than four times that of nine units of 1.1-MW plants separated geographically around
Alice Springs, and that the ratio was very similar for January (maximum load) and September
2014 (minimum load). The study was carried out in the context of the entire Alice Springs
network, and localised network impacts of EG due to other grid constraints, voltage rise and
frequency stability were not investigated. Another study [27] reported observations of close to
1-MW net load variability in Alice Springs (installed solar PV capacity of 2.1 MW) over periods of
minutes due to cloud-movement-induced solar PV fluctuations.
To avoid any adverse impact on network operation, we conclude that the increased net load
variability caused by higher penetration of intermittent EG must be taken into consideration and
investigated further for network planning and design.

3.2.2.3 Thermal loading
Most equipment in an electric power system has a thermal limit. The thermal limit can be
defined as a limit on the power carried by an electric power system that results from the
heating effects of the power carried by the devices.
Ageing network assets, along with increased accommodation of intermittent EG, can potentially
increase system stresses in active distribution networks. Traditional security assessment
incorporates deterministic (static) thermal limits of assets, which limit the extraction of hidden
impacts of variations in thermal capacities. Because dynamic thermal limits and variations in
weather conditions are aperiodic (irregular) events, their detailed modelling is beneficial for
determining the latent capacity of an active distribution network.
New stochastic (random) models are proposed in [28] to capture dynamic thermal limits of
assets, change in weather conditions, and their aperiodic patterns. The results suggest that
stochastic variations in weather patterns can affect the security of supply to customers
considerably more than can the effects of dynamic thermal limits. The study also mentions that
the change in weather patterns can considerably increase the costs of outages. Determining the
actual latent capacity of assets due to dynamic effects and its utilisation does not necessarily
reduce the impacts on security of supply in an active distribution network. The effectiveness of
modelling dynamic thermal limits to determine impacts is more beneficial and necessary in
stressed networks than in moderately loaded networks.
As an example, the United Kingdom generation profile is expected to change significantly in the
future; an increased penetration of wind power is anticipated to increase the share of renewable
energy sources to 20% by 2020. This will result in the need to transfer large amounts of
renewable energy from the north of Scotland to the mainland demand centres in the south. The
transmission capacity in the existing network cannot accommodate the increased power
transfer. The United Kingdom system operator and owners must therefore maximise the use of
existing transmission lines closer to their thermal limits to avoid constraining some generation
plants and to improve the stability limit. To enable this, a hierarchical framework for the future
stability control system has been proposed to improve stability and push the stability limit
beyond the static thermal limits [29] .
An interesting example of a customer asset that can cause thermal loading issues is a plug-in
electric vehicle (PEV). While not typically considered EG, electric vehicles (EVs) are highly
correlated loads that can also discharge to the grid through vehicle-to-grid (V2G) capability. A
study in the United States has identified the local distribution network as a likely area to be
adversely affected by unregulated PEV and EV charging, due to the characteristics of their
electric power generation [30] . The study looked at the impact of PEV charging on a local
distribution transformer insulation life, which is mostly affected by the hot-spot winding
temperature. A transformer model developed in the study was used to estimate the hot-spot
temperature, given the knowledge of load ratio and ambient temperature. Additionally,
different penetrations of PEVs were studied from the transformer insulation life aspect. The
results indicated that a high penetration of PEVs can significantly affect the power grid,
including potentially reducing the lifespan of transformers: particularly in the case of poor
coordination of charging times. Conversely, low penetration of PEVs is not detrimental to
transformer life, especially if charging is coordinated to some extent.
The integration of distributed EG in low-voltage (LV) networks was studied in [31] , which
studied the thermal loading impacts of both the transformer and the line connecting the
substation busbar to the first terminal along the feeder. Deterministic analysis of two limit cases
(units located at the furthest or nearest households to the substation bus) indicated that the
thermal loading of the transformer reached 100% of its rated value with an EV penetration level
of approximately 25%. The individual phases exceeded their rated loading capacity for EV
network penetrations of approximately 23—30%. This indicates that the thermal loading of
network components must also be considered as a barrier to the number of EVs that can charge
simultaneously on a particular network. However, the stochastic analysis (random location of
EVs, weather uncertainty and time-varying load profiles) revealed that the cable exceeded its
maximum rated loading less than 2% of the time with loading levels of 127% for 50% EV
penetration. This suggests that the distribution of EVs and associated charging profile in
distribution networks can have a profound impact on thermal loading, although the specific
impact in a given case will be heavily influenced by the network design and other contextual
considerations. The probability was the same for both winter and summer. The thermal loading
of transformers due to high penetrations of EV may be kept low by using off-peak or controlled
load charging mechanisms.
3.2.2.4 Losses
The transmission of electrical energy from power generation plants to consumers or end users
via the transmission and distribution networks is usually accompanied by voltage drop and power
losses. Power losses in electricity networks are caused by active and reactive power flows
through line resistances and transformer impedances. The magnitude of active power loss
depends on the amount of current flow and the line resistance, while reactive power losses can
be defined as the reactive power that is absorbed by network components such as line
inductance or transformers. Hence, reactive power must be supplied by generation in addition
to the reactive power consumed by network loads. These losses are inherent in the utility grid
system and can be minimised by controlling volt-amps reactive (VAR) sources, such as capacitor
banks and EG. One of the challenges in using EG to reduce or minimise losses in the electricity
network is the optimal sizing and placement of the generators in the electricity distribution
network.
Several algorithms have been developed [32] [33] for optimal allocation and sizing of EG in the
distribution network. While EG can introduce problems to the electricity network, it also has
benefits that need to be explored further.

3.2.2.5 Load prediction
When a significant amount of intermittent renewables-based EG is incorporated into a power

system, appropriate planning and management of load demand is essential to ensure adequate
power quality and reliable electricity supply. As seen throughout this report, the potential
mismatch between intermittent generation output and load may lead to inefficient operation of
conventional power systems, or otherwise necessitate renewable generation curtailment.
Accurate forecasting and prediction of load demand is essential for effective use and
management of EG sources on the network, and also to establish sustainable load management
systems for the smart grid.
Load prediction is broadly classified in the context of the following timeframes [59] :
 long-term forecasting (1—20 years), for applications in capacity expansion and long-term
capital investment return studies
 medium-term forecasting (1—12 months), used to prepare maintenance scheduling and to
plan for outages and major works in the power system
 short-term forecasting (1—4 weeks ahead), required for operation planning, unit
commitment and economic dispatching
 very short-term prediction (1—7 days ahead), used for load exchange and contracting
with neighbouring networks and to maintain a secure power system.
In Australia, operational generator dispatch and frequency control requires very short-term to
short-term load forecasts (next five minutes, 1—2 hours, day-ahead and week-ahead). Long-term
forecasts (year to years-ahead) are required for network planning and bulk generation
(centralised or distributed) planning. Accurate load prediction across all timeframes is very
important for appropriate investments, and hence the changing and challenging aspects of load
prediction stated earlier in this section would need to be considered.
For most of the eastern states of Australia, the National Electricity Market (NEM) provides a
central dispatch mechanism that adjusts power supply to meet load demand through the
dispatch of generation every five minutes. Shorter-term to instantaneous generator output is
regulated by the generator governors under AGC or similar systems, with some generators
assigned load-following duty to maintain system frequency.
In the Australian NEM, the Australian Energy Market Operator (AEMO) measures electricity
demand by metering supply to the network, rather than metering consumption. The advantage
to this method is its inclusion of electricity used by customers, network losses and the energy
used to generate the electricity (i.e. auxiliary loads). A high-level topology of the Australian
electricity transmission network that connects supply (generation) and demand (customers) is
shown in Figure 7. This figure also shows the different points at which various types of
generation are connected, and the points at which supply and demand are measured. As seen in
Figure 7, small-scale EG are attached to both transmission and distribution customers, which

reduces the amount of electricity that needs to be supplied by large-scale generation. For
example, a large increase in residential rooftop solar PV systems would reduce load demand.
Similarly, energy efficiency and load control initiatives would also reduce load. On the other
hand, an increased uptake of EVs or PEVs may increase the demand at customers’ locations, with
the potential of new peak demand created if EV charging times are not appropriately
coordinated. An increase in the uptake of battery storage systems and customers controlling
their loads at times of high prices would also contribute to the challenge of predicting customer
load demand, with the difficulty increasing when these activities are more widespread.
However, the growing use of smart meters may improve the ability to forecast the level of
customer load demand.
Power systems operate by ensuring the amount of load and generation at any given time is
matched in order to maintain the system frequency at the desired level. In scenarios with high
penetration of small-scale EG, the energy supplied by these generation sources would need to be
subtracted when calculating the energy supplied by generation controlled through the NEM
dispatch process (i.e. the scheduled and semi-scheduled generation sources). Reliable operation
of any power system and minimisation of impacts on the normal functions of the generation
operations (see Section 3.2.2.1), relies on accurate prediction of load demand. This is essential
to minimise operating spinning reserve and load-following generation capacity in excess of the
load. Forecasts would need to take into account small-scale EG as negative load, as well as all
other initiatives that are likely to change the load demand profile (e.g. battery storage systems,
time-of-use electricity tariffs, demand-side participation).
Figure 7 – Electricity network topology [37]

Internationally, the International Council on Large Electric Systems Working Group C1-24 state
that the increasing installation of EG in distribution networks, along with the changing
characteristics of that generation (particularly rooftop solar PV), are changing the utilisation
patterns of electricity networks. This is now becoming a global issue [38] .
Changes required to load prediction include:
 producing accurate minimum demand and energy consumption forecasts, in addition to
traditional peak demand forecasts
 making long-term forecasts more granular (on at least an hourly basis, possibly shorter)
 understanding the price elasticity of demand as it relates to demand-side response
 differentiating the energy consumed by customers from the electricity generated by their
own EG
 understanding the relationship between energy tariffs and customer behaviour and the
impact on network revenues.
Issues that make accurate load prediction challenging include the:
• changes to customer behaviour in response to electricity prices, and the availability of
low-cost EG systems such as rooftop solar PV arrays
• limited information available to system operators from beyond customer meters
• impact of government policies that encourage energy efficiency
• government and regulatory policies on tariffs, including FiTs
• government policies facilitating EG in distribution networks
• uncertainty in the economy.
The capital expenditure (capex) required to augment the capacity of the network depends
heavily on peak demand forecasts. Such forecasts are a large part of the capex programs in the
current regulatory control period of network service providers (NSPs).
As a percentage of total capex, demand-related capex is around 25% for distribution network
Ergon, but can be much larger for transmission networks: approximately 60% in one case [60] . If
demand growth exceeds the forecasts reflected in the regulatory determination, NSPs may need
to incur additional capex. Otherwise, if actual demand turns out to be less than that forecast,
expenditure might be set too high. A large proportion of capital outlay is related to asset
replacement and customer connections (distribution network service providers; DNSPs), rather
than to meet increases in peak demand [60] . If there is a significant growth in demand-side
participation, energy efficiency, individual onsite EG and uptake of cost-reflective pricing (e.g.
time-of-use tariffs), the proportion of capex for demand-driven augmentation could decline. An
increased addition of EG and time-of-use tariffs could potentially change customer demand
profiles significantly. This would need to be taken into consideration when forecasting load
demand.
Key Finding: Increasing penetration of EG, including greater uptake of rooftop solar
and other EG, is changing load profiles on Australian electricity networks and making
network load prediction more challenging. Appropriate planning and management of
network load demand across different timeframes is essential to ensure both adequate
power quality and a reliable supply of electricity of the network.

3.2.2.6 Bidirectional power flow
With the recent rise of EG — and in particular, residential rooftop solar — in many parts of
Australia’s energy networks, bidirectional power flow has moved from being a topic of research
curiosity to drawing the attention of Australian distributors. For example, Ergon noted in 2012
the ‘potential for reverse flows into the high voltage network as penetrations increase’ [14]
Energex has gone a step further, indicating that this has already occurred within their network
during times of peak solar generation [15] .
Several issues relate to bidirectional power flow in a radial distribution (and potentially
transmission) network. Of these, the highest immediate impact is likely to be the effect of
reverse power flow on voltage regulators such as step voltage regulators (SVRs), which are more
prevalent in rural networks than in urban networks. SVRs maintain voltage on a network segment
within specified tolerances. The specific operating modes of SVRs and the inherent limitations of
each are described in [16] and summarised below:
 Normal bidirectional mode
o this simplistic control strategy can include a flaw wherein reverse power flow can
cause the transformer to sense a high-voltage (HV) condition on the supply side
o this causes it to tap in the wrong direction, resulting in the opposite outcome than
intended and likely a major over-voltage on the LV side
 Co-generation mode
o supports reverse power flow without critical impacts
o any generation on a regulated line will cause the SVR to operate suboptimally due
to incorrect estimation of the regulated voltage point
 Reactive bidirectional mode
o can address most issues of reverse active power flow, but may be susceptible to
the same issues as normal directional mode when reactive power flows in the
reverse direction
o currently of limited concern, except on network segments with high capacitive
loads (e.g. shunt capacitor banks for power factor regulation), but may become
more significant as smart inverters or static VAR compensators (SVCs) perform
more volt-VAR control.
In addition to the issues relating to control strategy, the ability of SVRs to regulate voltage
correctly can also be affected by reverse power flow. To maintain voltage within the specified
range, an SVR must estimate the voltage along the length of line it is managing. This estimation
is typically formed through line drop compensation, and is based on the current at the SVR and
the known per-unit-length resistance and reactance of the line. The current through the line is
assumed to be proportional to the current at the SVR and the line length between the SVR and
the voltage regulation point. Significant generation in the line can invalidate this assumption,

causing the SVR to incorrectly detect a light-load condition. This can lead to poor voltage
regulation and deviations from the intended voltage range at the load connection point.
Some Australian DNSPs have provided anecdotal evidence stating that in rare circumstances on
some extreme network segments, the current peak is no longer caused by maximum demand,
but rather by maximum generation. This is due to the high penetration of strongly correlated
rooftop solar PV, in particular.
Beyond this, in smaller remote-area or off-grid networks, reverse power flow that is not
managed before reaching the generator can significantly affect the control and operation of the
generation unit, including over-excitation, over-voltage and frequency destabilisation [19] .
Other concerns relating to bidirectional power flow include the need to upgrade supervisory
control and data acquisition and other load-monitoring equipment to account for it where it was
not originally included in system planning.
Historically, electricity networks were designed for unidirectional flow of power from source to
load. However, EG can also result in power flow from load to source. This can adversely affect
network voltage and protection systems [34] . For example:
• In the presence of EG on the network, fault current is supplied both from the power
system and the EG(s). This can increase the short-circuit current, which can exceed
breaker capacity.
• Based on the location of the EG and fault, the power flow from the EG(s) can reduce
feeder current from the substation, which may decrease relay sensitivity to operate
during fault conditions.
• Non-optimal voltage control might reduce the voltage profile along the medium-voltage
(MV) feeder.
The protection-related issues of reverse power flow are detailed in the following section.
To integrate and accommodate high penetration levels of EG on distribution networks,
protection coordination is required that can sustain bidirectional power flows. At present, some
local supply authorities are limiting the maximum power of PV systems, which is one way of
reducing the possibility of net reverse power flow. Local DNSPs have observed reverse power
flow in their networks, some of which have resulted in voltage rise.
Some modern grids are designed to accommodate reverse power flow. For example, reverse
power flow is limited to 60% of the transformer nominal rating in Ontario, but is more typically
limited to 30% [35] . In a study of three Swedish distribution networks, the duration of reverse
power flow ranged from 0.1—3.5 hours a day through the MV/LV transformer substation when all
properties had a 1-kWp solar PV system installed [36] . When larger, 3-kWp systems were
installed on each property, the daily reverse power flow duration ranged from 1.1—8.9 hours.
In the next 20 years, the increasing prevalence of embedded solar generation may exacerbate
the above problems. This is noted in the Future Grid Forum’s scenario modelling [20] , which
indicates that by 2050 between 18 and 45% of energy may be supplied by onsite generation:
including a significant portion of distributed solar. Even the high end of this range may be
considered a conservative forecast for some networks’ for example, Energex already has 1 GW of
connected PV on a network with a combined peak of 5 GW, and government policy could see this
double by 2020. In such scenarios, the NEM, which is dominated by networks that operate on
near-identical time-zones and hence sunlight patterns (cloud cover notwithstanding), could be
expected to have a highly correlated solar resource. Such a time-correlated and intermittent
source may cause significant reverse power flow into upstream and HV networks unless carefully
managed.

Key Finding: Reverse power flow has been observed at the medium-voltage and high-
voltage network levels. In radial networks designed for unidirectional power flow, this
can affect the operation of protection and voltage regulation devices. In islanded
networks, such as some remote communities, it can also have a detrimental effect on
generator control and stability.
3.2.3 Protection
What is it?
Network protection is the switchgear and other equipment installed and operated to protect the
electrical system from faults and the plant from damage. It ensures the safety of workers and
the public while maximising the reliability of the network.
EG can affect several aspects of protection, from production of excess supply leading to reverse
power flow through protection elements to inadvertently supplying voltage in the absence of a
grid connection, which has significant safety implications.
Australian DNSPs have already seen some issues with protection schemes from existing EG. This
is only expected to increase as EG continues to penetrate networks.
Traditional radial networks were generally designed to facilitate unidirectional power flow under
normal system conditions. The increasing penetration of EG in distribution networks adds
complexity to the power flows within a network, by introducing issues such as bidirectional
power flow and variations in the scale of fault currents. In turn, these issues can complicate the
protection requirements for these systems.
As detailed below, several investigations in recent years have identified issues related to
increasing EG penetration. These include:
 current flowing in a network experiencing a fault condition, which may necessitate
review of protection settings
 reverse power flow in some networks, which can affect the operation of existing
protection schemes
 poorly controlled or mutually supporting local generation, which can retain a voltage
during a grid failure, leading to an islanded grid with potential safety concerns
 downstream generation affecting the operation of impedance relays, which rely on
accurate knowledge of local network impedances.
3.2.3.1 Fault detection in high-embedded generation systems
In a typical network based on centralised generation, fault detection is performed

predominantly through the use of either fuses (typical on LV networks), or over-current
protection relays, which trip circuit breakers. These trigger quickly on a high-current event, and
more slowly for lower current faults, to protect the cabling on that network segment.

When a short-circuit fault occurs, the impedance between the active and the neutral, or
between two actives, is reduced by several orders of magnitude. This greatly increases the
current flowing through the active. A low-impedance fault will generate a current that far
exceeds the standard load current of the network; when this occurs, the protection device will
activate and interrupt the current flow before any significant damage is caused. Various
schemes, such as sensitive earth fault and distance protection, can detect faults of magnitudes
less than load current and initiate trips after a longer time, which is required to ensure that a
fault exists.
EG can detrimentally affect the currents flowing in a network under fault conditions in several
ways, as detailed below.
Contribution of embedded generation to network fault currents
Low-impedance faults in distribution networks have traditionally been detected and eliminated
through the use of over-current relays and similar protection systems. Some concern has been
noted by industry [1] and in academic literature [2] about the capacity for EG to provide
additional fault current, such that the embedded generator effectively masks the fault from the
protection system, leading to the protection system failing to operate. However, there is limited
evidence of this occurring in actual network systems.
Reference [3] proposes that faults that draw current from both a grid supply and EG should not
cause protection devices to fail. They claim that the impedance seen between the fault and the
wider grid is not changed by the EG, and so the fault current through the grid-side protection
device should not change significantly. This is true for over-current protection. However,
distance protection must be set to take account of the remote end of EG, as it will hold up the
voltage and effect distance relay operation.
Reference [5] indicates that inverter-coupled systems are generally only able to supply
relatively small currents, even under short-circuit conditions. The authors found that these
currents are typically around two to four times the rated current of the inverter. In strong urban
or central business district (CBD) MV networks, this is relatively small in comparison to a typical
fault current. However, in weaker networks (e.g. long rural or single-wire earth return (SWER)
grids), network impedances may be high enough that protection systems need to be designed to
account for the more significant input from EG. In these cases, it is possible that inverter-
coupled EG may supply enough fault current such that current at the fault is high, but protection
devices do not trigger. In particular, the devices should not trigger within the time limits
required by AS/NZS 3000 and related standards for fault currents at the fault that would
normally cause a trip.
Synchronous and induction machines and transformer-connected systems can also contribute
fault currents significantly higher than inverter-connected systems ([5] [6] ). This is because the
fault current that can be supplied by these generators is limited only by the generator or
transformer impedance, and so may be sufficient to compromise some fault-detection schemes
if settings are not adjusted.
Network fault level
Many fault-detection schemes will also have a maximum fault current that can be interrupted
without the protection device (typically a circuit breaker) failing. This is sometimes termed the
maximum or ultimate short-circuit current. The maximum short-circuit current rating of devices
on a network is collectively known as the network design fault current for that segment of
network. This has been identified as an issue for a number of Australian networks; in particular,
Victorian DNSPs have identified this as a barrier to incorporating EG in CBD and commercial
networks.
If one or more EG units are connected within a network, the fault current will be the combined
currents of both the grid inertia and the EG. In the case of high fault current generation, such as
synchronous or induction machines and transformer-coupled inverter systems, the protection
device may exceed its maximum operating current and fail to operate correctly. This is why
embedded generators may not be allowed to connect to a network that is already approaching
its maximum design fault current.
Fault current limits on off-grid and microgrid systems
In the case of both off-grid and grid-connected microgrid systems, a protection system must still
interrupt faults, regardless of the presence of grid inertia to supply large fault currents. This has
proven challenging in the case of purely inverter-based systems [7] .
Solutions that have been proposed to this problem include incorporating energy storage
elements, such as flywheels, purely for the purpose of supplying large fault currents [9] , and
using machine intelligence techniques, such as artificial neural networks, to detect electrical
faults by performing pattern recognition on system states [10] . Such techniques also show
promise in differentiating between short-circuit faults and inrush currents on motors and
transformers [11] . This is still very much an open area of research.

Fault detection in self-islanding microgrids
Assessment of fault detection thus far has assumed that subnetworks including EG are essentially
of static design, and can be either grid-connected or islanded, but not both. However, there is
an emerging trend [12] of grid-connected microgrids that can change to isolated operation in
the presence of grid outages, increasing the continuity of supply.
In these ‘self-islanding’ microgrids, any internal protection mechanisms must be able to detect
and isolate faults reliably in both grid-connected and islanded modes of operation. When grid-
connected, the traditional methods of fault detection as described above should suffice,
typically through identification of low-impedance faults with over-current devices. However, the
limitations on the supply of fault current in off-grid microgrids will equally apply when the
microgrid is islanded. An ideal protection system could operate correctly in both situations.
While some work [13] has been undertaken in this area (in this case, using phase imbalance and
the thermal properties of the protected conductors to help identify faults), it remains an open
area of research.
Note that most Australian DNSPs put strict requirements on the operation of temporary islands in
or connected to their networks, and many do not allow temporary islands to operate at all.
3.2.3.2 Impacts of bidirectional power flow on protection systems
The degree to which bidirectional power flow affects network operations — and in particular,
protection mechanisms — is strongly tied to the type of network experiencing the effects. During
the Smart Grid Smart City project, Ausgrid trialled a significant penetration of EG on an urban
feeder in Newington. They did not identify any significant events relating to the feeder
protection. On the other hand, they noted that high-penetration EG could lead to reverse power
flow, which could significantly affect rural networks [18] : particularly in the area of reclosing
circuit breakers without having extinguished the fault, as described in Section 3.2.2.6.
Reverse power flow in a radial network can also trip some types of protection devices that are
specifically designed to protect unidirectional generators from upstream power flow [17] .
However, this only relates to types of relays that are rarely used in Australian networks, and as
such may not be a significant issue here.

3.2.3.3 Unintentional islanding
A concern with high-penetration EG, and particularly grid-connected microgrid systems, is their
potential to remain in operation in the event of a grid failure, and thus provide a voltage signal
for load (and other generation) equipment to connect to. This is a critical issue for DNSPs,
because the equipment may continue to energise the network while being worked on by
personnel, greatly increasing the risk of electric shock and arc-flash. There are two instances of
working on the power system: either ‘live line’ or ‘dead’ under a permit system. The potential
problem occurs in the latter, when a worker may assume the local power system is ‘dead’,
because they have undertaken isolation from the feeder supply points, but EG may still be back-
feeding. This risk can be addressed through standard operating procedures (e.g. testing dead,
applying earths, wearing personal protective equipment).
These situations are typically avoided by ensuring that EG is coupled to the network via a grid
protection device. This will detect when the main supply is de-energised, and disconnect the
generation (and potentially the entire microgrid, in the case of self-islanding systems) from the
grid. For example, the strict disconnect requirements in AS 4777.3-2005 require that inverter-
coupled systems cease energising their AC output within two seconds, and remain offline until a
stable grid supply has been detected for at least one minute. In addition, inverter-coupled
systems must also include at least one form of active islanding detection, whereby an inverter
will attempt to force a change in grid parameters; the degree to which this succeeds will
identify whether the system is grid-connected or off-grid. This reduces the possibility of closely-
coupled inverter systems reinforcing the supply seen by each other, leading to the formation of
an unintentional island. Note that these limits remain broadly similar in the current draft of the
upcoming AS/NZS4777.2-2015, although the field of application has been extended from
inverters rated up to 30 kilo-volt ampere (kVA) to any inverter connected to the LV network.
Even noting this requirement for active anti-islanding detection, there is some concern in the
literature [2] that closely-connected inverter systems at a high enough penetration may still
produce a stable supply beyond the restrictions described in AS(/NZS) 4777. This is usually
related to the requirement for inverters to provide some level of fault ride-through and grid
support. A potential scenario may involve a fault that is incorrectly identified by an inverter as a
voltage excursion; the inverter then attempts to ride through the LV by maintaining its own
output, creating a temporary islanded network. This can be exacerbated when multiple inverters
are co-located, such that they support one another, reducing each inverter’s ability to detect
the fault or off-grid condition and further extending the duration of the island.
If this type of islanding lasts long enough, it could cause safety concerns as described above. It
also poses a potential issue to automated recloser systems (see Section 3.2.2.6).
Summary of anti-islanding detection methods

Mechanisms for detecting an islanded electrical system have traditionally relied on passively
monitoring electrical parameters for deviations that indicate the network segment has been

isolated, and then commanding the inverter to stop exporting power. More recently, significant
effort has been made into researching active methods for identifying grid disconnections,
typically by attempting to vary the grid parameters.
Reference [3] identifies the following passive and active islanding detection mechanisms, along
with a review of the benefits and drawbacks of each specifically for the case of PV inverters.
Passive techniques
 Over/under-voltage and over/under-frequency detection
 Phase jump detection
 Direct monitoring of voltage and current harmonics
 Rate of change of frequency
 Rate of change of power output
Active techniques
 Impedance measurement
 Sliding mode frequency shift (active phase shift)
 Sandia frequency shift (active frequency shift with positive feedback)
 Reactive power export error detection
Reference [3] also describes a number of ‘remote techniques’, where a grid failure is detected
by the grid operator rather than by the inverter or its grid protection device. This is then
communicated to the inverter to force it to shut down. As these types of techniques rely on
communication with the inverter, they may be limited in applicability. However, the upcoming
update of AS/NZS 4777 Part 2 includes a demand reponse (DR)-type interface including a DRM0
command to isolate from the grid, which may encourage future adoption of these techniques.
No single technique was shown to demonstrate a superior set of functionality [3] . Rather, they
all trade off desirable and non-desirable aspects, including complexity of implementation, non-
detection zone, speed of detection and isolation, and rate of false positives/false negatives.
Automatic re-closing of non-synchronised network segments

When a fault is detected on a piece of network, a circuit breaker can be opened to isolate the
network segment in which the fault has occurred. In the case of overhead breakers with
reclosers, these faults are often quickly cleared, and the breaker is closed to once again provide
supply to the network segment. However, if an island has formed during the power disruption,
there may no longer be direct synchronisation of the voltages on either side of the breaker.
When the connection is made and the unsynchronised voltages clash, this can cause
unpredictable current flows. It can also potentially damage the EG, as well as customer loads
and appliances.
3.2.3.4 Impedance relay operation

Impedance relays can be used to only operate on specific regions within a feeder. These units
rely on knowledge of the impedance characteristics of the local network to estimate where
faults are occurring, based on fault current and voltage profiles. By varying the voltage profile
seen at a relay, EG may compromise the reliability of these systems.
Key Finding: EG has the potential to contribute current to any fault in the vicinity.
Protection system design and settings must be considered in rural and off-grid systems
to ensure faults are cleared quickly. In urban areas, the added fault level (the
maximum current that can flow in the network segment as a result of a fault) may
restrict the amount of EG that can be connected due to system design fault levels.
3.2.4 Network stability

What is it?
According to AEMO, ‘power system stability is the ability of the electric power system, for a
given initial operating condition, to regain a state of operating equilibrium after being subjected
to a physical or electrical disturbance, with system variables bounded so that practically the
entire power system remains intact’ [61] .
By varying the typical operating constraints under which a conventionally designed power system
operates, EG can have a detrimental effect on the voltage and frequency stability profiles of a
generating unit. In addition, circulating reactive power in islanded microgrids or off-grid power
systems can cause oscillatory effects in voltage at various points on the network.
Stability is of significant issue to all network entities, but in particular to those that own
generation assets. While most Australian DNSPs do not currently own significant generation
assets, this is becoming increasingly attractive in managing fringe-of-grid networks, as detailed
below.
The impact of EG on power system stability has been a concern for some distribution networks
for many years [61] . Until recently, this has typically been considered a primary concern for LV
and MV distribution networks. However, the significant increase in distributed solar in recent
years has affected the upstream HV network. With the uptake of EVs and battery storage
expected to greatly increase in the next 10—20 years, there is a strong possibility that
correlation of these new loads due to static tariff structures or consumer behaviour may
increase this effect.
Power system stability can be broadly categorised into stability at centralised generation (i.e.
power station stability), and stability within the transmission or distribution network. At a
central generation, such as a power station, generator output voltage, rotor frequency, rotor
angle and output phase angle are all critical to maintaining stability of power networks. Within
distribution networks, frequency and its related variables are primarily controlled by the
upstream generation, while voltage stability is of greater relevance.

3.2.4.1 Power station stability
In the traditional radial model of a power system, maintaining the stability of centralised
generation is key to ensuring consistent, reliable power for consumers, as required by the
Australian National Electricity Rules. Power station stability is strongly coupled to the load
demand on power networks, as detailed below. Increasing penetration of EG may actually
improve power system stability [63] . However, this is dependent on complex interactions
between system dynamics, and the role of large amounts of EG in this is not yet clear.
In a typical large-scale network power station, stability is the concern of the generator operator;
the DNSP is shielded from these issues to a degree. In addition, given the relatively low
penetration of EG when compared with the total load supplied by existing, large-scale,
centralised generation, it is easy to assume that Australian utilities will not face these problems
for some time — if ever. However, some Australian DNSPs (most notably Horizon Power) manage
significantly smaller generators that have already been exposed to significant penetrations of
EG. In addition, there is an increasing case being mounted for other DNSPs to be able to own and
operate small generation assets in islanded microgrids as an alternative mechanism for
supporting fringe-of-grid networks. In these situations, challenges such as those detailed below
may become significantly more problematic than in the current, centrally dominated generation
model.
At time scales of several cycles to several seconds, a balance is maintained between power
demand by customers and power supplied by generators. This occurs via the interaction between
the passive dynamics of the transmission and distribution networks, the primarily large
electromechanical machines generating power, and the regulation and control mechanisms on
the network designed to provide electrical power as drawn by customers. These system
dynamics are maintained within safe operating limits of the electrical system infrastructure.
The interaction among all the components of the electrical system, including all associated
regulatory mechanisms, should result in stable voltages and power flows — in the sense that
small departures from an equilibrium operating point will decay at time scales faster than
changes in load. At time scales of several cycles to several seconds, the stability of at least
three parameters are of engineering design interest: voltage stability, power system frequency
and synchronous generator rotor angle stability. Voltage stability, a primary concern for
distribution networks, is strongly coupled to the regulation of reactive power flows. Power
system frequency and synchronous generator rotor angle stability are both strongly coupled to
the regulation of real power.
Phase angle stability is also a significant factor. However, given its relationship to frequency
stability, as well as the limited relationship with EG (beyond EG acting as a negative load), this
report does not address it in detail.
Voltage stability

Voltage stability on electrical networks is maintained by control mechanisms that are designed
with the implicit assumption that the alternating current (AC) voltage at any given location in
the network will increase with the injection of reactive power, rather than decrease. The
transmission of too much real power across a transmission link — or the provision of insufficient
reactive power at various points within a network — can violate this condition, resulting in
voltage instability. Correspondingly, an insufficient quantity of generation providing voltage
support via the production of reactive power in response to a fall in voltage (a service that is
presently provided by synchronous machine generators) could lead to voltage instability.
The inclusion of EG will typically reduce the likelihood of voltage instability. This is due to
demands to transmit too much real power across a link, which will typically reduce real power
flows, rather than increase them, since it generates power closer to load. Inverter-based EG
typically supplies little or no reactive power, either pushing centralised generation to increase
reactive power outputs or risking voltage instability [64] .
Frequency and rotor angle stability
Frequency stability and rotor angle stability are both associated with the regulation of real
power supply, and refer to the behaviour of large-scale mechanical generators. Rotor angle is
the phase difference between voltage phases of the generator’s rotor and that of the grid. The
rotor angle increases with increasing power generation output requirements (up to a stability
limit), and its stability is defined relative to the grid. The frequency is that of the grid
(corresponding to that of the majority of the generators), and its stability is defined relative to
an external set-point reference. Grid frequency and generator rotor angle are not strictly
independent, in that frequency is the rate of change of phase. However, there is sufficient
separation of time scales and set point to ignore this dependence in stability analysis and
regulation design.
Frequency stability is achieved if the responsiveness of generators to rising and falling power
demand requirements is sufficient to ensure that the resulting transient phase errors do not
exceed the stability margin for any generator. Instantaneous net differences in power supplied
by mechanical generators and that delivered to the grid are met by changes in energy stored in
the generators in the form of rotational inertia, and hence rotor phase angle. Inverter-based
generators have negligible inertia, and would be typically incapable of increasing power
generation at the rates required to help maintain frequency stability. Embedded synchronous
generation, on the other hand, will be associated with corresponding rotational inertia, and may
help maintain grid frequency stability.
Different types of centralised generation plant have different power frequency dynamic
characteristics. These are influenced by their power control scheme, and constrained by the
inherent mechanical energetic dynamic characteristics of their physical mechanisms. Hydro-
electric turbines, synchronous and asynchronous generators, steam turbines that provide
mechanical power to synchronous generators, and generating devices with inverters all have
their own particular dynamics. Different types of load and transmission links also have their own

particular steady state and dynamic characteristics that relate real (and reactive) power
consumption to grid frequency and voltage supply. The interaction among these dynamic
responses determines whether the interacting electricity supply network is stable (on a local,
small-signal scale).
The steady-state response of synchronous machines is typically characterised by a (closed loop)
droop, with an approximately negatively proportional relationship between frequency and power
generation. Maintaining frequency stability of the system would likely be easier if EG mimicked
the dynamic power response to frequency deviations of centralised plant. Some inverter-based
EG uses droop [65] , but truly replicating the frequency response of large-scale power generators
would likely require significantly more inertia than is typical of EG.
Small-signal rotor angle stability refers to the stability characteristics of the regulated phase
angle of electromechanical generators. It appears to depend [66] on the interaction between
the torsional characteristics of a turbine generator (shaft) and the generator’s own excitation
controls (speed governors) with nearby direct current (DC) inverters. This type of stability seems
to be associated primarily with self-regulation exciting resonant electrical modes. It does not
appear to be strongly associated with interaction with other generators via the grid.
Large-scale rotor angle stability requires local stability of equilibrium stability. In addition, the
magnitude of disturbances from all sources (including faults) must not produce a response that
takes the rotor angle phase difference outside its stability region.
3.2.4.2 Circulating reactive power from embedded generation
The ability to export or import reactive power is a key benefit of embedded generators and can
be used in many aspects of network control, from simple reactive power support to optimisation
of local voltage through volt-VAR control.
Volt-VAR control, in particular, is a useful technique for ensuring that voltage deviations (e.g.
due to high-penetration solar EG) do not lead to inverter disconnection and loss of generation. In
this sense, integrated volt-VAR control in existing EG directly improves the stability of
distribution networks. However, if not properly controlled, reactive power loading and injection
can lead to a situation where VARs are supplied by local EG and absorbed by a nearby generator.
This inflates network currents and increases I2R losses, and can also lead to steady-state voltage
instability due to the strong coupling of voltage and reactive power.
As noted in the previous section, droop control is a useful mechanism for allowing EG to react to
frequency and voltage deviations in the same manner as traditional generation. This greatly
simplifies the integration of droop-capable EG units. However, both local voltage variations
within a network and increasing power output values can lead to poor reactive power sharing,
and resultant steady-state voltage instability [67] . This is particularly prevalent in microgrids,
where the ability to provide reactive power support is primarily or wholly the responsibility of
local EG [68] .

3.2.5 Power quality
What is it?
Power quality refers to how close a single or multiphase supply voltage comes to the intended
waveform. The properties that contribute to power quality include:
 waveform amplitude and variability
 phase imbalance
 voltage flicker
 voltage harmonics.
EG can significantly affect power quality by exporting (or in some cases importing) power on one
phase more than others, or at an inopportune time. It can also generate current harmonics,
which can have detrimental effects on the local voltage waveform and hence on other electrical
equipment.
Network operators in Australia are required to maintain the above aspects of power quality on
their networks within strict tolerances, and EG can pose a challenge to their management.
This section considers the potential technical challenges that affect quality of power:
specifically with regards to voltage fluctuation, harmonic distortion and reactive power
compensation. The increased penetration of renewable energy in the grid has caused growing
concern in distribution system planning regarding power quality issues, as observed by Australian
DNSPs. Therefore, the impact of EG integration must be carefully investigated to ensure
optimum system performance.
Common problems related to voltage fluctuations are over/under-voltage, voltage flicker and
power outages. These can disturb sensitive electric and electronic equipment, reducing its
lifespan and inconveniencing consumers. The variability of renewable generation over time is
not the only reason for these problems; grid-connection issues, load dynamics, faults during
operation and starting of load motors are also responsible. The following sections will address
these technical impacts. Several of the power quality impacts of EG are also covered in detail in
Section 5.2.3, where they relate to the commercial operations of networks.
3.2.5.1 Voltage intermittency and voltage rise
The integration of EG can result in over-voltage issues. This need not be a problem when EG is
connected to a system facing an LV problem; but for weakly loaded systems, it may result in HV
problems, thereby interfering with standard voltage regulation practices [39] . In a weak grid
with low short-circuit capacity and high grid impedance, the voltage rise problem may be
severe, the control capability of EG may need to be exploited [69] . DNSPs have had different
experiences of voltage rise, due to different network strengths and settings. However, in
general, the weak parts of a network are more likely to start experiencing such issues at
relatively modest penetration levels. Accordingly, rural feeders are more likely to experience
voltage rise problems because of their long span, which increases the feeder impedance values.
Specifically, due to the Ferranti effect, the voltage at the receiving end of lightly loaded long
transmission line becomes greater than the sending end because of the voltage drop across the
line inductance.
Recently, Ausgrid conducted a real-world trial on the strong part of a network, using batteries to
simulate different PV penetration levels [72] . No single threshold level of penetration was
observed at which voltage problems arose. However, Ausgrid noted that while higher
penetrations may cause no problem on a strong part of the network, it may on the weaker parts,
or on networks with considerably different operating characteristics.
As with most network trials, Ausgrid’s results are context dependent.For example, recent
studies by Ergon Energy and Central Queensland University [43] have explored the issues
involved in varying PV penetration levels on different SWER networks. If all customers (around
70) installed 5-kW PV systems, then the voltage rise would be around 8% on LV networks; if they
all installed 10-kW PV systems, it would be around 15%. Increasing the penetration levels of
rooftop PVs in residential networks of Perth Solar City could also cause a voltage rise above 6%
of the nominal voltage [40] [41] .
The effect of EG on power quality and stability was also investigated on an island microgrid on
Heron Island, Queensland [57] . The system has unique factors, such as small-inertia diesel
generators, high R/X ratio cables, high PV penetration, and a complete lack of coordination of
loads and power generation. The simulation results indicate approximately 10% increase
temporarily over-voltage and about 6% of steady-state over-voltage, when PV begins to increases
by 75% towards 98% of its operational 50-kW capacity. A wind turbine study observed a steady-
state voltage rise in the network due to wind speed variation, and voltage sag due to the
turbine’s high start-up current [46] .
The impacts of voltage intermittency relating to EG, and in particular how they influence the
commercial operations of networks, are discussed in Section 5.2.2.
Note that Queensland currently has a standard endpoint line-neutral voltage of 240 V +/–6%,
compared with the rest of the NEM, which has standardised on 230 V +10/–6%. Better aligning
both base voltage and voltage range would enable improved consistency of standards and
solutions to address EG issues.
3.2.5.2 Voltage imbalance

Voltage imbalance in the three-phase system is a condition in which the three-phase voltages
differ in amplitude, and/or do not have 120-degree phase difference. The voltage imbalance
factor has been defined using several methods, but the Institute of Electrical and Electronics
Engineers (IEEE) recommend |V-/V+|*100, where V- and V+ are the negative and positive
sequence voltage component, respectively [74] . This definition can be related to the maximum
voltage deviation from the average phase (or line) voltage, divided by the average phase (or
line) voltage. The allowable limit for voltage imbalance is 2% according to the IEEE standard [74]
. The compatibility level for voltage imbalance in Australian Standard AS61000.2.2 is a negative-
sequence component of 2% of the positive sequence component, but allows up to 3% where
there is a high level of single-phase loads, as is typical in residential areas. Note that the
standard does not refer to specific restriction on the zero sequence component.
Such voltage fluctuation, which can be caused by the intermittent nature of EG, is exacerbated
in single-phase systems, because voltage increases can occur on the connected-phase only —
hence contributing to phase imbalance in the network. This can be managed through phase-
swapping, but this comes with an associated cost.
Additionally, the physical location of rooftop PV is still largely dictated by customer choice. For
example, it is possible that 80% of customers on one of the three phases have installed PV, while
only 50% and 10% of customers on the other two phases have installed PV. In such a condition,
even if the voltage imbalance of the network was within the standard limits without any PV, it is
not guaranteed to remain so. Therefore, the possible PV installation number or rating on such
systems must be investigated in a way that retains the voltage imbalance within the standard
limit.
For example, Ergon state that solar PV has produced power quality issues such as voltage rise
and imbalance caused by single-phase connections [73] . Researchers at Queensland University
of Technology have shown that rooftop PV (with a rating of less than 5 kW) can increase network
voltage imbalance by 0.1% when installed at the beginning of the feeder, and by 0.25% when
installed at the end: specifically when the feeder supplied up to a 1-MW load [47] . Therefore,
rooftop PV installation will typically have a minor effect on the voltage imbalance at the
beginning of an MV feeder (less than 1%). However, multiple PV units at the end of the feeder
might increase to more than the 2% standard limit. The same study shows that depending on the
load of the phase in which the PV is installed, the voltage imbalance will increase or decrease
based on the location and rating of the PV. Also, if PV is installed on one low-voltage line, the
voltage imbalance will be modified on all other LV lines of the network. Some minor
improvement is suggested in [47] and completed in [48] .
In general, EG seems to pose little risk to the voltage imbalance on a system, according to
probabilistic load flow experiments [44] . As penetration levels increase from 1 to 50%, the
maximum imbalance remains unaffected, at around 3%. An exception is encountered if EG
provides VAR support; in this case, the EG risks increasing the voltage imbalance if not properly
controlled.
Another case of voltage imbalance of up to 5% occurs when wind energy farms are located in
rural areas with a weak local network or relatively long transmission lines. Such a voltage
imbalance can worsen the performance of doubly fed induction generators (DFIGs) by
introducing unwanted torque harmonics and inaccuracy in the generation of commanded
active/reactive power [46] . DFIGs are widely used in wind energy harvesting systems, due to
their smaller power converter, flexibility in autonomous control of the active and reactive
power, and relatively simple and rugged structure. The impacts of negative-sequence voltage on
DFIG system performance were recently investigated, and a collaborative control strategy based

on the reduced-order generalised integrator was developed to reduce such impacts during
network imbalance [70] .
3.2.5.3 Voltage flicker
Periodic disturbances to the network voltage are defined as ‘flicker’. The level of flicker is
quantified by the short-term flicker severity value, Pst, which is the perception of light flicker in
a short-term (10 minutes) interval. According to IEEE standard 1453, the allowable limit for the
flicker meter output is Pst=1. Also important is the flicker duration, which is the percentage of
time that the bus’ Pst is above a given value. In particular, overambient flicker (OAF) is a
measurement of the percentage of time that Pst is greater than or equal to 0.02. The concerns
associated with flicker are also related to voltage variations. In most cases, voltage flicker
consists of small, periodic voltage fluctuations with frequencies of less than about 30—35 Hz [46]
.
Reference [44] indicates that increased flicker caused by EG can vary based on the presence of
regulators in the system. On a feeder without regulators, the flicker manifests itself as direct
flicker, occurring at low values of Pst. In contrast, on a feeder with regulators, the flicker
manifests itself as indirect flicker, by increasing the length of time over which high flicker values
occur. Indirect flicker occurs at higher values of Pst (because of tap changes); hence, it is the
more concerning source of impact on network voltage. It is also the major driver for a large OAF,
and signifies increased operation of the voltage regulator tap changers and capacitor switches.
This opens the door to a centralised, intelligent control scheme for voltage-regulating devices on
feeders with high penetrations, which would increase device life and reduce the most noticeable
aspect of EG-induced flicker.
Since buses without voltage regulators are usually located at urban feeders and early parts of
rural feeders, the results in [44] indicate that urban feeders are likely to see EG flicker impacts
in the form of direct flicker with minor impact. Rural feeders are instead likely to see EG flicker
impacts as indirect flicker, because the buses often have voltage regulators. As EG penetration
levels increase from 1 to 50%, the average Pst value increases slightly from 0.008 to 0.009, while
the average OAF and maximum OAF change dramatically from 4.2 to 9.7% and 13.8 to 25.4%,
respectively [44] .These values approximately double in a weak network, showing the sensitivity
of a weak network to voltage flicker. These results match intuition on EG impacts. The study
also demonstrates that EG’s direct flicker is unlikely to have a major impact on network voltage,
since it concentrates near lower Pst values. However, the magnitude and probability of the Pst
value noticeably increase as EG penetration increases.
While an overambient flicker value is defined as any flicker greater than or equal to 0.02, a
large OAF does not imply a flicker problem on the system. For there to be a detectable, harmful
flicker problem, OAF would generally need to be greater than 1.0, which does not occur in any
scenario in [44] . The largest Pst observed was 0.6053, which occurred during the most
susceptible test with 50% EG penetration on a weak system. Based on these results, the impact
of EG on flicker appears to be relatively minor. However, the flicker measured here is on the
main feeder only; the flicker on the laterals and customer services is likely to be higher.
Additionally, the weaker system tends to have higher flicker metrics. The result of weakening
the system on OAF is dramatic, causing the weaker system to experience more than double the
OAF of the strong system for equivalent EG penetrations [44] .
In wind turbines, flicker emissions are produced during switching and start-up, as well as during
continuous operation (due to variations in wind speed, the tower shadow effect and mechanical
properties of the wind turbine). Flicker emissions from variable-speed wind turbines, such as
DFIG, are low (around 0.05) compared with those from fixed-speed wind turbines, such as
squirrel-cage induction generators, which are around 0.16 [46] .
Flicker emissions increase in both fixed-speed and variable-speed wind turbines at higher wind
speeds, due to higher turbulence intensity in the wind. However, there is a fundamental
difference in flicker level between the two types of turbine; while it strictly increases at higher
wind speeds for fixed-speed turbine, it first increases but then decreases at rated wind speed
for the variable-speed turbine. As the wind turbine reaches rated power, the variable-speed
system will smooth out the power fluctuations, and thereby limit the flicker [46] . The authors
of [49] also studied flicker emissions in a DFIG-based wind farm. They found that wind generator
control strategies, wind generator reactive power capability and the operating point of the
power curve all affected flicker emissions.
Flicker may also occur in PV modules during fast alternations of clouds and sunshine, but
reference [45] suggests this has no noticeable impact on voltage.
3.2.5.4 Harmonics
Harmonics are one of the most dominant attributes that need to be kept to a minimum to ensure
good network power quality. Harmonic distortion can be minimised by good control algorithm
design in the current control loop. According to IEEE standard 519, harmonics in the power
system should be limited for both the harmonic current that a user can inject into the network
at the point of common coupling (PCC), and the harmonic voltage that the utility can supply to
any customer at the PCC. The limits of different harmonic orders for LV-connected inverters are
specified in Australian Standard AS 4777.
Power electronic devices, together with operation of nonlinear loads, inject current harmonics
into the grid, which may potentially create voltage distortion problems. These effects may
increase power system heat losses and reduce the lifespan of nearby connected equipment.
Harmonic current flowing through the impedances of a distribution network causes voltage
distortion. Harmonic voltages also cause voltage distortion and zero-crossing noise in the
network. The degree of distortion of an AC voltage or current is known as the total harmonic
distortion (THD). It is defined as the ratio of the square root of the sum of the squares of the
individual harmonics to the fundamental harmonics.

A comprehensive study analysed the harmonic impacts on both winter and summer seasons with
10-kW PV penetration on the distribution network [50] . The measured results provided basic
guidelines that include the following for this case: THD of the PV current was low during high
generation, and high in the period of low generation. Also, odd current harmonics affect the
distribution network significantly more than do even harmonics.
Experiments at CSIRO’s Renewable Energy Integration Facility in Newcastle, Australia showed
that increasing PV penetration from 7.5 to 11.3 kW increases the harmonics injection into the
network [75] . The fast Fourier transform (FFT) analysis found that current harmonic injections
from all even harmonics are within the range stated in the AS4777 standard. However, the third
and ninth harmonics exceeded the regulatory standard, while injections from the seventh and
15th harmonics just reached the threshold levels.
Although fixed-speed wind turbines do not employ any power electronic converters, they can
cause harmonic distortion through the interaction of power factor correction capacitors and
system impedances. A small amount of harmonics is also produced during the starting up of the
turbines, while wind turbines with converters inject harmonics into the network during their
operation. The design of power electronic converters and filters is the influencing factor of
harmonics produced by wind turbines [46] .
Characteristics of harmonics injected into a wind energy integrated power system were
investigated with variety of configuration and operating conditions in [51] and [52] . Voltage
distortion is expected to be observed in the distribution network due to the rotating machine
characteristics [51] and the design of power electronic interface [52] in the wind energy
systems, which lead to harmonics in the air-gap. As a result, the stator and rotor will contribute
to current harmonics. Harmonics are also observed in the main field passing over the stator slot,
due to the inherent characteristic of the machine; hence, these are in need of a different
approach if they are to be managed [51] .
A recent study [53] investigated the problem of a frequency converter in a wind energy system
causing harmonics in the line current, leading to harmonic voltages in the network. With the
application and guidance of International Electrotechnical Commission (IEC) regulatory standards
IEC 61000-3-6 and IEC 61000-4-21, this study integrates several appropriate harmonic load
models representative of harmonic network elements and harmonic sources. Potential voltage
distortion was observed, caused by coincidence of the first harmonic resonance frequency with
the peak harmonic spectrum of the wind turbine current. The harmonic distortion of the voltage
increased at minimum load, as well as being more pronounced closer to the wind farm. An HV
distortion appeared at the MV busbars, which is concerning, while minimal distortion occurs at
the HV side [53]
Reference [54] explored the characteristics of self excitation and harmonics generated by a
fixed-speed induction generator used in wind turbine generation. The saturation of the magnetic
circuit in the transformer and the resonance circuit between the capacitor compensation and
the rest of the circuit causes harmonic injection, which makes the power system vulnerable. The
tap changer, capacitor compensator, power factor and level of generation all influence the
intensity of saturation, as well as the characteristics and amount of the harmonic source
injection. Power quality behaviours — in particular, voltage sags and harmonics injection into
the network — were investigated in a study on integrating wind energy into LV and MV networks
[55] . The results showed low injection of voltage sags for all three case study scenarios.
Observed THD was within the safety limit; however, it increased with greater wind energy
penetration into the system.

The harmonic issues arising from PV systems coupled by multiple inverters have been
investigated at CSIRO’s Renewable Energy Integration Facility [56] . They showed that harmonic
current emission is related to the output power level of the PV inverter. In addition, the current
and voltage waveform distortion at the PCC to the grid can be significantly increased; however,
this is most prevalent when the net power flow through the bus is small.
3.2.5.5 Voltage regulation
There has recently been much interest and concern in the industry regarding allowing EG to
participate in VAR support, and therefore regulate the voltage at their bus, in the same way as a
conventional generator.
For example, in wind turbines, the consumption of reactive power by an induction generator is a
common problem affecting grid power quality. An induction generator requires an increasing
amount of reactive power as the power factor improves at high loading, and the reactive power
must be supplied locally as close as possible to the demand levels. Due to fluctuations in the
active and reactive power, the voltage also fluctuates at PCCs.
A recent study [44] exemplifies why allowing EG to participate in VAR regulation requires
careful consideration. The effect of using VAR-regulating EG to help shape the voltage profile of
a feeder is profound, and the possible benefits are vast (see Section 3.3.4). However, it is clear
that without careful management of VAR injections, unexpected, less-desirable results in
voltage quality can easily occur. For example, the maximum regulated voltage limits would be
expected to fall as more VAR capacity is available to the system. This is the case when
comparing metrics at the same EG penetration level: which is, in this experiment, the installed
capacity of EG on the system over the peak-load demand. For example, in [44] , at 25% EG
penetration level, the maximum voltage falls from 1.069 per unit (pu) with 20% VAR support to
1.066 pu with 100% VAR support. At 50% EG penetration level, maximum voltage falls from
1.083 pu (20% VAR support) to 1.076 pu (100% VAR support). However, the maximum voltage
limit increases over the cases without any VAR support (e.g. at 50% EG penetration level without
any VAR support, it has a 1.073 pu voltage limit). These findings may be due to an inconsistent
amount of VAR injection across the feeder.
Another important issue to address when considering EG reactive power participation is how
much the inverter must be oversized. The authors of [47] found a substantial decrease in voltage
imbalance with a modest 16% oversizing of the EG.
The capability diagram of grid-tied microgrids is defined in [58] as the import/export capability
of active/reactive power. This research analysed the impacts of individual EG capability limits,
load modelling, voltage regulation, EG outages, and plug-in hybrid EV operation on the limits of
microgrid capability diagram. Due to time-varying load and generation patterns, the capability
diagram of a microgrid will vary over time, and it is not practical for the microgrid central
controller to derive capability diagrams for an entire day.

3.2.5.6 Power quality impact by technology
Table 1 and Table 2 summarise the power quality impacts identified in the literature for inverter
technology and rotating machines.
Table 1 – Power quality impacts of inverter technology
Consideration Impact Note
Voltage rise Network configuration May exceed the standard limit As penetration level increases, the average voltage
goes up, but minimum and maximum voltages are
No. of feeders
almost stable
Voltage regulation method
Voltage Geographical and electrical 1) Minor impact 1) Network phase imbalance worsens in single-
imbalance distribution of photovoltaic (PV) phase systems due to inconsistent embedded
2) May exceed the standard
installation generation (EG) distribution
limit at the end of the feeder
Intermittent nature of PV 2) Imbalance increases when EG installed at the
end of the feeder
Flicker Grid impedance No noticeable impact 1) Weak network has higher flicker metrics
2) Direct flicker on a feeder without voltage

regulator (urban feeders) has minor impact (low
metric values)
3) Rural feeders (with regulator) are likely to see

indirect flicker impact
Harmonics Harmonic content of the grid Normally below the standard 1) Odd current harmonics generally have more
(total harmonic voltage limit, but may exceed the limit impact than even harmonics
distortion)
Series impedance of the grid 2) Total harmonic distortion increases with
increasing EG penetration
PV inverter topology
Current control loop technology
Voltage Inconsistent volt-amps reactive Needs careful management As VAR capacity increases, minimum voltage rises
regulation (VAR) injection and maximum voltage falls
Table 2 – Power quality impacts of rotating machines
Consideration Impact Note
Voltage Network configuration May exceed the standard limit

rise/drop
Number of feeders
Voltage regulation method
Voltage Wind farm location May exceed the standard limit, Small wind turbines in suburban and downtown
imbalance depending on turbine location areas may help
Rural areas with weak networks
Long transmission lines

Flicker Wind speed variation May exceed the standard limit 1) Flicker emission of variable-speed turbine
(doubly fed induction generator) is lower than
Tower shadow effect
fixed-speed (squirrel-cage induction generator)
Mechanical properties of wind
2) Flicker emission increases at higher wind speed
turbines
Start-up operation of generators
Harmonics Power electronics Normally bellow the standard Fixed-speed turbines have no power electronic
(total harmonic limit, but may exceed the limit converters, but may cause harmonic distortions
Rotating machine characteristics
distortion) due to power factor correction capacitor
Turbines with converters (variable-speed turbines)

inject harmonics
Small amount of harmonics during start-up
Harmonic distortion of the voltage is increased at

minimum load and closer to wind farm
Voltage Inconsistent volt-amps reactive Needs careful management As VAR capacity increases, minimum voltage rises
regulation (VAR) injection and maximum voltage falls

References
[1] ‘Standard for Connection of Embedded Generators in the Ergon Energy Distribution
Network’, Ergon Energy.
[2] Cornforth, D., Sayeef, S., Moore, T., 2001, ‘Beyond overcurrent protection: Distributed
generation in the future grid’, ISGT Asia, 2011.
[3] Ahmad, K.N.E.K., Selvaraj, J., Rahim, N.A., ‘A review of the islanding detection methods
in grid-connected PV inverters’, Renewable and Sustainable Energy Reviews, vol. 21, pp.
756-766, May 2013.
[4] ‘Intelligent Grid: A value proposition for distributed energy in Australia’, CSIRO, 2009.
[5] Keller, J., Kroposki, B., 2010, ‘Understanding Fault Characteristics of Inverter-Based
Distributed Energy Resources’, National Renewable Energy Laboratory.
[6] Nimpitiwan, N., Heydt, G.T., Ayyanar, R., Suryanarayanan, S., 2007, ‘Fault Current
Contribution From Synchronous Machine and Inverter Based Distributed Generators’,
Power Delivery, IEEE Transactions on, vol. 22, no.1, pp. 634-641.
[7] Loix, T., Wijnhoven, T., Deconinck, G., 2009, ‘Protection of Microgrids with a High
Penetration of Inverter-Coupled Energy Sources’, IEEE PES/CIGRE Symposium. Calgary,
Canada, 29–31 Jul. 2009.
[8] Roughan, T. (2013, October 11-12). ‘Workshop on Microgrid Technologies and
Applications.’ Retrieved from Rensselaer Polytechnic Institute:
http://www.rpi.edu/cfes/Workshop%20on%20Microgrid/C3%20Roughan%20NGrid.pdf.
[9] Jayawarna, N., Jenkins, N., Barnes, M., Lorentzou, M., Papthanassiou, S., Hatziargyriou,
N., 2005, ‘Safety analysis of a microgrid’, in Proc. Int. Conf. Future Power Syst.,
Amsterdam, The Netherlands, Nov. 16–18, 2005, pp. 1–7.
[10] Flores, A., Quiles, E., Emilio García, E., Morant, F., Correcher, A., 2014, ‘A Modular
Neural Network Scheme Applied to Fault Diagnosis in Electric Power Systems,’ The
Scientific World Journal, vol. 2014, Article ID 176463, 13 pages.
doi:10.1155/2014/176463
[11] Paraskar, S.R., Beg, M.A., Dhole, G.M., 2011, ‘Discrimination between Inrush and Fault
in Transformer: ANN Approach’, International Journal of Advancements in Technology,
vol. 2, no. 2, pp. 306–318.
[12] ‘Your own energy “island”? ORNL microgrid could standardize small, self-sustaining
electric grid’, http://www.ornl.gov/ornl/news/features/2014/a47cc734-54f3-4aec-95ba-
df1bbbdb25ad, accessed on 28 April 2015.
[13] Nikkhajoei, H., Lasseter, R.H., 2007, ‘Microgrid Protection’, Power Engineering Society
General Meeting, 2007. IEEE, pp.1-6. Tampa, Florida, 24-28 June 2007.
[14] Ergon submission to ‘QCA's Issues Paper - Estimating a Fair and Reasonable Solar Feed-in
Tariff for Queensland’, Sept 2012.
[15] ‘Energex Power Quality Strategic Plan, 2015-20’, Oct 2014.
[16] Walling, R.A., Saint, R., Dugan, R.C., Burke, J., Kojovic, L.A., 2008, ‘Summary of
Distributed Resources Impact on Power Delivery Systems’, Power Delivery, IEEE
Transactions on, vol.23, no.3, pp.1636-1644.
[17] ‘Addressing Grid-interconnection Issues with Variable Renewable Energy Sources’,
prepared for APEC by IT Power, 2010.
[18] ARUP. Ausgrid smart grid smart city microgrid protection research (accessed 09-07-2015)
https://ich.smartgridsmartcity.com.au.
[19] Taskin, S., 2010, ‘Power Plants Monitoring For Reverse Power Flow Evaluation’, Journal
Of Electrical and Electronics Engineering, vol. 10, no. 1, pp.1153-1162.
[20] ‘Change and choice: The Future Grid Forum’s analysis of Australia’s potential
electricity pathways to 2050’, CSIRO, 2014.
[21] Wan, Y., Parsons, B. K., 1993, ‘Factors Relevant to Utility Integration of Intermittent
Renewable Technologies’, NREL/TP-463-4953.

[22] Sayeef, S., Heslop, S., Cornforth, D., Moore, T., Percy, S., Ward, J.K., Berry, A., Rowe,
D., 2012, ‘Solar intermittency: Australia’s clean energy challenge – Characterising the
effect of high penetration solar intermittency on Australian electricity networks’, CSIRO
report.
[23] Bebic, J., 2008, ‘Power System Planning: Emerging Practices Suitable for Evaluating the
Impact of High-Penetration Photovoltaics’, NREL Technical Report (NREL/SR-581-42297).
[24] California Energy Commission, 2007, ‘Intermittency Analysis Project: Appendix B Impact
of Intermittent Generation on Operation of California Power Grid’, prepared by GE
Energy Consulting (CEC-500-2007-081-APB).
[25] The Energy Supply Association of Australia (ESAA) - Electricity Gas Australia 2015 report.
[26] CAT Projects and ARENA, 2015, ‘Investigating the Impact of Solar Variability on Grid
Stability’.
[27] UNSW Centre for Energy and Environmental Markets (CEEM), 2011, ‘Alice Springs – A
Case Study of Increasing Levels of PV Penetration in an Electricity Supply System’.
[28] Jayaweera, D., Islam, S., 2014, ‘Security of energy supply with change in weather
conditions and dynamic thermal limits’, IEEE Transactions on Smart Grid, vol. 5, no. 5,
pp. 2246-2254.
[29] Kerahroudi, S.K., Zobaa, A.F., Taylor, G.A., Bradley, M., Yang, L., 2012, ‘Investigating
the impact of stability constraints on the future GB transmission system’, Proceedings of
the Universities Power Engineering Conference, Article number 6398448. London, 4-7
Sep. 2012.
[30] Gong, Q., Midlam-Mohler, S., Marano, V., Rizzoni, G., 2012, ‘Study of PEV charging on
residential distribution transformer life’, IEEE Transactions on Smart Grid, vol. 3, no. 1,
pp. 404-412.
[31] Richardson, P., 2012, ‘Integration of Distributed Energy Resources in Low Voltage
Electricity Networks’, PhD thesis, School of Electrical, Electronic and Communications
Engineering, University College Dublin.
[32] Wang, C., Nehrir, M. H., 2004, ‘Analytical approaches for optimal placement of
distributed generation sources in power systems’, IEEE Transactions on Power Systems,
vol. 19, no. 4, pp. 2068-2076.
[33] Acharya, N., Mahat, P., Mithulananthan, N., 2006, ‘An analytic approach for DG
allocation in primary distribution network’, International Journal of Electrical Power &
Energy Systems, vol. 28, no. 10, pp. 669-678.
[34] Paliwal, P., Patidar, N.P., Nema, R.K., 2014,‘Planning of grid integrated distributed
generators: A review of technology, objectives and techniques’, Renewable and
Sustainable Energy Reviews no. 40, pp. 557-570.
[35] El Nozahy, M.S., Salama, M.M.A., 2014, ‘A comprehensive study of the impacts of PHEVs
on residential distribution networks’, IEEE Transactions on Sustainable Energy, vol. 5, pp.
332-342.
[36] Widén, J., Wäckelgård, E., Paatero, J., Lund, P., 2010, ‘Impacts of distributed
photovoltaics on network voltages: Stochastic simulations of three Swedish low-voltage
distribution grids.’ Electric Power Systems Research, vol. 80, pp. 1562-1571.
[37] Australian Energy Market Operator, 2012, ‘2012 National Electricity Forecasting
Report’.
[38] CIGRE Australia – available at: http://cigreaustralia.org.au/news/in-the-loop-
news/future-proofing-electricity-forecasting/.
[39] Paliwal, P., Patidar, N.P. & Nema, R.K. 2014, ‘Planning of grid integrated distributed
generators: A review of technology, objectives and techniques’, Renewable and
Sustainable Energy Reviews, vol. 40, pp. 557-570.
[40] Geiles, T.J., Islam, S.M., 2013, ‘Impact of PEV charging and rooftop PV penetration on
distribution transformer life’, IEEE Power and Energy Society General Meeting.

[41] Islam, S.M., Giles, T.J., 2012, ‘Distribution transformer life optimisation in a smart
feeder embedded with photovoltaic power generation’, 7th International Conference on
Electrical and Computer Engineering, pp. 953-957. Dhaka, Bangladesh, 20-22 Dec. 2012.
[42] Albarracin, R., Amaris, H., 2009, ‘Power quality in distribution power networks with
photovoltaic energy sources’, In Proceedings of the 8th International Conference on
Environment and Electrical Engineering. Karpacz, Poland, 10-13 May 2009.
[43] Guinane, A., Shafiullah, G.M., Oo, M.T.A., Harvey, B., 2012, ‘Voltage fluctuations in PV
penetration on SWER networks - A case study for regional Australia’, Proceedings of the
IEEE PES General meeting. San Diego, United States, 22-26 July 2012.
[44] Sexauer, J.M., Mohagheghi, S., 2013, ‘Voltage quality assessment in a distribution
system with distributed generation - a probabilistic load flow approach’, IEEE
Transactions on Power Delivery, vol. 28, no. 3, pp. 1652-1662.
[45] Shafiullah, G.M., Oo, A.M.T., Ali, A.B.M.S., Wolfs, P., Arif, M.T., 2013, ‘Renewable
energy integration: Opportunities and challenges’, Green Energy and Technology, vol.
132, pp. 45-76.
[46] Shafiullah, G.M., Oo, A.M.T., Ali., A.B.M.S, Wolfs, P., 2013, ‘Potential challenges of
integrating large-scale wind energy into the power grid - A review’, Renewable and
[47] Shahnia, F., Majumder, R., Ghosh, A., Ledwich, G., Zare, F., 2011, ‘Voltage imbalance
analysis in residential low voltage distribution networks with rooftop PVs’, Electric Power
Systems Research, vol. 81, no. 9, pp. 1805-1814.
[48] Shahnia, F., Ghosh, A., Ledwich, G., Zare, F., 2014, ‘Voltage unbalance improvement in
low voltage residential feeders with rooftop PVs using custom power
devices’, International Journal of Electrical Power and Energy Systems, vol. 55, pp. 362-
377.
[49] Meegahapola, L., Perera, S., 2014, ‘Capability constraints to mitigate voltage
fluctuations from DFIG wind farms when delivering ancillary services to the
network’, International Journal of Electrical Power and Energy Systems, vol. 62, pp. 152-
162.
[50] Fekete, K. Klaic, Z., Majdandzic, L., 2012, ‘Expansion of the residential photovoltaic
systems and its harmonic impact on the distribution grid’, Int. J. Renewable Energy, vol.
43, pp. 140–148.
[51] Lindholm, M., Rasmussen, T., 2003, ‘Harmonic analysis of doubly fed induction
generators’, Proceedings of the 5th International Conference on Power Electronics and
Drive Systems (PEDS), pp. 837–841. Singapore, 17-20 Nov. 2003.
[52] Jorgensen, P., Tande, J.O., Vikkelso, A., Norgard, P., Christensen, J.S., Sorensen, P.,
1997, ‘Power quality and grid connection of wind turbines’, Proceedings of the 14th
International Conference and Exhibition on Electricity Distribution (CRIED). Birmingham,
United Kingdom, 2-5 June 1997.
[53] Papathanassiou, S.A., Papadopoulos, M.P., 2006, ‘Harmonic analysis in a power system
with wind generation’, IEEE Transactions on Power Delivery, vol. 21, no. 4, pp. 2006-
2016.
[54] Muljadi, E., Butterfield, C.P., 2005, ‘Self excitation and harmonics in wind power
generation’, Proceedings of the 43rd AIAA Aerospace Science Meeting and Exhibit, Reno,
United States, 10 January 2005.
[55] Vilchez, E., Stenzel, J., 2008, ‘Wind energy integration into 380 kV system – impact on
power quality of MV and LV networks’, Transmission and Distribution Conference and
Exposition: Latin America, 2008 IEEE/PES, Bogota, 13-15 Aug. 2008.
[56] Du, Y., Lu, D.D., Cornforth, D., James, G., 2011, ‘A study on the harmonic issues at
CSIRO Microgrid’, Proceedings of the International Conference on Power Electronics and
Drive Systems, pp. 203-207. Singapore, 5-8 Dec. 2011.

[57] Franklin, A., Yan, R., Saha, T.K., 2013, ‘Investigation of a micro-grid operation: A case
study of Heron Island network, Australia’, 2013 Australasian Universities Power
Engineering Conference, AUPEC 2013. Hobart, Australia, 29 Sep. - 3 Oct. 2013.
[58] Jayawardena, A.V., Meegahapola, L.G., Robinson, D.A., Perera, S., 2015, ‘Microgrid
capability diagram: A tool for optimal grid-tied operation’, Renewable Energy, vol. 74,
pp. 497-504.
[59] Elakrmi, F., Abu Shikhah, N., 2011, ‘Electricity Load Forecasting – Science and
Practices’, Jordan International Energy Conference, September 2011.
[60] Australian Energy Market Commission, 2013, ‘Advice to SCER – Consideration of
Differences in Actual Compared to Forecast Demand in Network Regulation’.
[61] Australian Energy Market Operator, 2012, ‘Power System Stability Guidelines’.
[62] Jenkins, N., Strbac, G., 1997, ‘Impact of embedded generation on distribution system
voltage stability,’ Voltage Collapse (Digest No: 1997/101), IEE Colloquium on, pp.9/1-
9/4. London, United Kingdom, 24 Apr. 1997.
[63] Newman, D.E., Carreras, B.A., Kirchner, M., Dobson, I., 2011,‘The Impact of Distributed
Generation on Power Transmission Grid Dynamics,’ System Sciences (HICSS), 44th Hawaii
International Conference on, pp.1-8, 4-7 Jan. 2011.
[64] Amarasekara, H.W.K.M., Meegahapola, L., Agalgaonkar, A.P., Perera, S., 2013, ‘Impact
of renewable power integration on VQ stability margin,’ Power Engineering Conference
(AUPEC), 2013 Australasian Universities, pp.1–6, Hobart, Australia, 29 Sep. - 3 Oct. 2013.
[65] Vasquez, J.C., Guerrero, J.M., Gregorio, E., Rodriguez, P., Teodorescu, R., Blaabjerg,
F., 2008, ‘Adaptive droop control applied to distributed generation inverters connected
to the grid,’ IEEE International Symposium on Industrial Electronics, pp.2420-2425.
Cambridge, United Kingdom, 30 June - 2 July 2008.
[66] P. Kundur, N.J. Balu, M.G. Lauby, 1994, ‘Power System Stability and Control’, Chapter
15, McGraw-Hill Publishing.
[67] Rowe, C.N., Summers, T.J., Betz, R.E., Cornforth, D.J., Moore, T.G., 2013, ‘Arctan
Power–Frequency Droop for Improved Microgrid Stability’, IEEE Transactions on Power
Electronics, vol.28, no.8, pp.3747-3759.
[68] Lidula N.W.A., Rajapakse, A.D., 2011, ‘Microgrids research: a review of experimental
microgrids and test systems’, Renewable and Sustainable Energy Reviews, 15, pp. 186–
202.
[69] Maryam Hashemi, N., Agelidis, V., 2013, ‘Evaluation of voltage regulation mitigation
methods due to high penetration of PV generation in residential areas’, Proceedings of
2013 International Conference on Renewable Energy Research and Applications, pp. 1180.
Madrid, Spain, 20-23 Oct. 2013.
[70] Cheng, P., Nian, H., 2015, ‘Collaborative control of DFIG system during network
unbalance using reduced-order generalized integrators’, IEEE Transactions on Energy
Conversion, vol. 30, no. 2, pp. 453-464.
[71] He, J., Li, Y.W., Wang, R., Zhang, C., 2015, ‘Analysis and mitigation of resonance
propagation in grid-connected and islanding microgrids’, IEEE Transactions on Energy
[72] Ausgrid, 2014, ‘Smart Grid, Smart City: Shaping Australia’s Energy Future - National Cost
Benefit Assessment’.
[73] Ergon Energy, 2012,‘Submission on the issues paper estimating a fair and reasonable
solar feed-in tariff for Queensland’, Queensland Competition Authority.
[74] IEEE recommended practice for monitoring electric power quality, IEEE standard 1159-
1995.
[75] Shafiullah, G.M., Oo, A.M.T., Ali, S., Azad, S., Arif, M., Moore, T., 2012, ‘Experimental
analysis of harmonics on utility grid with PV penetration’, In: Proceedings of the
International Conference on Electrical and Computer Systems, Ottawa, Canada, 22–
24 Aug 2012.

3.3 Options for mitigating network impacts and risks
3.3.1 Introduction
Though it is clear that embedded generation carries a set of real issues that may inhibit its
capacity to play a meaningful and beneficial role in Australia’s electricity networks, the solution
set is rich. The range of mitigation options looks set to grow as storage technologies mature,
forecasting improves, and inverters become smarter — and as systems emerge that can unify
these elements to deliver highly responsive load and generation. This section explores such
options, highlighting opportunities for maximising EG performance and the studies that have
shown the real-world value of pursuing these opportunities. The document map on the following
page summarises the resultant findings, indicating the types of network where benefit is likely
to be seen, when uptake will likely deliver meaningful benefit to the network, and the
significance of the opportunity.

Document map
Mitigation Chapter
option section
Energy storage
3.3.2.1
(grid-scale storage)
Energy storage
3.3.2.2
(distributed storage)
Ancillary services 3.3.3
Intelligent networks
(static 3.3.4.1
compensators)
(passive static VAR 3.3.4.1
compensators)
3.3.4.1
(smart inverters)
3.3.4.2
(demand response)
(ramp rate 3.3.4.3
management)
Enabling
technologies
3.3.5.1
(renewable energy
prediction)
Enabling
technologies
(voltage flicker 3.3.5.2
identification and
control)
Enabling
technologies
(separate 3.3.5.4
generation and load
metering)
Enabling
technologies
3.3.5.5
(improved models
for load prediction)

3.3.2 Energy storage
Energy storage is often seen as a critical component in mitigating the effects of intermittency in
some EG. It is also important for storing bulk energy to reduce the likelihood of reverse power
flow in the transmission and distribution networks. This section highlights mechanisms by which
grid-scale and distributed storage can offset key issues caused by EG.
The Australian Energy Market Commission (AEMC) and CSIRO have recently published a detailed
report on the integration of energy storage into Australian networks [1] . The report describes
types of electrical energy storage and existing Australian trials and deployments, and analyses
technologies considered of particular interest for Australian network entities. It provides
additional context for the information given below.
3.3.2.1 Grid-scale storage
Grid-scale storage encompasses several different technologies with a wide range of

characteristics. These include pumped hydroelectric storage (PHS), compressed air energy
storage (CAES), batteries, flywheels, superconducting magnetic energy storage (SMES) and
supercapacitors.
It is not yet clear which type of energy storage will be the most cost-effective. Many people
automatically think of batteries as the default form of electricity storage, but at an industrial
scale, they require many improvements in the coming decades. PHS is at a mature stage of
development, but locations where they can be built are limited. Other types of storage, such as
flywheel and CAES, are promising technologies for particular applications and locations. For
example, the remote towns of Marble Bar and Nullagine in Western Australia previously required
the delivery of more than one million litres of diesel by truck each year to meet their energy
needs. Through the combined use of solar power and two 500-kW flywheel storage by Horizon
Power, this has been reduced by more than 400,000 litres [74] .
The potential market for energy storage devices could be 3000 MW by 2030, even without any
specific government policies to support storage, but several barriers might restrict this growth
[74] . The Australian Energy Storage Roadmap [75] outlines initiatives to better define and
address the safety, environmental, technical, commercial and informational barriers to the
deployment of both large and small-scale energy storage technologies.
The essential characteristics of typical grid-scale energy storage mediums can be described in
terms of the metrics listed below [2] :
Metric Unit Description
Nominal energy storage capacity kWh Amount of energy that can be stored
| 63
Usable energy storage capacity kWh Usable amount of energy that can be
(kWh): stored, allowing for limits on the
practical depth of discharge without
undesirably shortening the life of the
storage facility
Typical power output MW Amount of power that can be discharged

within the typical discharge duration
Energy density Wh/L Nominal storage energy per unit volume,

i.e. the volumetric energy density
Power density W/L Maximum available power per unit

volume
Charge/discharge duration Time Time needed for the storage to fully

charge or discharge
Response time Time Time needed for the storage to start

providing power output
Lifetime Years or Number of years or cycles that a storage

cycles technology will continue to operate. The
rating in terms of years versus cycles
depends on the specific storage
technology
Roundtrip efficiency % Ratio of energy discharged by the system

to the energy required (including losses)
to charge the system over each cycle
Capital cost $/kW or Upfront investment costs of a storage

$/kWh technology per unit of power discharge
($/kW) or energy storage capacity
($/kWh)
Grid-scale energy storages can provide many grid services, which can be broadly classified based
on their timescale as power (short-duration) or energy (long-duration) services. Common grid
services are described below [2] :
 Power quality services
o support use of electric energy without interference or interruption
o generally refer to maintaining voltage levels within bounds
 Transient stability services
o help maintain synchronous operation of the grid when the system is subject to
sudden (potentially large) disturbances
 Regulation services
o correct short-term power imbalances that might affect system stability
o generally used for frequency synchronisation
| 64
 Spinning reserves
o provide online reserve capacity that is ready to meet electric demand within
10 minutes
 Voltage control
o provides the ability to produce or absorb reactive power
o provides the ability to maintain a specific voltage level
 Energy arbitrage
o using power that is produced during low-cost periods to serve load during high-
cost periods
o i.e., energy storage charges during off-peak times and discharges during peak
times in order to provide load levelling/load shifting
 Load following (balancing)
o adjusting power output as demand fluctuates to maintain power balance in the
system
 Firm capacity services
o provide energy capacity to meet peak power demand
 Congestion relief
o reduces network flows in transmission constrained systems
o done by either increasing the capacity of the lines or providing alternative
pathways for the electricity [3] [4]
 Upgrade deferral
o deferring either generation or transmission asset upgrades
o e.g. by using energy storage to reduce peak loading on the system.
The commercial impacts and benefits associated with grid-scale storage are further explored in
Section 5.4.5.
Table 3 summarises the advantages and disadvantages of several different storage technologies,
while
Table 4 lists the general requirements of energy storage applications.
| 65
Table 3 – Advantages and disadvantages of storage technologies [10] . (Reproduced courtesy of Elsevier)
| 66
Table 4 – General energy storage application requirements1, Electric Power Research Institute (EPRI) [13] .
The characteristics of different technologies and their appropriate grid services are listed in
| 67
Table 5, which highlights that characteristics and applications can vary drastically for different
storage technologies. For example, the limited energy capacity of flywheels constrains these
devices to provide power services, such as regulation. In contrast, technologies such as PHS and
CAES have a high energy capacity; therefore, they have historically been used for services with
longer time scales, such as energy arbitrage. However, recent improvements in the response
time, cycle efficiency and power-to-energy ratios have expanded the potential applications of
PHS and CAES technologies. For example, variable-speed pumps on a PHS system can provide
frequency regulation [5] .
Table 5 also shows that different battery technologies have a wide range of energy capacity
levels, charge/discharge durations, response times and efficiencies. This variety makes batteries
suitable for many energy and power services. Details about the costs of different technologies
are outlined in [14] .
| 68
Table 5 – Grid-scale storage technologies [2] . (Reproduced courtesy of Elsevier)
| 69
In general, energy storage has the inherent flexibility to be used as a generation, transmission,
or renewable energy integration asset, or a combination of these. In a power system with
renewable energy resources, grid-scale storage can reduce effects of intermittent, uncertain
and non-dispatchable power sources in several ways (see e.g. [6] [7] [8] [9] . As discussed
throughout Section 3.2, the high penetration of intermittent renewable resource, together with
demand variations, has introduced many challenges to distribution systems. These include power
fluctuations, reverse power flow, voltage rise and LV stability. In such cases, large-scale storage
can prevent loss-of-load by providing firm capacity. It can also suppress power output
fluctuations by improving power quality or providing ancillary services such as regulation. When
the availability of wind or solar power does not align with demand, storage can facilitate time
shifting of loads.
For example, PV sources have been integrated with battery energy storage (BES) in a commercial
distribution system to create a dispatchable PV-BES source [6] . Advances in dispatchable BES
technologies provide an opportunity to make such non-dispatchable PV sources, such as
conventional generators, dispatchable. Reference [7] developed a mitigation strategy for solar
PV impacts by an effective use of distributed energy storage integrated with solar PV in the LV
networks. The storage consumes surplus solar PV power during PV peaks, and uses the stored
energy in the evening for the peak-load support. A charging/discharging control strategy
considers the current state of charge of the storage and the intended discharge period. It can
also reduce the impact of sudden changes in the PV output on the grid by absorbing generation
fluctuations during peak output periods.
Another critical role of storage in renewable energy applications is in offsetting the types of
voltage rise (or fall) caused by sharp renewable ramp rates (as seen in Section 3.2.5.1). This
action is often accomplished by co-locating storage with the renewable power source [11] . Co-
locating storage with wind farms can stabilise power output and regulate power levels in
response to either market or operational conditions [10] .
Coupling power electronics with energy storage can improve basic storage system capabilities,
power quality and grid stability [8] [9] . When power electronics include the capability to
provide both real and reactive power (VAR), the integrated storage system can provide voltage
support by both injecting or absorbing VAR to control voltage levels [12] . Compensating for
large voltage surges or drops (referred to as HV and LV ride-through, respectively) helps
maintain power quality and stability in the system [14] . Power electronics can also control
storage behaviour in response to AGC signals to correct frequency deviations [10] .
Recent work has investigated coupling energy storage with flexible AC transmission system
devices to provide a wider range of control options. For example, the importance of voltage-VAR
instability is analysed in [6] . Figure 8 – highlights how storage and VAR injection can improve
power quality and reduce the negative impacts of renewable energy. In this experiment, the PV
size is 4.33 MW and 2.39 MVAR, the BES power rating is 1.8 MW and the BES energy capacity is
13.5 MWh.
| 70
Figure 8 –Voltage profile of three different scenarios, without storage (No PV-BES), storage without VAR (PV-BES, unity
PF) and storage with VAR (PV-BES, optimal PF) [6] . BES = battery energy storage; PF = power factor; PV =
photovoltaic; VAR = volt-amps reactive
While energy storage systems can significantly contribute to planning and operation in power
systems, their widespread use has been limited to date. This is due to:
 relatively high technology costs
 a lack of deployment experience
 barriers and uncertainties caused by the present electricity market and regulatory
structures that were designed for conventional electricity systems
 technical limitations, such as the maximum limit of energy a given storage can hold.
For a recent international review of the implications of regulatory and electricity market
structures on the emergence of grid-scale electricity storage, refer to [15] and Section 4.
Key Finding: Grid-scale storage has the potential to provide a range of network
benefits relating to EG, including voltage management, energy balancing, and
improving network stability. These benefits can increase the flexibility, reliability and
efficiency of power delivery to consumers.
| 71
3.3.2.2 Distributed storage
Distributed storage is defined as an energy storage system that is connected deep within the
distribution network (typically at LV and often behind the meter). It is often connected for the
purpose of providing energy services to end customers, where the storage system can generate
in parallel with the grid. Services can range from storing bulk energy from local generation (e.g.
rooftop solar PV) and off-peak charging (energy arbitrage) to managing short-term intermittency
of EG and improving voltage and power quality with inverter-based control techniques.
In this section, we address the technical benefits relating in particular to managing the negative
impacts relating to EG. Section 5.4.3 details commercial and strategic benefits associated with
residential storage, in particular.
Many of the mitigation strategies for grid-scale storage discussed in Section 3.3.2.1 are also valid
for distributed storage. A strong case can be made for distributed storage to resolve EG-related
issues, by storing energy from a local embedded generator for use at a later time. This can be
broadly broken into two categories:
 fast storage, for smoothing intermittent local supply (both generation and load) and any
resultant voltage variability
 bulk energy shifting, to better balance load and generation (particularly for peak-
shifting).
As discussed in Section 3.2.2.6, anecdotal evidence from some DNSPs indicates that in certain
parts of the LV network, extreme penetrations of rooftop solar have led to daytime generation
causing peak reverse power flows, rather than the usual peak net load. In these cases, options
for managing these peak flows without significant direct network investment include: curtailing
the generation (either by using the energy in discretionary loads or simply dumping it), or using
distributed storage to store the energy for later discharge.
Some Australian DNSPs place limits on the maximum charging and discharging current allowed to
an energy storage system. This can restrict some undesired impacts of these systems, such as
their contribution to peak load or generation if not properly managed. However, this may also
limit the versatility and capability of the system, particularly in managing the intermittency (and
accompanying voltage effects) of high-penetration renewable generation, such as solar PV.
Distributed storage is well placed to reduce the impacts of intermittent supply or peak power
flow issues, due to its potential for co-location with the power flow to be balanced. This allows
local constraints to drive the control strategy for the storage. This may be compared with the
grid-scale case, in which the capability of storage to directly reduce impacts is restricted to
those occurring at the level where the storage is connected (or higher). Grid-scale storage does
have some capability to provide lower-network benefits; however, these are far more limited
than that of distributed storage. For example, in the case of competing voltage and power factor
impacts at different LV network points, a single MV-connected storage unit may not be able to
support all of these functions, while targeted storage could.
| 72
When distributed storage is either directly or indirectly under the control of a customer, it could
be expected to benefit the customer. In some cases, this might be expected to be to the
detriment of the local NSP; however, economic drivers of the end-customer can align with those
of the DNSP under many circumstances. In such cases, an energy storage system may provide
significant value to the network at no infrastructure cost. To make the best use of distributed
storage, utilities must provide incentives that will direct the customer towards controlling
storage in a manner that does not conflict with, and ideally supports, the network and any issues
it may be experiencing (EG-related or not). This can only occur when such incentives are
permitted under the regulatory environment.
For residential battery systems to have discharge-to-grid capability, they require a grid-
connected inverter (compliant to AS 4777) to convert the direct current supplied by the battery
into 240/415 V AC power for the grid. As with the grid-scale storage described above, modern
inverters have significant additional capability to improve network power quality through
techniques such as power factor correction and volt-VAR support. Any bidirectional battery
system should therefore be similarly able to assist with these functions. More information on the
capability of inverters to improve network power quality can be found in Section 3.3.4.1.
Network management of distributed storage

Several options are available for allowing DNSPs to manage distributed, grid-connected storage
assets for their own benefit. They are likely to receive the most direct benefit of an energy
storage unit if they can direct its charging and discharging directly. Such a control strategy is
used in [56] , where the power flows in and out of distributed energy storage systems are
controlled by a DNSP in conjunction with tap-changing transformers to manage voltage rise due
to high-penetration solar PV.
However, such central-control strategies — and distributed control strategies, which are
managed solely by the DNSP or other central agency — are unlikely to encourage consumer
uptake of these systems. This is key to enabling strong penetration of distributed storage
without requiring massive investment from DNSPs. A more likely scenario would involve a
network customer installing a battery system (perhaps coupled with a new or existing solar PV or
other generation system) to manage their own energy use and maximise the economic gain from
any EG they may own. A business case could be made to these customers whereby they may
cede some level of control over their storage system to the DNSP, perhaps in exchange for off-
peak charging tariffs for example. These systems can be used to provide network support as
described above without requiring large capital investment.
Such a form of control requires an interface that enables a DNSP to take temporary control of a
customer’s energy storage system. The AS/NZS 4755 suite of DR standards is currently
developing Part 3.5 – Operational instructions and connections for grid-connected electrical
energy storage (EES) Systems. This will expand the existing 4755 architecture to include
distributed energy storage systems. Adding a 4755-compliant demand response enabling device
(DRED) to a residential energy storage system will allow customers to use their system for their
own benefit the majority of the time, as determined by their contract with their DNSP). At the
same time, this allows the network to intervene to limit (or encourage) charging or discharging,
or provide reactive power support, when it is beneficial to the network to do so. A draft issue of
this standard for public comment may occur before the end of 2015. The current draft of AS/NZS
4777 Part 2 includes similar DR functionality, which will apply to all inverter energy systems and
include battery-coupled inverters, This standard is expected to be published in late 2015.
| 73
Electric vehicle storage
Distributed storage can be considered to include charge/discharge-controllable EVs, with or
without a bidirectional charging system. A residential EV charging station that supports vehicle-
to-grid discharge of the car’s internal battery can provide similar benefits to stationary
distributed storage, in the context of ensuring that the EV’s state of charge is sufficient when
the owner intends to drive it.
The ability for EVs to provide energy storage services for network support through their
participation in ancillary services markets is detailed in Section 3.3.3.2.
Storage devices as loads

Some types of energy storage devices do not (and in many cases, cannot) generally discharge to
the grid. Examples of these include uninterruptible power supplies (UPS), consumer electronics
such as laptops and mobile phones, and non-discharge-capable EVs. Such devices may have a
significant effect on networks, particularly when taken as aggregated loads; however, they are
generally similar to other power electronics systems, as detailed in Section 3.3.4.1. They may
also provide benefits if properly network integrated and under some level of network control, by
acting as discretionary load for the purposes of managing EG.
Some of these devices may develop into systems that overlap with the above definition of
distributed storage in the future. Hybrid UPS/backup power systems already fulfil some of the
role of distributed storage, as described above. As noted above and in Section 3.3.3.2, V2G-
capable EVs have been developed for use in network support. However, these are currently
struggling to find either manufacturer support or a market for their capabilities.
Australian trials of distributed storage

Very few trials of distributed storage have been conducted in Australia. Recently, Ergon
deployed 10 residential energy storage systems alongside home energy management systems
with alternative tariffs to investigate how these tools might affect consumer behaviour and
reduce peak-load concerns. The trial is still in the preliminary stages, and no conclusive results
have yet been published.
As part of their Smart Grid Smart City technology trial, Ausgrid deployed 60 10-kWh zinc bromine
distributed energy storage systems through the Newington, Elermore Vale and Upper Gundy
areas of eastern New South Wales. The results from this trial were mixed, and noted significant
technical issues relating to the maturity of the battery technology that negatively affected the
outcomes of the project. These issues included:
 high passive energy consumption
 difficulty matching peak load due to limited capacity, particularly for long-duration peaks
 poor battery performance at high ambient temperature
 reduced system performance after the first five months.
The trial identified that the systems deployed were smaller than would have been ideal for a
residential system targeting peak-load shifting. Consequently, the results have better informed
industry regarding appropriate sizing of the systems, as well as provided guidance about
technology selection. While no negative power quality impacts were identified as part of this
trial, the dominating factor was determined to be the strength of the Ausgrid network in the
trial locations.
Although the trial’s results imply significant limitations of the technology, these issues are in
part due to the technology’s immaturity. In particular, the reduced system performance after
| 74
prolonged deployment and the significant reduction in capacity at high temperatures may have
related to technical issues that could be improved upon or potentially solved by further
development.
Key Finding: Distributed storage can provide numerous benefits to networks, including
improved management of voltage and power flows, peak load and generation
management, and reactive power support. Customer-owned storage has the capacity to
provide these benefits alongside direct benefits to the customer, despite not being
network-owned.
3.3.3 Ancillary services
This section discusses how ancillary services can be implemented to address some of the
network impacts outlined in Section 3.2. The impacts that may be mitigated by additional
ancillary services include:
 energy balance
 ramp rates
 thermal loading
 losses
 demand (load) prediction
 bidirectional power flow
 phase imbalance.
Traditional ancillary services, such as fuel-powered operating reserves, are expensive to run.
This section will explore contemporary alternatives, such as using EG or EV-distributed batteries
for grid balancing, increasing spinning reserve capacity and load following.
3.3.3.1 Embedded generation
With increasing penetration of EG, particularly intermittent renewables such as solar and wild,
system stability issues such as those discussed in Section 3.2.4 will become more likely, possibly
causing power outages and damage to electronic devices [28] [29] . These issues may be
addressed through additional ancillary services to ensure optimal grid operation.
A simulation study of a standard IEEE 13-bus network [30] investigated the feasibility of wind
and solar-based EG to provide voltage and stability support ancillary services to a microgrid with
70% of total generation from renewables. The study found that for microgrids to operate safely
(either independently or as part of a larger grid), each EG unit must contribute to ancillary
| 75
services for frequency and damping support. The simulation showed that the voltage stability of
the network improves during a disturbance or increase in load demand.
Reference [31] proposes that household rooftop PV systems could provide active power ancillary
services. The study identified that even with the intermittency of PV generation and uncertainty
related to household energy consumption patterns, there are potential economic benefits for
households to provide active power support.
3.3.3.2 Electric vehicles
Several studies [32] [35] have explored customer charging behaviours and driving patterns, as
well as the technical performance of EVs, to assess the ability of EV fleets to provide a range of
ancillary services. They show that V2G-capable EVs may profitably provide power to the grid and
can be more attractive than conventional ancillary services. They can reduce operating costs, as
well as generate revenue for utilities. EVs can provide the following services:
 voltage and frequency regulation
 load balancing
 spinning reserve
 load following.
Traditionally, spinning reserve is supplied by the free generating capacity of committed units.
Spinning reserves are online, synchronised generators that can rapidly increase their generation
to respond immediately to unexpected power fluctuations (e.g. fast changes in load demand or a
sudden loss of generation). The intermittent nature of renewable energy technologies means
that more reserve may be required to maintain the stability and reliability of the power system.
The expected uptake of EVs in the next 10—20 years promises the opportunity for distributed
batteries in EVs to link to a smart grid, thereby increasing the power system spinning reserve
capacity. This could be achieved by either dynamically reducing the EV charging load on the
grid, or by increasing V2G generation.
EV-based reserves have several advantages over conventional fuel-powered reserves, including
high ramp rates and negligible start-up costs. Because EVs can act as both a load (while
charging) and a source (V2G), they could be used as load-following generators. If there is an
increase in the power generated from renewables or a drop in demand, EVs can store the surplus
energy through battery charging to minimise the mismatch between generation and loads. This
flexibility of EVs may reduce many EG network impacts. As Australia’s EV uptake grows, they can
facilitate and enable higher penetrations of EG (including but not limited to renewables).
Significant environmental benefits would also result from using EVs (ideally powered from
renewable sources) instead of traditional fuel-powered reserves.
One of the technical barriers to EV-based ancillary services is the accelerated degradation of
batteries through charging and discharging. Consequently, most EV manufacturers do not
significantly support V2G capability. However, although V2G may reduce the lifetime of
batteries, this cost is outweighed by the value that EV-based ancillary services can bring to
networks [33] . With coordinated and efficient charging/recharging strategies, EV-based
ancillary services are overall more economical for both utilities and customers [33] , [35] .
There are several other barriers to increasing the penetration of V2G operations in Australia. For
example, the charging and discharging current for EVs is typically limited by Australian DNSPs to
20 A for a single-phase charger, to minimise the likelihood and potential for peak-load effects,
particularly in high-penetration areas. EVs with larger capacities (e.g. Tesla) can use
| 76
significantly larger charging and discharging capacities, which may improve the economic case
for these systems in Australia.
To export stored energy from EVs to the grid, an inverter compliant with AS 4777 is also
required. This comes at substantial additional cost. While many houses will already have a
compliant inverter on the premises due to rooftop solar PV, these are generally designed with a
single supply point and no charging capability, indicating they are not able to be retrofitted to a
V2G system. There are also significant challenges in metering and communications for control of
the devices. Standards such as the now-paused AS/NZS 4755.3.4 (see Section 3.5.2.3) and the
international Open Automated Demand Response (OpenADR) will hopefully assist in future
development of these EV-based ancillary services.
3.3.4 Intelligent networks

This section covers a broad range of topics that fall under the traditional area of the
‘smart grid’. Some of these techniques have previously been proposed to support the network in
other ways. However, their relevance to the rising concern of reducing the impact of EG
warrants their renewed attention.
3.3.4.1 Power electronics solutions
Power electronics solutions have been regularly touted as having the capacity to strongly control
voltage variability and improve power quality at many points in the distribution network. Several
power electronics devices may be able to help manage voltage and power quality issues in
distribution networks. Of these, static synchronous compensators (STATCOMs), smart inverters
and SVCs are particularly capable in this area, and are detailed below.
STATCOMs
STATCOMs are power electronics devices typically based on a voltage-source inverter, which is
controlled in such a way as to provide one of a variety of power quality improvements in
distribution networks. STATCOMs have been trialled numerous times in recent years in Australia
and worldwide. Trials have ranged from relatively small deployments, in the order of tens of kVA
[58] , to multi mega-volt ampere (MVA) systems connected at the substation level for feeder-
level support [61] . An Australian trial [59] demonstrated significant benefits after installing a
20-kVA STATCOM in the LV network, including reduced voltage variation and reduced voltage
imbalance.
STATCOMs can provide simple reactive power support, slow and fast-reactive voltage control and
counter-harmonics, and also improve voltage balance between phases [62] . They can be
configured to include (or be paired with) energy storage such as batteries, which allow the unit
to provide real power support at peak-load times.
| 77
Passive static VAR compensators
While typically consisting of dynamically switched passive components, rather than active power
electronics, passive SVCs can be applied in a manner that performs many of the benefits of
STATCOMs and smart inverters described above. SVCs are well established and understood
devices that are already being used in networks to provide both voltage support and power
factor correction at transmission and distribution level. Many passive SVCs include limited
capability for filtering harmonics. However, these filters are often sized to deal directly with the
harmonics injected by the SVC itself, and therefore do not provide any real harmonic
improvements to the wider network.
Unfortunately, the capital cost associated with passive SVCs may make them prohibitive at a
small-scale distribution level. They are worth considering at larger scales, particularly when
they can be co-located with intermittent embedded generators that cannot provide their own
reactive power support. This may be due either to the function not being available, or because
the inverters are already operating at rated power and so cannot expend any for VAR support.
Smart inverters
With the increasing penetration of inverter-coupled EG in distribution networks comes a host of

grid-connected inverters. While these are often seen as causing significant issues for DNSPs,
most inverters are able to perform many of the functions of a STATCOM, without the upfront
infrastructure cost. In this way, intermittent or uncontrolled generation such as rooftop solar PV
can be modified to facilitate its own high penetration. Inverters that can improve power factor
and support voltage are often termed ‘smart’ inverters.
Most currently available inverters cannot act in this fashion, purely because their firmware does
not support a control strategy to do so. To support this capability, a firmware upgrade is often
required. We can learning from the German experience of attempting to simply retrofit the
existing inverter fleet with a modified LV disconnection limit, wherein cost became a near-
prohibitive issue. Instead, it may be simpler to encourage incoming embedded generators to
include smart inverters as a baseline capability, and to replace failed existing inverters primarily
with smart inverters.
The current revision of AS/NZS 4777 Part 2 includes the capability to offset voltage rise caused
(at least in part) by the energy system the inverter is connected to. This is initially done by
gradually increasing the amount of reactive power absorbed by the inverter until a nominal kVA
| 78
threshold is reached. If voltage issues continue once this threshold is reached, the inverter can
move to ramp down its power output, wasting potential supply energy. This is intended as a last
resort, since foregone energy can represent a significant cost to the system owner.
One limitation of using distributed inverters in this manner is that voltage control is often most
required at times of high export of energy, when inverters are likely to be exporting close to
their rated apparent power output. This is particularly the case for high-penetration rooftop
solar and the geographic correlation of power output. While this limits the capacity for smart
inverters to improve power quality, it does not eliminate it. For example, a system of voltage-
source PV inverters has been used overnight to control voltage fluctuations caused by the
intermittency of a co-located wind farm [63] .
Key Finding: Power electronics solutions, such as STATCOMs and smart inverters, have
the capability to mitigate power quality issues relating to EG, including managing
voltage ramp-rates and excursions. In some cases, they can also reduce harmonic
content.
3.3.4.2 Demand response
DR provides an excellent resource for managing power flow issues within distribution networks.
For example, the PeakSmart controlled air-conditioner trials by Energex and Ergon demonstrated
that with appropriate incentives, end customers were often willing to allow DNSPs to take
control of their air-conditioning systems for relatively short periods of high demand.
DR is at different levels of maturity across Australia; timing may be midterm for some
jurisdictions, while others are current. Recent standards developed in Australia and
internationally have provided significant guidance into how DR can be enabled technically.
Hundreds of models of air conditioners from 15 manufacturers [64] can now support the AS/NZS
4755 DR architecture. Similar standards have been developed for pool pump controllers and
electric hot water systems, with the prospect of battery storage and EV chargers also being
investigated for inclusion. Note that the EV standard AS/NZS 4755.3.4 has been put on hold to
review international development (see Section 3.5.2.3. In addition, some EV manufacturers
indicate a preference for incentivised charging when beneficial for networks through appropriate
tariffs or other price signals, rather than through direct charge control by DNSPs.
The primary opportunity for DR to address the impacts of EG is its ability to control discretionary
loads at times when it is beneficial to the network. EG acting as a load (such as user-controlled
storage charging) may otherwise cause peak-load conditions at times when this may not typically
be the case. In this circumstance, being able to reduce load through DR could be a significant
benefit.
| 79
A more likely case is significant excess energy in the distribution network due to oversupply of
EG output, such as from rooftop solar PV. In these cases, the ability to take advantage of this
oversupply by enabling discretionary loads to be switched on could benefit the network and the
customer. The network could better manage issues such as voltage rise and reverse power flow,
while the customer may be able to use the energy generated by their solar system in their own
home.
Currently, the only AS/NZS 4755-compliant systems that can turn on discretionary loads are
electric hot water heaters and pool pumps. It is envisaged that this capability will also be
included in the respective energy storage and EV standards, if and when they are published.
Unfortunately, electric hot water and pool pump systems have yet to see significant uptake of
the 4755-style control architecture, unlike air conditioners.
Reference [65] shows that air conditioners have significant capability to act as a discretionary
load. They can pre-cool residences at times when energy may otherwise be dumped, as well as
improve user comfort and reduce overall energy consumed. The current version of AS/NZS 4755
precludes the capability for air conditioners to push on load in this way. However, there may be
scope to provide these capabilities to DR enabled air conditioners outside the AS/NZS 4755
architecture, or to modify the standard so that air conditioners that do support these
capabilities are not deemed non-compliant. While the latter may appear to be a better solution
from a market perspective, concerns have been raised about consumer issues, with DNSPs
switching on air conditioners while the occupants are not home or do not require it. Such cases
should be taken up to address these issues.
Other issues also affect the uptake and support of regulators in enabling DR for DNSPs, or that
limit DNSPs’ benefits relating to DR. For example, if not properly managed, the return of
generation or load at the end of a peak event can induce a peak load of a similar magnitude
(and conceptually higher) than if no demand event has been called.
DNSPs should use DR tools to address issues that are appropriate for the scale of the technology.
A localised demand or generation peak at a distribution substation can be effectively managed
by calling a targeted DR event, provided that sufficient penetration of DR-enabled loads or
generators are available on the small network segment being managed. Managing a much
broader peak (e.g. a zone substation) would require a similar proportional penetration on a
much larger network segment, and may have unintended commercial consequences due to the
impact on consumers. This highlights the need to critically analyse the requirements of a given
peak, and the potential implications of managing it using DR.
3.3.4.3 Ramp-rate management
Methods to minimise or reduce the impacts of increased ramp rates of net load as a result of
high penetration, intermittent EG include energy storage, advanced inverter control, active
power curtailment, load control and spatial diversity of EG resources. Increasing system
| 80
flexibility (or decreasing flexibility requirements) will also support increased levels of
intermittent EG.
System flexibility
System flexibility can be described as the ability of an aggregated set of generators connected
to a network to respond to the variation and uncertainty in net load [69] . Not including
techniques such as energy storage and load shedding, measures of system flexibility [70]
include:
• Minimum load
o the lowest power output that conventional generation sources can manage
without being forced to turn off
o the lower the minimum load, the less likely generation will be forced to shut
down during periods of light net load brought on by high penetration of EG
• Start/stop speed
o the time required for generators to start-up and shut down
• Ramp rates
o the rate at which generation can increase or decrease output
• Level of spinning reserve
o the more spinning reserve, the greater the ability to accommodate short-term
fluctuations in net load (generally at a higher cost)
• Load balancing area
o the area over which generation output is matched to demand
o for the Australian NEM, this is the entire east coast of Australia, including South
Australia and Tasmania.
Ways to increase system flexibility include balancing the generation portfolio and introducing
more flexible conventional generation. Power systems can also be redesigned to handle reverse
power flow from distributed EG.
Reducing net load variability would reduce the need for increased flexibility of the system.
Measures suggested in [23] include:
• energy storage
• load control
• increased control and communication
• ability to curtail intermittent EG
• spatial diversity of the resource.
Experience with systems containing large amounts of intermittent renewable generation have
shown that flexibility of the portfolio balance is crucial for economic and stable operation [71] .
It is predicted that generators that can vary their output as well as cycle on and off quickly will
deliver high value in future energy systems, because they are best able to respond to the
increasingly variable outputs caused by growing renewable penetration. Reference [70] that
Californian grid operators should plan for a combination of flexible generation and import-export
agreements to allow for a smaller minimum net load (load minus distributed EG) and greater net
load variability. The report also suggested that the evaluation of the generation flexibility of a
system should be done at the ‘load-following’ time scale in relation to the variability of the net
load.
| 81
Curtailment
To manage the increased fluctuations in net load introduced by intermittent generation,
curtailment of EG generation would be required if the generation portfolio is not sufficiently
flexible, and other measures such as those mentioned above are insufficient.
A study into the degree of curtailment of variable generation required for different mixes of
wind and solar generation for varying levels of system flexibility was performed on the Electric
Reliability Council of Texas (ERCOT) system [69] . The curtailment requirements for varying
levels of minimum load for penetration levels of up to 80% renewables were analysed. The
simulation results of two scenarios are shown in Figure 9: one in which the minimum load
capability is 21 GW (approximately 35% of annual peak load), and another in which minimum
load capability is 13 W (approximately 22% of annual peak load). In the first case, 21% of
generation from renewable energy sources had to be curtailed, due to the minimum generation
constraints of inflexible wind and solar generation, which contributed only 20% of the energy
demand. In the second case, less than 3% curtailment was required by increasing flexible
generation, with renewables contributing 25% of the system’s annual energy. For a wind
penetration level of 50%, curtailment dropped from 50 to 20% for a 10% increase in system
flexibility (i.e. reduced minimum loading capability). Inputs into the study would have included
the load profile for ERCOT network, the local wind generation profile and the theoretical
minimum load levels.
The effect of energy storage on increasing system flexibility was also investigated in the ERCOT
study, the results of which can be seen in [22] . Insufficient transmission capacity can reduce
flexibility, as shown by a real-world example in which insufficient transmission from west Texas
to loads in the east resulted in 17% curtailment of wind generation in 2009.
The ERCOT study used theoretical values for system flexibility. For this study to be applicable in
Australia, it would need to use the local generation mix and models of solar and wind generation
based on local wind and irradiance data. Work of this type would be of great assistance in
determining possible penetration levels of intermittent EG for local power system networks. A
study could also investigate if, and to what extent, the introduction of energy storage would
increase system flexibility compared with introducing more flexible generation.
| 82
Figure 9 – Impact of system flexibility on curtailed energy, Electric Reliability Council of Texas [69]
| 83
3.3.5 Enabling technologies
Some mechanisms for reducing the impacts of EG require supporting techniques to function
optimally (or in some cases, at all). This section details the technologies that could best enable
the integration and uptake of the mechanisms covered in previous sections of this report.
3.3.5.1 Renewable energy prediction
Managing increased levels of intermittent renewable energy will require accurate prediction of
solar and wind availability. Reliable forecasts are important for managing both the intermittency
and uncertainty of renewable generation, and should be incorporated into network planning and
grid and market operation, including accurate generator unit commitment scheduling.
Prediction models can be broadly categorised as either statistical and physical [19] . Statistical
models apply statistical methods on existing time series of the resource, and do not involve any
physical modelling. In contrast, physical models include a physical modelling of the atmosphere
based on different types of atmospheric data. Hybrids of physical and statistical models are also
common.
Figure 10 illustrates the recommended prediction methods for solar, wind and waves on
different temporal and spatial domains; some of these are also listed in Table 6. As indicated,
physical models are preferred to statistical models for long-term forecasts; however, there are
some exceptions for solar. Total sky images are sometimes applied for short-term,
high-resolution forecasting.
| 84
Figure 10 – Recommended prediction methods for solar, wind and wave resources in different temporal and spatial
domains. [19] TSI = total sky images
Prediction at various timescales is also required. For example, improved day-ahead prediction of
renewable resources is required for more accurate unit commitment, and accurate forecasts are
needed to predict the rapid ramping of solar system output.
Table 6 – Some renewable resource prediction models reviewed in [19]
Resource Method Site Metrics
Solar Physical: satellite images (SI) and NWP, Six sites across the United States RMSE
(NDFD)
Solar Physical: NWP (WRF) Andalusia, Spain RMSE
Solar Statistical: ANN Ajaccio, France RMSE
Solar Statistical: ANN + GA California RMSE
Solar Hybrid: ANN + SI + lagged GHI Spain RMSE
Wave Physical: SWAN statistical: spectral model Four different sites in Atlantic and one MAE
in Pacific Ocean
Wave Statistical: mix of neural network and Four different locations in the Pacific Mean
regressions. Two different time horizons Ocean absolute
analysed: short term and long term percent
error
Wind Statistical: Markov-switching Two different offshore parks in RMSE

autoregressive model Denmark
Wind Physical: combined physical and statistical 11 wind farms in Ireland RMSE
(Fuzzy-NN)
Wind Statistical: ARIMA-GARCH 64 wind farms in Ireland RMSE
ANN = artificial neural network; GA = genetic algorithm; MAE = mean absolute error; NWP = numerical weather
prediction; RMSE = root mean squared error; SI = solar irradiance;
| 85
In Australia, forecasts of solar generation would be required up to two years out, as is currently
required for wind generation. Forecasts should include information about the expected output,
such as the degree of uncertainty, to indicate particularly volatile periods. Since 2008, AEMO has
used the Australian Wind Energy Forecasting System, which provides a wind forecast every five
minutes. The system was developed in response to the growth in intermittent generation in the
NEM, and to facilitate operation of the market.
In May 2014, the first phase of the Australian Solar Energy Forecasting System was installed and
commissioned into the live market system at AEMO [18] , while the scientific program to design
improvements was completed in June 2015. This first phase provided an operational system that
uses basic forecasting based on weather forecast products and statistical techniques, to cover all
of the AEMO-required prediction timeframes (from five minutes to two years). The system caters
for large-scale PV and solar thermal plants, enabling the integration of solar energy generation
at all scales into the national grid and allowing operators of larger systems to participate in the
NEM.
The solar forecasting system is configured as an extension to the wind forecasting system, which
has been successfully operating within AEMO market systems [18] . Without such forecasting
systems, wind and solar renewable energy generation will be subject to increasing levels of
curtailment. This will undermine both their viability and their significant contribution to
greenhouse gas reduction. Note that extensions to forecasting systems to cover embedded
renewable generation are likely to be more pressing as penetration levels increase.
Solar intermittency is believed to be more predictable than wind intermittency, because it is
affected by fewer climatic elements [16] . As discussed throughout this report, however, the
ramp rates for solar (mainly PV) are potentially significantly higher than wind, due to the lack of
system inertia. Solar output is more predictable in the very short term, because the movement
of clouds is visible; however, accurate prediction of long-term solar output is needed to
determine ways to effectively compensate for solar intermittency.
Key elements that may help to drive solar prediction across different time scales include [16] :
 temperature, irradiance and location
 correlation between weather patterns in nearby geographical locations
 ability to obtain samples at all required time intervals (i.e. seconds, minutes, hours,
days)
 accuracy of data
 cross-checking of data with satellite data
 seasonal data to cater for seasonal variations
 sky-view cameras for large-scale solar farms to monitor cloud movement
 a large network of ground-based monitoring stations that correlate satellite data with
their output, which should be maintained for several years.
Key Finding: Improved prediction of renewable generation, informed through

dedicated metering, can improve EG integration by increasing the utilisation of assets
such as storage and solar photovoltaic (PV)-controlled air conditioning. It can also
inform decisions on power system unit commitment and network planning. These
benefits can be broadened beyond EG to medium and large-scale generation.
| 86
3.3.5.2 Voltage flicker identification and control
Typically, wind farms are connected to MV distribution feeders (e.g. 33 kV). A short-term flicker
severity (Pst) of 0.35 is considered acceptable for wind farms installed in distribution networks,
while for wind farms connected to transmission networks, the Pst at the PCC should be below
0.30 [20] . Refer to Section 3.2.5.3 for the definition of Pst.
Although this section summarises results only from wind farms, the general findings are
applicable to other types of EG.
When flicker emissions are higher than the established limits, strategies for mitigation are
required. Table 7 depicts strategies for reducing the flicker emission of wind farms, highlighting
the main advantages and disadvantages of each. The first two strategies do not actively reduce
the magnitude of voltage variations, although they do allow it to be reduced. However, these
strategies are expensive and can generate environmental impact. Strategies that actively reduce
the magnitude of voltage variations are more effective. These include using storage devices to
smooth the net active power injected to the grid, and exchanging reactive power with the grid.
The latter is preferred, because it does not need any further investment in additional
equipment.
Table 7 – Strategies for flicker emission mitigation reviewed in [22]
Strategy Advantage Disadvantage
Strengthen the grid and increase the Low losses High cost and environmental impact
voltage level
Increase the inertia of the wind Smoothing of the power generated High cost
turbines
Energy storage in the DC link of back- Low cost Low energy cpacity
to-back power converter of variable-
speed turbines
Short-term storage devices in the wind Fast response and active power The need of investments in additional
farm regulation capability equipment
Compensate voltage fluctuation by Fast response and active power The need of investments in additional
reactive power using FACTS devices regulation capability equipment
Compensate voltage fluctuation by Fast response and no need of Possible oversizing of the grid-side
reactive power using wind turbine additional equipment power converter of wind turbines
DC = direct current; FACTS = flexible alternating current transmission system
| 87
The authors of [21] and [22] studied flicker emissions in a DFIG-based wind farm, and found
that wind generator control strategies, wind generator reactive power capability and the
operating point of the power curve had large impacts on flicker emission. They propose a flicker
mitigation strategy that injects compensating reactive power to the network; its value is
affected by the transmission line ratio that connects the wind farm substation to the grid. This
strategy can be implemented as an integrated control approach with the main control strategy,
without significantly compromising the performance of the main control scheme.
The weighting between two strategies can be adjusted by varying the control parameters.
Hence, a wind farm developer can adjust the mitigation level and performance requirement
according to the utility’s standards. The short-term flicker severity and average reactive power
for a range of mean wind speeds for three main control strategies before and after implementing
the flicker mitigation strategy are shown in Figure 11.
Figure 11 - Flicker emission comparison between wind generator main control strategies for different wind speeds,
before (left) and after (right) implementing the flicker mitigation strategy [21]
In Figure 11, high short-term flicker severity can be seen for the reactive power dispatch
strategy compared with the voltage and power factor control strategies before and after the
flicker mitigation. As an example, at a mean wind speed of 14 m/s, the flicker severity (Pst) is
0.42 for the reactive power dispatch strategy before the flicker mitigation, while for the voltage
and power factor control strategy, it is only 0.016 and 0.10, respectively. In the reactive power
dispatch strategy, the DFIG reactive power capability is limited to 0.78 pu at full operating
speed, hence the full active power output. This leads to reactive power deficit under certain
wind conditions.
Figure 11 also shows that the reactive power requirement for the voltage control strategy has
progressively decreased, since higher wind speeds imply higher active power output. However,
the reactive power capability to maintain the voltage at a higher value (i.e. 1.05 pu) has also
| 88
decreased. Furthermore, reactive power for power factor control strategy increases at higher
wind speeds, due to the greater active power output of the wind generator. Flicker emissions
substantially decrease after implementing the flicker mitigation strategy. In particular, for the
reactive power dispatch strategy, short-term flicker severity (Pst) falls from 0.42 to 0.03.
References [23] and [24] propose individual pitch control and generation torque control
methods, respectively, to mitigate the flicker emission at different wind speeds in DFIG wind
turbines. The generator active power oscillation caused by turbulence, wind shear and tower
shadow effects, which leads to flicker emission, is significantly damped by these control
methods at both high and low wind speeds. The results that show the effectiveness of these
control methods are given in Figure 12.
Figure 12 – Comparison of flicker values after individual pitch control (IPC) and generation torque control (GTC) with
the original case (without flicker mitigation) [23] , [24] (Reproduced by permission of the Institution of Engineering &
Technology)
Voltage flicker evaluation

Another aspect of flicker mitigation is usually hidden behind the flicker control strategies:
effective flicker evaluation. According to the IEC standard 61000-4-15, the general functions of
flicker meter can be divided into several blocks, as shown in Figure 13 [25] .
Figure 13 – Block diagram of flicker meter functions, as shown in [25] (Reproduced courtesy of IEEE)
The whole solution procedure shown in Figure 13 is complicated and successive. Some inaccurate
estimation results may propagate measurement errors. An accurate solution procedure, based on
conventional Prony’s method, is proposed in [26] to deal with this problem. The main
characteristics of the proposed mechanism are:
 The variation trajectory of flicker components can be accurately tracked.
 The calculation of the evaluation index is robust to the interferences of harmonics,
interharmonics, fundamental frequency deviation and variations of flicker components.
| 89
 There is no need to tune parameters or compensate for the proposed mechanism.
 The computational burden associated with the conventional method is reduced.
The proposed solution procedure could perform an accurate calculation of flicker severity index
to meet the requirements in IEC 61000-4-15, where the maximum acceptable relative error is
5%. The relative error of the new method for different case studies is within 0.5%, compared
with around 8% and 11%, respectively, for conventional flicker meters and FFT methods.
Finally, a decrease in the power quality of renewable generation due to voltage fluctuations may
be mitigated by foreknowledge of its occurrence. To relate the quality of power produced from
small-scale renewable energy installations to the meteorological values that drive them, several
neural network methods are used [27] . Data from an experimental, small-scale, residential
installation that includes PV and wind power are used to build the models, and are then used to
estimate the values of short and long-term flicker severity. The error of short-term flicker values
is around 20% and 60% for 1 minute and 10 minute forecasts, respectively.
3.3.5.3 Smart meters
Smart meters are advanced interval-type metering devices that record customer energy
consumption (typically reported in half-hour intervals). They provide more information than
conventional meters and allow digital communications with electricity distributors.
The growing use of smart meters in Australia can enable the EG mitigation options discussed in
this report, and therefore facilitate higher penetrations of EG. Applications for smart meters
include:
 improved load prediction for residential and commercial customers, especially the
accuracy of short-term prediction [36] [37] , which allows utilities to plan resources and
improve the balance of energy throughout the grid
 DR to shift loads to times of peak renewable output, and thereby reduce reverse power
flow. Smart meters assist with both:
o DR programs for remote on/off control of loads; and
o flexible pricing, including time-of-use tariff structures
 outage identification and confirmation, which reduces the overall duration of the
interruption and damage to assets
 supply of data to manage voltage levels [28] [37] .
3.3.5.4 Separate metering for embedded generation and loads
As discussed in Section 3.3.5.1, accurate prediction of loads ensures the reliability and stability
of power systems, which is especially important under high penetrations of intermittent
| 90
renewable EG. Separate metering of EG and loads can improve the ability to forecast loads,
because the exact consumption can be measured instead of the net exports/imports.
Australia includes a significant penetration of both gross and net-metering systems. Separate
metering of load and generation, as facilitated by gross metering, provides considerably greater
insight into the makeup of network load and better facilitates independent prediction of load
and EG production. Net metering does not support this. New load-plus-generation techniques
would therefore need to be developed, as discussed in Section 3.3.5.5.
In 2001, the Australian Greenhouse Office published a document discussing different mechanisms
for metering embedded generators in Australia [38] , and the benefits and drawbacks of each.
While the advantages of gross metering (which allows direct measurement of EG) were
discussed, the report was broadly supportive of a true net-metering scheme to simplify
connection and reduce cost. As indicated above, such a scheme would not support the proposed
separate prediction of load and generation. The report did not appear to include the potential
benefit of separate metering in improving load prediction, most likely because it was published
before the rise of household PV in Australia, when embedded generators had low levels of
network impact.
3.3.5.5 Improving models for load prediction
An accurate estimation of net load is required to determine the necessary flexibility of

generation for a given penetration level of EG. Models characterising net load need to clearly
represent the expected variability during the timeframes of interest.
In the NEM, generators are scheduled and dispatched into production to match supply every five
minutes [72] , as shown in Figure 14. Under increasing renewable EG, NEM planners must
consider how much generation flexibility is required to manage intermittent renewable
variability. To assist in this process, the amount or extent of variability in net load that the
network might experience in a five-minute interval needs to be determined.
Figure 14 – A day in the National Electricity Market (NEM) [72]
For effective operation of the power system, and to ensure adequate power quality and reliable
supply of electricity, generation output must match load demand at all times. This is managed in
| 91
real time by the generator governors and AGC or similar systems, together with some online
generators providing spinning reserve. Accurate prediction of load demand is essential to
effectively manage and use both EG and conventional generation sources on the network, and to
establish suitable load management systems for any smart grid. Many studies of load prediction
consider EG as negative load, and defined net load as the system load minus the EG. The net
load is the total load that needs to be met by conventional forms of generation.
Various models have been developed to predict load demand over different time intervals.
These include neural network training using algorithms such as the steepest gradient descent
algorithm, Newton algorithms and Levenberg-Marquardt algorithm for short-term load prediction
for microgrids [73] . Alternatives to neural network approaches include statistically based
multiple-regression models, which are more transparent and readily describable than black-box
neural approaches.
Factors contributing to changing use patterns of electricity networks need to be taken into
account for accurate prediction of load demand, irrespective of approach. These include:
• increased penetration of intermittent renewables-based EG
• changes to customer behaviour in response to electricity prices and new tariff structures
(e.g. time-of-use tariffs)
• increased uptake of energy efficiency measures, partly encouraged by government
policies
• potential increase in the uptake and use of EVs and energy storage.
Forecasts of generator dispatch long-term load demand need to be more granular: at least on an
hourly basis, and possibly shorter. Better understanding of the factors listed above requires
customer load statistics, detailed load profiles and EG output data for individual homes and
buildings. These data also need to be more granular to be implemented in generator dispatch
load prediction models, thereby allowing accurate forecasts of load demand at short time
intervals.
Key Finding: Improved network load prediction techniques that incorporate increased
penetration levels of EG are necessary to retain the current benefits of short, medium
and long-term prediction of net load in a high-penetration EG environment. The
prediction techniques also need to incorporate any other factors that may be
contributing to changing utilisation patterns of electricity networks.
| 92
References
[1] Cavanagh, K., Ward, J.K., Behrens, S., Bhatt, A.I., Ratnam, E., Oliver, E., Hayward, J.,
2015, ‘Electrical energy storage: technology overview and applications,’ CSIRO, Australia.
[2] Castillo, A., Gayme, D.F., 2014, ‘Grid-scale energy storage applications in renewable
energy integration: A survey’, Energy Conversion and Management, vol. 87, pp. 885-894.
[3] ‘Use of energy storage to increase transmission capacity,’ EPRI, Palo Alto, CA, United
States, 2011, 1024586.
[4] ‘Application of storage technology for transmission system support: Interim report,’ EPRI,
Palo Alto, CA, United States, 2012, 1025418.
[5] Scaini, V., 2012, ‘Grid support stability for reliable, renewable power,’ Technical report
WP083002EN. Eaton Corporation.
[6] Hung, D.Q., Mithulananthan, N., Bansal, R.C., 2014, ‘Integration of PV and BES units in
commercial distribution systems considering energy loss and voltage stability’, Applied
Energy, vol. 113, pp. 162-170.
[7] Alam, M.J.E., Muttaqi, K.M., Sutanto, D., 2013, ‘Mitigation of rooftop solar PV impacts
and evening peak support by managing available capacity of distributed energy storage
systems’, IEEE Transactions on Power Systems, vol. 28, no. 4, pp. 3874-3884.
[8] Amarasekara, H.W.K.M., Meegahapola, L., Agalgaonkar, A.P., Perera, S., 2013, ‘Impact of
renewable power integration on VQ stability margin’, 2013 Australasian Universities
Power Engineering Conference, AUPEC 2013, pp. 1–6. Hobart, Australia, 29 Sep. - 3 Oct.
2013.
[9] Collins, L., Ward, J.K., 2015, ‘Real and reactive power control of distributed PV inverters
for over-voltage prevention and increased renewable generation hosting
capacity’, Renewable Energy, vol. 81, pp. 464-471.
[10] Ayodele, T.R., Ogunjuyigbe, A.S.O., 2015, ‘Mitigation of wind power intermittency:
Storage technology approach’, Renewable and Sustainable Energy Reviews, vol. 44, pp.
447-456.
[11] Ibrahim, H., Ilinca, A., 2013, ‘Techno-economic analysis of different energy storage
technologies,’ In: Energy storage-technologies and applications, InTech. Rijeka, Croatia,
pp. 1–40.
[12] Enslin, J.H.R., 2014, ‘Integration of photovoltaic solar power - The quest towards
dispatchability’, IEEE Instrumentation and Measurement Magazine, vol. 17, no. 2, pp. 21-
26.
[13] Rastler, D., 2010, ‘Electricity Energy Storage Technology Options: A white paper primer
on applications, costs and benefits,’ Technical report 1020676. Electric Power Research
Institute (EPRI). Palo Alto, CA, United States.
[14] Carnegie, R., Gotham, D., Nderitu, D., Preckel, P.V., 2013, ‘Utility Scale Energy Storage
Systems: Benefits, Applications and Technologies,’ Technical Report, State Utility
Forecasting Group.
[15] Anuta, O.H., Taylor, P., Jones, D., McEntee, T., Wade, N., 2014, ‘An international
review of the implications of regulatory and electricity market structures on the
emergence of grid scale electricity storage’, Renewable and Sustainable Energy
Reviews, vol. 38, pp. 489-508.
[16] Sayeef, S., Heslop, S., Cornforth, D., Moore, T., Percy, S., Ward, J. K., Berry, A., Rowe,
D., 2012, ‘Solar intermittency: Australia’s clean energy challenge; Characterising the
effect of high penetration solar intermittency on Australian electricity networks,’
Technical report, CSIRO.
[17] California Energy Commission, July 2007, ‘Intermittency Analysis Project: Appendix B
Impact of Intermittent Generation on Operation of California Power Grid’, prepared by
GE Energy Consulting (CEC-500-2007-081-APB).
| 93
[18] Marchment Hill Consulting, 2014, ‘Integration of Renewables into the Grid; Stocktake
Results’, prepared for the Australian Renewable Energy Agency.
[19] Widén, J., Carpman, N., Castellucci, V., Lingfors, D., Olauson, J., Remouit, F.,
Bergkvist, M., Grabbe, M., Waters, R., 2015, ‘Variability assessment and forecasting of
renewables: A review for solar, wind, wave and tidal resources’, Renewable and
[20] International Electrotechnical Commission, IEC 61000-3-7. Electromagnetic
Compatibility (EMC), Part 3: Limits – Section 7: assessment of emission limits for the
connection of fluctuating installations to MV, HV and EHV power systems, 2nd ed.; 2008.
[21] Meegahapola, L., Perera, S., 2014, ‘Capability constraints to mitigate voltage
fluctuations from DFIG wind farms when delivering ancillary services to the
network’, International Journal of Electrical Power and Energy Systems, vol. 62, pp. 152-
162.
[22] Girbau-Llistuella, F., Sumper, A., Díaz-González, F., Galceran-Arellano, S., 2014,
‘Flicker mitigation by reactive power control in wind farm with doubly fed induction
generators’, International Journal of Electrical Power and Energy Systems, vol. 55, pp.
285-296.
[23] Zhang, Y., Chen, Z., Hu, W., Cheng, M., 2014, ‘Flicker mitigation by individual pitch
control of variable speed wind turbines with DFIG’, IEEE Transactions on Energy
[24] Zhang, Y., Hu, W., Chen, Z., Cheng, M., Hu, Y., 2014, ‘Flicker mitigation strategy for a
doubly fed induction generator by torque control’, IET Renewable Power Generation, vol.
8, no. 2, pp. 91-99.
[25] IEEE Recommended Practice Adoption of IEC 61000-4-15:2010, Electromagnetic
Compatibility (EMC)—Testing and Measurement Techniques- Flickermeter-Functional and
Design Specifications, IEEE Std. 1453-2011, 2011.
[26] Chen, C-I., Chen, Y-C., Chang, Y-R., Lee, Y-D., 2014, ‘An accurate solution procedure
for calculation of voltage flicker components’, IEEE Transactions on Industrial
Electronics, vol. 61, no. 5, pp. 2370-2377.
[27] Rodway, J., Musilek, P., Misak, S., Prokop, L., 2014, ‘Prediction of voltage related
power quality values from a small renewable energy installation’, Proceedings of 2014
Electrical Power and Energy Conference, pp. 76. Calgary, Canada. 12–14 Nov. 2014.
[28] Römer, B., Reichhart, P., Kranz, J., Picot, A., 2012, ‘The role of smart metering and
decentralized electricity storage for smart grids: The importance of positive
externalities,’ Energy Policy, vol. 50, pp. 486–495.
[29] Tuohy, A., Kaun, B., Entriken, R., 2014, ‘Storage and demand-side options for
integrating wind power,’ Wiley Interdiscip. Rev. Energy Environ., vol. 3, no. 1, pp. 93–
109.
[30] Alsokhiry, F., Adam, G.P., Lo, K.L., 2012, ‘Contribution of distributed generation to
ancillary services,’ 27th Int. Univ. Power Eng. Conf. (UPEC), pp. 1–5. London, United
Kingdom, 4–7 Sept. 2012.
[31] Clastres, C., Ha Pham, T.T., Wurtz, F., Bacha, S., 2010, ‘Ancillary services and optimal
household energy management with photovoltaic production,’ Energy, vol. 35, no. 1, pp.
55–64.
[32] Chukwu U.C., Mahajan, S.M., 2011, ‘V2G electric power capacity estimation and
ancillary service market evaluation,’ Power Energy Soc. Gen. Meet. IEEE, pp. 1–8. San
Diego, United States, 24–29 Jul. 2011.
[33] Yilmaz, M., Krein, P.T., 2013, ‘Review of the impact of vehicle-to-grid technologies on
distribution systems and utility interfaces,’ Power Electron. IEEE Trans., vol. 28, no. 12,
pp. 5673–5689.
| 94
[34] Kempton, W., Kubo, T., 2000, ‘Electric-drive vehicles for peak power in Japan,’ Energy
Policy, vol. 28, no. 1, pp. 9–18.
[35] Kempton, W., Letendre, S.E., 1997, ‘Electric vehicles as a new power source for electric
utilities,’ Transp. Res. Part D Transp. Environ., vol. 2, no. 3, pp. 157–175.
[36] Gajowniczek, K., Ząbkowski, T., 2014, ‘Short Term Electricity Forecasting Using
Individual Smart Meter Data,’ Procedia Comput. Sci., vol. 35, pp. 589–597.
[37] Day, P., Fabian, M., Noble, D., Ruwisch, G., Spencer, R., Stevenson, J., Thoppay, R.,
2014, ‘Residential Power Load Forecasting,’ Procedia Comput. Sci., vol. 28, pp. 457–464.
[38] Roche, D., Coelacanth Consulting, 2001, ‘Metering of embedded generators in
Australia’, Australian Greenhouse Office.
[39] Depuru, S.S.S.R., Wang, L., Devabhaktuni, V., 2011, ‘Smart meters for power grid:
Challenges, issues, advantages and status,’ Renew. Sustain. Energy Rev., vol. 15, no. 6,
pp. 2736–2742.
[40] Jackson, J., 2014, ‘Smart grids: an optimised electric power system’ Future Energy, pp.
633–651, 2014.
[41] Overbye, T.J., 2014, ‘Smart-grid-enabled distributed reactive power support with
Conservation Voltage Reduction,’ in 2014 Power and Energy Conference at Illinois (PECI),
pp. 1–5. Illinois, United States, 28 Feb. – 1 Mar. 2014.
[42] Park, G.L., Colony, R., 1977, ‘Voltage reduction as a means of reducing distribution
load,’ IEEE Trans. Power Appar. Syst., vol. 96, no. 2, pp. 628–634.
[43] Ellens, W., Berry, A., West, S., 2012, ‘A quantification of the energy savings by
Conservation Voltage Reduction,’ in 2012 IEEE International Conference on Power System
Technology (POWERCON), pp. 1–6. Auckland, New Zealand, 30 Oct. – 2 Nov. 2012.
[44] Amin, M., 2013, ‘IEEE Smart Grid,’[Online]. Available: http://smartgrid.ieee.org/march-
2013/813-the-self-healing-grid-a-concept-two-decades-in-the-making. [Accessed
14 May 2015].
[45] Amin, M., 2001, ‘Toward self-healing energy infrastructure systems,’ IEEE Computer
Applications in Power, vol. 14, no. 1, pp. 20–28.
[46] Moslehi, K., Kumar, R., 2010, ‘A Reliability Perspective of the Smart Grid,’ IEEE
Transactions on Smart Grid, vol. 1, no. 1, pp. 57–64.
[47] Angelo, C., Selejan, P., 2013, ‘Technologies of the self healing grid,’ in 22nd
International Conference on Electricity Distribution. Stockholm, Sweden, 10–13 Jun.
2013.
[48] Li, J., Liu, C.-C., Schneider, P., 2010, ‘Controlled partitioning of a power network
considering real and reactive power balance,’ IEEE Transactions on Smart Grid, vol.1, no.
3, pp. 261–269.
[49] Wang, S., Rodriguez, G., 2010, ‘Smart RAS (Remedial Action Scheme),’ in Integrated
Smart Grid Technologies (ISGT), Gaithersburg, United States, 19–21 Jan. 2010.
[50] Kezunovic, M., 2011, ‘Smart fault location for smart grids,’ IEEE Transactions on Smart
Grid, vol. 2, no. 1, pp. 11–22.
[51] Kazemi, S., Lehtonen, M., Fotuhi-Firuzabad, M., 2012, ‘Impacts of fault diagnosis
schemes on distribution system reliability,’ IEEE Transactions on Smart Grid, vol. 3, no.
2, pp.720–727.
[52] Metke, A., Ekl, R., 2010, ‘Security technology for smart grid networks,’ IEEE
Transactions on Smart Grid, vol. 1, no. 1, pp. 99–107.
[53] Jianhua, Z., Yue, Y., Rong, S., Shouxiang, W., 2014, ‘Switch configuration optimization
meeting the requirement of self-healing,’ in China International Conference on
Electricity Distribution (CICED), pp. 1524-1528. Shenzhen, China, 23-26 Sep. 2014.
[54] Holmlund, J., Kostiainen, A., Hakola, T., Gulich, O., Oy, A., ‘Practical experience of a
self-healing power system by means of the zone concept’.
[55] Kotagiri, A., Rose, A., 2011, ‘The future of energy storage in Australia’, QSI Online.
| 95
[56] Xiaohu, L., Aichhorn, A., Liming, L., Hui, Li., 2012,‘Coordinated Control of Distributed
Energy Storage System With Tap Changer Transformers for Voltage Rise Mitigation Under
High Photovoltaic Penetration,’ IEEE Transactions on Smart Grid, vol.3, no.2, pp.897–906.
[57] Ausgrid, 2014,‘Distributed Generation and Distributed Storage (DGDS)’, Smart Grid,
Smart City: Shaping Australia’s Energy Future - National Cost Benefit Assessment.
[58] Australian Energy Market Operator, 2011, ‘Network Support and Control Ancillary
Service (NSCAS) Description’.
[59] Energex, 2013, ‘STATCOM Study Trial Report’.
[60] Ergon Energy, 2014, ‘Ergon Energy Demand Management Innovation Allowance Report
2013-14’.
[61] Kato, T., Ito, T., Aihara, T., Namatame, S., 2007, ‘Development of a 20-MVA STATCOM
for Flicker Suppression’, Hitachi Review, vol. 56, no. 4, pp. 133-137.
[62] Betz, R., Palmer, R., Summers, T.J., Cook, B., Hogg, P., 2006, ‘H-Bridge Multilevel
STATCOM for Mining Applications’, Proceedings 2006 Australian Mining Technology
Conference.
[63] Varma, R.K., Khadkikar, V., Seethapathy, R., 2009, ‘Night-time application of PV solar
farm as STATCOM to regulate grid voltage’, IEEE Transactions on Energy Conversion, vol.
24, no. 4, pp. 983-985.
[64] Energex, https://www.energex.com.au/residential-and-business/positive-
payback/positive-payback-for-households/households/current-peaksmart-air-conditioner-
models, accessed 19 May 2015.
[65] Wall, J., Matthews, A., 2014, ‘Residential Air Conditioning and Demand Response: Status
of Standards in Australia and Future Opportunities’, International Journal of Application
or Innovation in Engineering & Management, vol. 3, no. 6, pp. 338–347.
[66] Sayeef, S., Heslop, S., Cornforth, D., Moore, T., Percy, S., Ward, J.K., Berry, A., Rowe,
D., 2012, ‘Solar intermittency: Australia’s clean energy challenge – Characterising the
effect of high penetration solar intermittency on Australian electricity networks’. CSIRO
report.
[67] Bebic, J., 2008, ‘Power System Planning: Emerging Practices Suitable for Evaluating the
Impact of High-Penetration Photovoltaics’, NREL Technical Report (NREL/SR-581-42297).
[69] Denholm, P., Hand, M., 2011, ‘Grid flexibility and Storage Required to Achieve Very High
Penetration of Variable Renewable Electricity’, Energy Policy, vol. 39, no. 3, pp. 1817–
1830.
[71] Lew, D., Miller, N., Clark, K., Jordan, G., Gao, Z., 2010, ‘Impact of High Solar
Penetration in the Western Interconnection’, NREL Technical Report (NREL/TP-5500-
49667).
[72] Australian Energy Market Operator, 2010, ‘An Introduction to Australia’s National
Electricity Market’.
[73] Jin, L., Cong, D., Guangyi, L., Jilai, Y., 2014, ‘Short-term net feeder load forecasting of
microgrid considering weather conditions’, Proceedings of Energycon 2014 Conference,
Cavtat, Croatia, 13–16 May, 2014.
[74] Marchment Hill Consulting, 2012, ‘Energy Storage in Australia - Commercial
Opportunities, Barriers and Policy Options’.
[75] Clean Energy Council, 2015, ‘Australian Energy Storage Roadmap’.
| 96
3.4 Options for maximising network benefits
3.4.1 Introduction
The previous sections of this chapter focus on the issues that EG may present to Australian
networks, and the mechanisms available to alleviate these issues now and into the future. This
could imply that EG is a problem to be solved, rather than an opportunity to be embraced.
As a reponse to this implication, this section underlines the many network benefits that well-
managed EG will deliver. It draws extensively on real-world trials in both domestic and
international markets to emphasise the direct benefits afforded by EG in:
 addressing peak load
 deferring network upgrades
 supporting frequency
 managing voltage
 delivering a high-reliability grid.
The document map on the following page summarises these benefits and many others,
highlighting the significance of each and where and when they are most likely to deliver
meaningful outcomes for Australian networks.
| 97
Document map
Chapter
Benefit Network level Network type Impact Timing
section
PV and peak-load
3.4.4
reduction
Power and voltage

3.4.5
levelling
Frequency support 3.4.6
Dispatchability 3.4.7
Electrical
infrastructure life 3.4.8
extension
Reduced operating
3.4.9.1
costs
Investment deferral 3.4.9.2
Power quality and

network service 3.4.9.3
provision
| 98
3.4.2 Overview of relevant literature
Given the detailed nature of the literature associated with this section, we have written a brief
summary of references to allow the reader to better understand the results therein. In the
interests of brevity, this work is contained in Appendix C – Maximising network benefits:
overview of relevant literature.
3.4.3 Embedded generation technologies

This report discusses many different types of EG technologies. These include ‘secure’ energy
source types, such as diesel engines, gas and steam turbines, and combined heat and power (CHP)
units, as well as ‘insecure’ intermittent energy sources, such as PV and wind. PV and/or wind
combined with energy storage, which offsets their intermittent nature, is also considered an EG
technology. Storage is not required for secure EG.
3.4.3.1 Storage
Forms of energy storage include batteries, pumped hydro, flywheels, supercapacitors and fuel
cells. The majority of studies investigating storage as EG, including all those referred to in this
report, focus on batteries.
Battery storage systems can be categorised as either a battery energy storage system (BESS) or
hybrid energy storage system (HESS). A HESS combines battery storage with a more responsive
technology, and in one case referred to here, also uses a STATCOM [11].
Battery energy storage systems

As noted in Section 3.3.2, BESSs can reduce the costs of peaking plants by reducing variable
operation (ramping requirements) and minimising start and stop operations. This increases plant
efficiency and reduces maintenance costs [10], and is a strategic benefit of energy storage that
needs to be recognised. Properly deployed energy storage will allow utilities to participate in
novel business opportunities, protecting them against inevitable challenges to their existing
business models [14].
BESSs can absorb and deliver both real and reactive power with subsecond response times. With
these capabilities, BESSs can mitigate solar power generation issues such as ramp rate,
frequency and voltage issues [7].
A BESS typically consists of a:
 battery bank
 control system
 protective circuitry
 power electronics interface, for AC-DC power conversion
 transformer, to convert the BESS output to the distribution or transmission system
voltage level.
The four-quadrant power electronic converter on a BESS can inject reactive power to maintain
either a power factor or voltage set-point on the bus. The benefits of improved power factor are
reduced power system losses and reduced line current upstream of the reactive power source
[7].
| 99
BESS battery types are summarised in [11]. Lead-acid batteries are the most common type of
BESS, because they are the cheapest. Their drawback is their lifespan, which is considered short
at typically less than 1000 cycles and is shortened by deep cycling. Lithium-ion batteries are
expensive, but have a longer lifespan than lead-acid batteries, and are not affected by deep
cycling. They are also more efficient (the efficiency of a battery is the ratio of the possible
energy output from a battery (discharging) for an amount of input energy (charging) [21]).
Sodium sulfur batteries are also more efficient than lead-acid batteries, have a longer lifespan
and are less affected by deep cycling (2500 cycles). They can also be scaled to the MW range,
but pose a safety risk due to their high operating temperature.
The vanadium redox battery (VRB) has a long service life and is considered an excellent
candidate for large-scale PV applications, although it is expensive. The power and energy ratings
of VRBs are independent of each other, which increases the battery’s scalability for PV systems
of different power ratings. The one major drawback of the VRB is its low efficiency when
operated at less than 20% of its rated power. Hence, one of the objectives of the VRB power
management system is to avoid its operation at low power levels [11].
Hybrid energy storage system

The benefit of a HESS over a BESS is that the strengths of each technology are combined to
improve system performance.
The challenge facing HESSs lies in the management of power sharing between the technologies.
Power sharing is assigned according to the response time of each technology. A smaller storage
device with a fast response time manages small, high-frequency power fluctuations, while a
larger storage device with a slower response time manages large, low-frequency power
fluctuations [11].
3.4.3.2 PV Combined with Storage
Figure 15 presents a simple one-line diagram depicting a BESS in parallel with a solar PV system.
Power conversion can typically operate in all four power quadrants, able to inject real power
and both inject and absorb reactive power. A BESS/PV system can also perform load shifting.
This reduces congestion, line losses, and pollution from inefficient peaking power plants that
only operate at times of peak demand [7].
Figure 15 - One-line diagram of battery energy storage system and photovoltaic (PV) system [7] (Reproduced courtesy
of IEEE)
The technical benefits of a PV/STATCOM/BESS system, summarised in [10], include:

• ancillary services
| 100
• voltage regulation support
• provision of MW and MVAR support during critical contingencies
• provision of LV ride-through by PV inverters during faults
• postponement of transmission upgrades in constrained circuits.
3.4.4 Photovoltaics and peak-load reduction
This section explores the extent to which PV generation coincides with peak load, and the
subsequent ability of PV to contribute to peak load reduction.
The ability for PV generation to reduce peak load depends on the naturally occurring
coincidence between the two. Load profiles vary according to load type; residential peak load
tends to occur during late afternoon/early evening, while commercial peak load tends to occur
near the middle of the day. Both load and generation profiles vary according to location and
season. As a result, the coincidence between PV generation and peak load varies according to
location and season.
Reference [4] shows how PV can reduce peak commercial loads. Figure 16 illustrates an
approximate reduction in peak load of 500 kVA for a 10% PV penetration level, and a 1.5 MVA
reduction for 20% PV penetration.
Figure 16 - Reduction in peak commercial load due to photovoltaic (PV) penetration [4] (Reproduced courtesy of IEEE)
AEMO has provided information about the contribution of rooftop PV data to reducing summer and
winter maximum peak demand in different regions of the NEM [22]. The data produced for summer
is analysed in Figure 17. No results for winter are shown, because the reduction of winter peak
demand due to PV is negligible. The summer results show that reduction varies according to year
and state. For example, in 2005—06, South Australia had the highest proportion of installed PV
| 101
contributing to reducing maximum peak demand, at nearly 50% of the installed PV. This value
dropped to less than 20% for 2012—13. New South Wales recorded the highest contribution, at
nearly 30% in 2012—13. The variation in contribution is due to varying customer load demand;
reduced contribution indicates that maximum peak demand is being pushed back in the day.
Figure 17 – Percentage of installed photovoltaic (PV) capacity contribution to summer maximum peak demand (%)
[22]. NSW = New South Wales; QLD = Queensland; SA = South Australia; TAS = Tasmania; VIC = Victoria
The following sections review DNSP studies of the effects of PV generation on peak demand in
their networks (including projections for higher PV penetration levels) and outline the extent to
which PV can reduce peak load in geographically diverse jurisdictions. The studies provide
evidence of reduction in peak load due to solar PV installations, with variation between DNSP
jurisdictions. Observed and projected peak-load reductions range from 0.3—2.95% depending on
geographical location and PV penetration level. Larger reductions (~2.95%) are seen for
simulated networks running at 50% PV penetration.
3.4.4.1 Western Power
Western Power estimates that PV systems reduced the 2011 system peak load by 0.62% (24 MW)
and the substation peak load by as much as 1.75% in some locations [23]. In 2012, a decrease in
peak demand of 1.33—1.72% due to PV was observed.
Increasing uptake of PV systems across the network is expected to reduce the system peak even
further. Western Power has undertaken various PV saturation trials and simulations, estimating a
projected peak reduction of 2.95% (135 MW) by 2017 (Figure 18) under higher penetrations of
PV. The peak load is reduced, and the new peak occurs slightly later in the day.
| 102
Figure 18 - Projections of peak load in 2017 [23]
Table 8 outlines Western Power’s summary of average impacts and recommended adjustment to
system peak forecasts. The ability of PV to reduce peak demand increases under higher
forecasted penetrations out to 2017.
Table 8 - Summary of average impacts and recommended adjustment to system peak forecasts [23]
Year Photovoltaic Forecast peak (MW) Peak reduction (MW) Peak reduction (%)
capacity (MW)
2013 243 4093.09 72.26 1.77
2014 299 4208.01 88.52 2.10
2015 357 4324.27 105.36 2.44
2016 414 4438.31 121.90 2.75
2017 472 4584.49 135.33 2.95
3.4.4.2 Ausgrid
Ausgrid [24] investigated reductions in peak demand resulting from the New South Wales Solar
Bonus Scheme. They concluded that the reduction was not significant enough to warrant
deferring investment in the electricity network. The greatest potential for network investment
deferral was identified at Charmhaven Zone Substation on the Central Coast. However, the
above average PV penetration was still estimated to be insufficient to offset the need for an
additional zone substation transformer. Ausgrid found that if the uptake of PV on this substation
were increased threefold, there could possibly be enough peak-load reduction to defer
upgrading the transformer for one year. The value of this deferral was estimated as equivalent
to a feed-in tariff of one cent per kWh over 10 years for only the PV customers supplied by
Charmhaven Zone Substation.
| 103
Table 9 and Table 10 list the estimated impact of PV and associated percentage reduction in
peak demand for different summer peak days and across the top five zone substations in 2011.
The reduction is as much as 1.2% at Avoca Zone Substation.
Table 9 - Estimated solar impact on Ausgrid summer peaks for summer 2010—11 [24]
Summer Date Day Time of peak Actual system Estimated % reduction

peak (Eastern peak (MW) solar impact
day standard time) (MW)
1 3/02/2011 Thursday 4:00 pm 6072 18.0 0.30
2 1/02/2011 Tuesday 4:00 pm 5922 22.5 0.38
3 2/02/2011 Wednesday 4:00 pm 5802 21.6 0.37
4 4/02/2011 Friday 3:30 pm 5553 23.9 0.43
5 31/01/2011 Monday 4:00 pm 5423 22.9 0.42
Table 10 - Summary of solar impact on summer peak 2011 at top five zone substations [24]
Zone Zone peak Zone peak Rated capacity of Estimated solar Estimated %
(MVA) solar connected impact at time of peak reduction
date and time at time of summer peak
summer peak 2010/11 (MW)
2010/11 (MW)
Pennant Hills 5/02/2011 17:30 83.75 1.33 0.38 0.5
Pennant Hills 3/02/2011 17:30 79.84 1.33 0.33 0.4
Pennant Hills 2/02/2011 17:00 79.85 1.33 0.49 0.6
Pennant Hills 4/02/2011 17:00 75.64 1.33 0.54 0.7
Avoca 5/02/2011 16:00 43.27 0.93 0.50 1.2
Avoca 3/02/2011 18:00 40.50 0.93 0.18 0.4
Avoca 1/02/2011 18:30 39.71 0.93 0.11 0.3
Avoca 2/02/2011 18:30 36.70 0.93 0.11 0.3
Nelson Bay 3/02/2011 17:30 44.28 0.97 0.31 0.7
Nelson Bay 3/02/2011 18:00 42.64 0.97 0.26 0.6
Nelson Bay 5/02/2011 17:30 38.76 0.97 0.35 0.9
Nelson Bay 2/02/2011 17:30 37.73 0.97 0.34 0.9
Sefton 1/02/2011 17:00 76.06 0.68 0.27 0.4
Sefton 2/02/2011 16:30 72.95 0.68 0.33 0.5
Sefton 3/02/2011 17:00 72.40 0.68 0.27 0.4
Sefton 31/01/2011 16:30 70.39 0.68 0.33 0.5
| 104
Charmhaven 3/02/2011 16:30 44.24 0.93 0.24 0.5
Charmhaven 1/02/2011 17:30 43.73 0.93 0.31 0.7
Charmhaven 5/02/2011 17:30 43.50 0.93 0.31 0.7
Charmhaven 2/02/2011 18:00 39.77 0.93 0.24 0.6
Ausgrid has also shared half-hour electricity data for 300 homes with rooftop solar systems that
are measured by a gross meter [25]. The meter measures and records the total amount of solar
power generated every 30 minutes. Ausgrid examined the 2014/15 season and estimated a
reduction in summer peak demand at the zone substation level of about 50 MW from the 240 MW
of generation capacity. Each of the 187 substation zones were split into three categories (Table
11). The zone peak between 11:00 am and 4:00 pm is categorised as ‘High', between 4:00 pm
and 6:00 pm as 'Medium' and after 6:00 pm as 'Low'.
Table 11 – Reduction in peak demand for each zone substation category, 2014—15 [25]
Solar Year Total system Rated capacity Estimated solar % solar No. of
effectiveness load (MVA) of connected contribution at effectiveness zones
solar (MVA) time of peak
(MVA)
High 2015 2483 59 32 54 66
Medium 2015 1943 110 18 17 63
Low 2015 633 72 0 1 58
Total — 5059 241 51 21 187
| 105
A histogram of this analysis shows that the contribution from solar at the zone substation level
averages about 1% and finds a maximum of 6% (Figure 19).
Figure 19 - Rooftop solar contribution to maximum peak reduction of analysis of 2014/15 [25]
Ausgrid notes that 2015 was generally mild, with no extreme temperatures. When there are
higher temperatures and higher demand, peaks at zone substations are typically pushed back in
the day while residential air-conditioning ramps up, with a resultant lower contribution from
solar. A coarse estimate on the impact of very hot days on PV and maximum peak coincidence
indicates that the solar effectiveness drops to 15—20%.
Analysis of the 2010/11 data from [25] is summarised in Figure 20 and Figure 21. The total load
for 300 homes was combined, and the daily maximum peak demand (kW) identified (the blue
curve in Figure 20). The total PV generation for all 300 homes was also combined; this value was
subtracted from the total load. The reduction in the daily maximum peak demand due to PV was
then identified (Figure 20, red line). Figure 21 gives the percentage decrease in the daily
maximum load due to PV.
The figures show that maximum peak reduction only occurs during the summer months.
Figure 20 indicates that on higher demand (hotter) days, the reduction in maximum peak
demand due to PV is also higher. On the two days where the highest percentage reduction
occurred (>20%), the maximum peak demand was 600 kW. Maximum peak demand occurred
during early February, ranging between 600 and 840 kW. During this period, maximum peak
reduction ranged between 3 and 20%.
| 106
Figure 20 - Maximum peak demand (left y-axis, blue) and reduction in peak demand (right y-axis, red), 2010—11 [25]
Figure 21 - Percentage reduction in maximum peak demand, 2010—11 [25]
3.4.4.3 Energex
Energex state that the impacts on domestic peak demand for upstream assets tend to be
negligible across their network [2]. They acknowledge some benefit from a small number of solar
PV installations on some commercial buildings attached to commercial and industrial day
peaking substations and 11-kV feeders. On a typical day, these installations may reduce the
overall network peak demand. Figure 22 shows the system peak occurring at around 2:00 pm for
the load, not including solar PV generation. On this day, Energex estimated a peak reduction of
more than 335 MW and peak-load shift to 3:00 pm. However, they acknowledge that factors
contributing to this reduction may include temperature, humidity and degree of cloud cover. In
different circumstances, PV may have no impact on peak demand, particularly at local levels.
| 107
Figure 22 - Energex system demand, solar photovoltaic (PV) impact, 22 January 2014 [2]
3.4.4.4 Endeavour Energy
Endeavour Energy identified a negligible effect of PV on peak demand, even at 50% penetration
of 1.5-kW systems [26]. Figure 23 was included in Endeavour Energy’s submission to the New
South Wales Independent Pricing and Regulatory Tribunal in 2010. It compares the theoretical
50% penetration PV output with the 2010 summer peak demand at a major zone substation in
Glenmore Park. The figure shows only a very slight overlap between PV generation and peak
demand.
Figure 23 - Theoretical solar output compared to substation demand, 2010 [26]
| 108
3.4.4.5 Panel orientation
Peak load may be reduced further by orienting solar panels to the west. This would reduce the
mismatch between PV generation and load profiles by increasing late afternoon PV generation,
thereby increasing the chance of overlap with peak network load. The extent to which panel
orientation may influence peak load was not found in published literature. However, it may be
worth exploring further. If the reduction were significant enough, it may be worthwhile offering
incentives to customers installing west-facing panels.
3.4.5 Power and voltage levelling
Power and voltage profile levelling involves minimising both high-frequency (seconds) and
low-frequency (hours) fluctuation in voltage and power. Levelling reduces peak load and
alleviates congestion, reduces line losses and increases the life of electrical equipment. This
enables the deferral of equipment capacity upgrades where capacity is constrained. It also
reduces the operation of expensive peaking plants, eases voltage management requirements and
improves electrical system reliability.
Levelling in microgrids significantly reduces wear and tear on the diesel generators supplying the
rest of the grid, and helps the thermal units maintain power balance and the system electrical
frequency [7].
An illustrative Australian LV distribution network was used to test the effectiveness of a
PV/storage system at reducing peak load [5] (see Figure 24). The circuit length was 350 m; pole-
to-pole distance, 30—40 m; conductor type, 7/3 AAC; transformer size, 160 kVA; PV size, 2—
4 kW; and battery size was 250 Ah.
Figure 24(a) shows how redirecting PV real power to charge the battery reduces voltage rise.
The algorithm is also designed to minimise voltage fluctuation due to PV; the effectiveness of
this control is also illustrated in Figure 24(a). Figure 24 (b) shows how reverse power flow is
reduced. In Figure 24(a), the rise in voltage around the peak-load period is an indication of
peak-load support from battery discharging.
| 109
a) b)
Figure 24 - Performance of battery energy storage system. Top row, no storage; bottom row, with storage [5]
(Reproduced courtesy of IEEE)
In reference [6], a BESS is integrated into each PV bus of the General Electric distribution power
system model [27] (see Figure 25). Each BESS is modelled as a lithium iron phosphate battery
and incorporates ageing effects to take account of battery lifetime. On-load tap changer
(OLTC)/SVR operation positions, peak-load reduction and peaking power generation are recorded
over the life of the BESS system. The BESS usage, lifetime and system performance, in terms of
battery size, are then analysed on every bus. Actual annual load information, PV power profile
and temperature data is used for the analysis.
Figure 25 - General Electric feeder model [6]
The study in [6] focuses on feeder 2. The peak load is 11 MVA; the seven loads range between
0.3 and 0.5 MW, with a 0.92 power factor; primary voltage is 12.5 kV, and secondary is 240 V.
Although load power factors in Australia are typically higher than 0.92, the findings of the paper
are still relevant to Australian electrical networks. Using rules-based control, a central unit is
used to manage the actions of the BESS for voltage regulation and peak-load reduction.
| 110
Figure 26 presents the benefits provided by BESS operation over a 24 hour period. The results
illustrate a smoother voltage profile and reduction in both peak load and SVR/OLTC operation.
Figure 27(a) shows the annual reduction in SVR/OLTC operations, while Figure 27(b) gives the
annual reduction in peaking power generation for a given PV penetration and BESS size in the
study. PV penetration is the ratio of PV generation to peak load and BESS size is measured in
peak-load hours. The reduction in OLTC and SVR operation, as well as peaking power generation,
is clear.
a) b)
Figure 26 - Power and voltage profile, step voltage regulator (SVR) and on-load tap changer (OLTC) operation. a)
Without a battery energy storage system (BESS); b) with BESS [6]
a) b)
Figure 27 - (a) Reduction in step voltage regulator (SVR) and on-load tap changer (OLTC) operations; (b) reduction in
peaking power generation [6] BESS = battery energy storage system
For networks with sufficient levels of wind penetration, power generation from wind at night —
when load is much lower than during the day — may reverse power flow at the PCC and cause
voltage rise. STATCOMs or SVCs would normally be used to manage this voltage rise, but large
| 111
voltage-source inverters, normally used to interface PV solar farms to the grid, could be used
instead [28]. The capacitor of the voltage-source inverter is operated in ‘self-supporting’ mode,
and therefore doesn’t require an external DC power source. Instead, the reverse power from the
wind turbines provides the required power. Simulation testing shows that the voltage-source
inverter helps maintain voltage levels within regulation limits.
A PV-coupled BESS using the Xtreme Power-Dynamic Power Resource (XP-DPR) has been
developed for renewable energy applications [7]. The MW-scale system, manufactured by
Xtreme Power in Texas, is currently operating in a solar-coupled mode on 12.47 kV power
systems in the Hawaiian Islands, as well as at a solar technology testing facility in Colorado
under the auspices of Xcel Energy and the National Renewable Energy Laboratory. Most BESS
control systems can be operated via AGC signals, much like a conventional utility generation
asset, or in a solar-coupled mode. In the latter, real and reactive power commands for the
converter are generated many times per second based on real-time PV output and power system
data. Control algorithms implemented in the XP-DPR provide control of ramp rates, frequency
support and voltage/reactive power support, as well as services designed to optimise the
financial returns of the PV installation, including peak-shifting and levelling.
Figure 28 depicts the operation of an XP-DPR BESS smoothing the volatile power output of a
1-MW solar farm. Note that the system ramp rate is maintained at less than 50 kW/min, whereas
the solar resource alone had a maximum second-to-second ramp rate of more than 4 MW/min.
Figure 28 - Ramp-rate control to 50 kW/min for a 1-MW PV installation and a 1.5-MW/1-MWh battery energy storage
system. (a) Full day; (b) details of largest event [7] (Reproduced courtesy of IEEE)
BESS units can also apply a time-shift algorithm optimised for a given set of solar generation and
load forecasts. The algorithm can charge the BESS from the grid at night, or from some
percentage of the solar generation during the day. The BESS can then be discharged in the
afternoon hours, coincident with the evening peak. Figure 29 shows the morning PV generation
being used to charge the BESS for one day and for one hour; the BESS is then discharged to a
consistent 750 kW output from 2:00—4:00 pm.
| 112
a) b)
Figure 29 - (a) Full day operation of time-shift algorithm; (b) 1-hour of time-shift algorithm [7] (Reproduced courtesy
of IEEE)
3.4.5.1 Storage – Australian trials currently in progress
Three DNSPs are currently participating in battery storage trials: Ausgrid, Ausnet and Ergon
Energy.
i) Ausgrid has trialled a 60-kW battery storage system in Newington, Sydney, to test how battery
storage can help manage summer peak demand events. As part of the Smart Grid, Smart City
program, Ausgrid also completed a trial of 60 5-kW battery storage systems installed in homes in
suburban Newcastle. The aim was to understand the technical impacts of adding battery storage
to the grid, and the potential of batteries to power local areas during maintenance or outages.
ii) AusNet is currently trialling a 1-MW battery to support the electricity grid during peak
demand periods. Named the Grid Energy Storage System, the battery is located at Thomastown
industrial estate. The system also includes a 1-MW diesel generator. It provides active and
reactive power support, and can also enter islanding mode when isolated from the grid.
iii) Ergon energy is currently trialling their Grid Utility Support System. Figure 30 shows that the
system reduces LV excursion over 48 hours, covering two morning and two afternoon peak
periods.
Figure 30 - Voltage profile or Ergon Energy trial, with and without Grid Utility Support System (GUSS). CES =
centralised energy storage
| 113
3.4.5.2 Virtual power plant
This section discusses work that considers the power of several small-scale, distributed EG
devices in aggregate. The power generated from all devices, when considered in aggregate, may
be controlled to replicate large-scale dispatchable generation: a ‘virtual power plant’. The
virtual power plant systems discussed in this section are PV combined with storage, and
standalone storage.
In a virtual power plant, groups of EG have system visibility, controllability and impact similar to
a transmission-connected generator, with similar parameters (e.g. scheduled output, ramp
rates, voltage regulation capability, reserve) [29]. The aggregate power generated from the EG
is controlled and dispatchable.
In 2005, American Electric Power (AEP) pushed energy storage to the distribution level in the
form of sodium sulfur batteries located at distribution substations. The storage sizes are 1, 2 and
4 MW, with 7-hour discharge capability. AEP is developing and installing smaller, more broadly
distributed energy storage units at the edge of its grid, at customer sites. Termed community
energy storage (CES), the units are 25-kW operated as fleets to provide aggregate benefits at the
MW scale using existing communication and control technologies [14].
AEP found that power system reliability was improved when storage is pushed further out on the
system. Each battery installation is equipped with ‘dynamic islanding’, a feature that allowed
AEP to serve hundreds of customers during a power outage.
AEP list numerous advantages for the CES system, including the capacity to:
 provide buffering of customer-owned renewable generation
 improve power quality
 enable load levelling for capital deferral and optimised grid operation
 enable frequency regulation and other ancillary services.
To install and operate a fleet of hundreds of CES units, standardisation and battery cost
reduction are critical. Through an effort sponsored by the Electric Power Research Institute
(EPRI), with collaboration of more than 20 utilities, vendors and manufacturers, AEP has
developed detailed functional specifications for CES. The specifications are open source and
publically available [30].
AEP terms CES as a ‘virtual substation battery’, but claims it has distinct advantages over an
actual substation battery. These include:
• more reliable backup power to customers
• more effective voltage and VAR support
• more scalable, flexible implementation
• likelier to be a standardised commodity
• more efficient buffering of customer renewable sources
• more synergy with EV batteries
• easier installation and maintenance
• less critical effect of unit outage on the grid
• lower resistive loss in wires
• better fit into smart grids.
The potential for PV combined with storage to improve the voltage profile of residential LV
circuits is examined in [8]. The aim is to mitigate both voltage rise due to PV and voltage drop
during peak demand periods. All EG devices are in communication, and control is coordinated. A
potential deficiency of using reactive power absorption (for voltage rise) and reactive power
| 114
injection (for voltage drop) for rural Australian feeders due to the cables’ low X/R ratio (ratio of
reactance to resistance) is also pointed out in [8]. For rural areas, the authors propose using
storage integrated with PV as well as reactive power controls to manage voltage levels. Battery
charging and discharging is managed through voltage-dependent droop control, while constant
and variable droop control is considered. The proposed control method was tested on urban and
rural distribution feeders located in Brisbane, Queensland (see Figure 31).
a) b)
Figure 31 - (a) Urban feeder; (b) rural feeder [8]
For the urban case shown in Figure 31, the houses are 10—35 m apart and cable X/R ratios vary
between 0.8 and 1.1. For the rural feeder, houses are 80—200 m apart and cable X/R ratios are
around 0.17. Load data is taken from the distribution network operator. The average peak load
is 3 kW and PV size is assumed to be twice this to simulate severe over-voltage conditions. The
voltage for the house at the end of the feeder is examined. The PV inverters are capable of
providing 3.6 kVA during the day and 6 kVA during the evenings, running on battery power.
Figure 32(a) shows the change in voltage profile for urban and rural cases when only reactive
power support is provided; breaches of the lower voltage limit still occur in the rural case.
Figure 32(b) shows the rural feeder voltage when real power from the battery is also injected
into the feeder. In this case, voltage levels remain above the lower voltage limit.
a) b)
Figure 32 - (a) Voltage profile for urban and rural feeders, with and without Q compensation; (b) rural feeder voltage
when real power is also supplied from the battery [8]
The study found that variable droop control is as effective as constant droop control, while at
the same time reducing the required storage capacity for the house at the end of the feeder [8].
With variable droop control, houses towards the end of the feeder have a lower power response
to a voltage change than those closer to the head of a feeder. Storage at these locations can
therefore be reduced.
| 115
Key Finding: Utilising EG for power and voltage profile levelling reduces the operation
of expensive peaking plants. It can also ease voltage management requirements,
improve power system reliability and provide frequency support.
Key Finding: Insecure, non-dispatchable energy source types of EG (predominantly PV

and wind) alone provide only limited power and voltage levelling and frequency
support. Including storage offsets the insecure, intermittent nature of PV and wind,
allowing EG to provide these services effectively.
3.4.6 Frequency support
This section examines the contribution of EG to frequency control. It refers to PV technology

combined with energy storage, in which the PV + storage system is controlled to replicate the
frequency support response of a synchronous generator. The same support could also be
provided by wind combined with storage and actual synchronous generators such as diesel
engines or gas turbines.
Reference [31] describes the design of a ‘grid-friendly’ PV plant. The plant includes a controller
that manages real and reactive power flow and allows it to contribute to the reliability and
stability of the electrical transmission and distribution system. The PV plant’s features include:
 dynamic voltage and/or power factor regulation of the solar plant at the point of
interconnection
 real power output curtailment to ensure upper voltage is not breached
 ramp-rate controls to ensure that the plant output does not ramp up or down faster than
a specified ramp-rate limit
 frequency control to lower plant output in the case of an over-frequency event, or
increase plant output (if possible) in the case of under frequency
 start-up and shut down control.
Generator emulation controls are discussed in [32]. In these systems, the inverter of the grid-
connected PV system is controlled to mimic the behaviour and dynamics of a synchronous
machine. The generator emulation control allows the inverter to supply reactive power and
current harmonics to local loads, provide voltage support through volt/VAR control, and perform
frequency regulation through hertz/watts control.
| 116
The XP-DPR BESS [7] provides frequency support using proportional control to deliver or absorb
power in support of grid frequency stabilisation. This control is equivalent to conventional
generator governors equipped with a speed-droop or regulation characteristic. The droop
response of the XP-DPR is a function of the mechanical inertia of the conventional generation
(rotating machines) in the network. The smaller the inertia (installed capacity of conventional
generation), the more the frequency is affected by a step change in load or generation, and the
greater the required change in real power generated by the BESS for a given change in
frequency. R is defined by equation 1, where 1/R is the gain of the proportional controller:
𝑝𝑒𝑟𝑐𝑒𝑛𝑡 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 𝑐ℎ𝑎𝑛𝑔𝑒
%𝑅 = 𝑝𝑒𝑟𝑐𝑒𝑛𝑡 𝑝𝑜𝑤𝑒𝑟 𝑜𝑢𝑡𝑝𝑢𝑡 𝑐ℎ𝑎𝑛𝑔𝑒 × 100 (1)
The concept of large-scale, centralised PV (LSCPV) plant providing frequency support

functionality, frequency-sensitive mode and inertial response was tested in [9]. A LSCPV plant
was integrated into a standard IEEE 12-bus system with penetration levels of 5% and 20%.
Frequency-sensitive mode and inertial response were primarily provided though the control of
active and reactive power. Sufficient active power reserve was required to ensure the LSCPV
plant can operate in frequency-sensitive mode and provide inertial response. Active power
reserve was realised by setting the maximum power point set-point below its maximum, and the
reserve was released during transients.
The performance of the LSCPV was tested by initiating a frequency event (major loss of
generation) on the IEEE system. For the two test scenarios, with the LSCPV plant providing
frequency-sensitive mode and inertial response support, the system recovered more quickly. A
daily load profile was also applied to the system and the frequency deviation across the day
recorded. The load profile was divided among all the loads present, and the generators acted
together to compensate for any power mismatches, while AGC was managed by the largest
generator in the system. The frequency distribution was shown to have less deviation when
LSCPV plant frequency support was present [9].
3.4.7 Dispatchability
Many papers that discuss the potential of EG to provide power, voltage levelling and frequency
support also discuss dispatchability. Dispatchable EG increases the pool of generation that can
be called on to meet load demand, increasing generator portfolio flexibility.
Individually, EG is relatively small in scale, but may be large when considered in aggregate. It is
also likely to have a fast response time, enabling rapid response to both local and network load
changes. This section refers to PV combined with storage, and wind combined with storage, as
dispatchable systems.
Reference [11] proposes combining a supercapacitor bank (SCB) and a VRB to smooth out the
power fluctuation of a 1-MW grid-connected PV plant. The power management of the HESS was
designed to reduce the power requirement of the SCB to one-fifth of the VRB rating, and also to
| 117
avoid operating the VRB at reduced power levels. SCBs experience cell voltage unbalancing
issues above the MW level; minimising the required size of the SCB (relative to the VRB) avoids
this issue.
The power management system controls the HESS such that it meets the Australian 5-minute
semi-scheduled dispatch requirements [33], and maintains operation of the VRB and SCB with
rating constraints. A test case was simulated using a 250-kWh VRB and 50-kW SCB (the capacity
of the VRB was not specified). High-speed (~1 second) insolation measurements taken from the
St. Lucia campus of the University of Queensland were used for PV production modelling. A
30-min window of solar insolation and a power cap was carefully selected to ensure that all
operating modes (normal, deficit and excess power mode) of the HESS were triggered. Figure 33
shows that the operating constraints were breached on only three occasions (200, 400 and
900 seconds) [11].
Figure 33 - Hybrid energy storage system (HESS) operation [11]
The HESS was also tested using a week’s worth of worst-case insolation data, believed to have
been during a period of high variability in insolation [11]. The average PV power generated for
each 5-min period was selected as the power cap. The results show that the HESS does operate
outside of constraints on occasions, but that the power management system quickly returns
power to within constraints. The test case shows that the proposed HESS and power
management system can control the power generated from a 1-MW PV such that its operation
meets Australian grid code requirements. The SCB was also kept at one-fifth the size the VRB,
and VRB operation at low levels is avoided.
The work presented on rule-based control in reference [12] is similar to [11], with a focus on
BESS in place of HESS. A control strategy was designed to minimise the power fluctuations of
variable renewable EG (wind or solar) and make its power dispatchable. The effectiveness of this
control strategy and BESS was tested using an actual PV system and wind farm data. As in [11],
the results show that the normally variable power generation is smoothed considerably, making
dispatch possible [12].
The BESS was connected at the system’s PCC, and charged/discharged through a power
converter to smooth the net power injected to the system. The performance of the BESS and
control system was tested at hourly intervals (consistent with United States dispatch windows).
Maintaining constant power for 1 hour is more difficult than maintaining power for 5 min, as is
required in Australia [11].
Prediction is not mentioned in [11], and is unlikely to be useful. For solar, at 5-min intervals,
cloud transients are the source of unexpected variability in insolation, and prediction of either a
| 118
break in the clouds on an overcast day or random passing cloud on a sunny day is not possible
using standard regression methods. Reference [12] assumes that prediction would be used to
determine the target power generation for the upcoming hour. When testing the system, the
target power selected was the average of the upcoming hour wind or solar generation. A random
error with zero mean and 0.1 standard deviation was introduced to represent prediction error.
A simple rule-based system was used to control the BESS [12]. The battery’s state of charge must
remain within 30 and 100%, and current flow in and out of the battery must not exceed
maximum charge or discharge rates for the battery. The BESS and control system will fail to
meet the target power when any of these operational constraints are reached.
The system was tested with 1.5 MW solar and 50 MW of wind. For the solar case, the battery is
sized at 300 kWh, while for wind, the battery is sized at 10 MWh. Looking at the HESS system in
[11], the relative battery size for the solar case in [12] is smaller: 20% of PV capacity compared
with 25%. It might be expected that the battery size should be larger, because power is to be
maintained over 1 hour, instead of 5 min. The authors of [11] may have oversized the VRB to
ensure performance. An iterative process is used to select the size of the battery in [12]; the
process is not explained, but may optimise battery size selection.
For the 1.5-MW solar case, the effectiveness of the BESS at meeting the target power is best
illustrated by histograms of the difference in actual power and target power before and after
the BESS. Figure 34 shows that the BESS greatly reduces deviation from the target power;
undesirable deviation is considered anything greater than 100 kW, positive or negative. The
percentage of these occurrences reduces from 16 to 4% with the BESS [12].
a) b)
Figure 34 - Histogram of difference in actual power and target power for a 1.5 MW solar photovoltaic system. (a)
Without a battery energy storage system (BESS); (b) with BESS [12]
3.4.8 Electrical infrastructure – life extension and investment deferral
| 119
This section considers how the installation of EG affects the life of electrical infrastructure and
contributes to investment deferral.
It is apparent that any asset life extension is economically beneficial to DNSPs, although there is
no consensus about the extent of these benefits in monetary terms [15].
The growth of EG and a growing realisation that investment here may defer network
augmentation means that ‘non-wire solutions’ are now considered a realistic alternative to
traditional network upgrades. The benefit of EG in this context is straightforward: it will take
more time for the current on-load transformers or feeders to reach their technical limits if EG
can be effectively deployed and controlled to manage current flows. Therefore, the first step
towards quantifying the benefit of transformer or feeder investment deferral is to measure the
impact of the EG output on the currents across the distribution network [34].
Transformer life is dependent upon insulation life. The main factors determining insulation life
of an oil-immersed transformer are:
 load
 ambient temperature
 the number and magnitude of faults it has endured
 moisture and oxygen content of the oil.
The rate of ageing of transformer insulation is primarily temperature dependent (references [3,
35] detail a transformer heating model). The thermal response is a function of the transformer
constants and load. The loss-of-life (LOL) calculation is based on the transformer insulation
condition, which is a function of insulation temperature.
PV reduces daytime load, and as a result, limits the temperature rise of the transformer.
However, the transformer temperature reached at peak demand is the key concern. For the
midday peak, PV can directly reduce maximum transformer temperatures. For the typical late
afternoon/early evening peaks seen in Australia, the transformer temperature at the start of the
peak demand period will often be reduced by afternoon PV output, as discussed in Section 3.4.4.
As a result, the transformer temperature reached at peak demand is also reduced. This leads to
the conclusion that PV can extend transformer life (at least from the perspective of temperature
stress).
Energex claims that PV has contributed to a rating increase of 10% for their substation
transformers [2]. As of March 2014, 49 substations had been uprated, with a total capacity
increase of 309.5 MVA. Following a review of the 2013—14 summer load profiles on all mixed and
industrial substations, a further 19 substations were uprated by an additional 148.5 MVA.
Energex expects further upratings beyond July 2014, dependent on the rate of future solar PV
system connections.
In reference [4], for a commercial load, the number of hours a substation transformer
experiences overload conditions is calculated each year using an assumed load growth rate of
4%. The number of overload hours is again calculated with PV penetration levels of 10% and 20%.
For a projected 2016 load, the number of hours that a substation transformer is overloaded
decreases from 510 to 179 for 10% penetration and to 35 for 20% penetration. This defers the
transformer upgrade by two and four years, respectively.
| 120
Reference [13] analyses transformers in a 102-node, 400/230-V distribution network connected
to the South West Interconnected System (SWIS). These transformers are non-thermally
upgraded paper, with a life of 30 years. The distribution transformer (DTx) used for the analysis
is unbalanced [13]. Out of 77 connected residential consumers, 13 are connected to phase-A, 17
connected to phase-B and 21 to phase-C; 26 of the premises have a three-phase connection. A
total of 64 kW of PV is connected to the DTx. In Australia, it is standard practice to load a
transformer up to 1.4 times its rating for a short period of time in a given year [36]. The LOL of
a DTx is therefore investigated for loading up to 1.4 times its rated power (280 kVA, in this
case). Figure 35 gives the temperature difference in hot-spot temperature, loading and
reduction in LOL as a result of PV generation. The saving in LOL is 0.2, 14, and 160 days for
phases A, B and C respectively. Table 12 shows the increase in LOL reduction with increasing PV
penetration.
a) b)
c)
Figure 35 - At a loading of 1.4, (a) distribution transformer (DTx) hot-spot temperature (HST) with and without
photovoltaics (PV); (b) DTx loading with and without PV; (c) DTx loss-of-life (LOL) with and without PV [13]
Table 12 - Loss-of-life improvement with photovoltaic (PV) growth [13]
Transformer loading/phase C PV Total (max) (C) Hot-spot Loss-of-life (days)

installation temperature (max)
(C)
1.1 PU-26 kW PV 79.5 137.9 22
1.1 PU-35 kW PV 77.1 130.6 14
1.2 PU-26 kW PV 85 151.7 85
1.2 PU-40 kW PV 82.1 141.9 40
| 121
Reference [15] explores a method for quantifying the economic benefits of DTx life extension
through EG, applied to actual DTx installed in five sample cities in Iran to provide realistic
estimates. A DTx exhibits time-varying profiles of load and ambient temperature over the days
of its operation history. To remove the time dependence in the profiles of these two variables,
an approach was adopted from a previous work by the authors [17] to classify the operation
history of any analysed DTx into eight typical days corresponding to seasons, weekdays and
weekends. Each typical day was thought of as a sequence of 15-minute intervals during which
transformer load and ambient temperature were kept near-constant.
The study calculated the DTx LOL for each of the eight typical days. DTx load profiles were
generated through stochastic models of customer load and EG production, taking into account
atmospheric conditions (e.g. temperature, insolation, wind).Figure 36 is a schematic diagram of
the modelling process. The EG units considered include PV, wind, CHP units and microturbine
(MT) units. Load data consists of characteristics adopted from [37] (historical Iranian load data
with a noise input of 15%).
Figure 36 - Distribution transformer loss-of-life (LOL) modelling process [15]. DG = distributed generation (Reproduced
courtesy of IEEE)
Ageing formulas for oil-immersed transformers in [3] were used in the analysis. Table 13 gives
the parameters of cities examined for the analysis.
Table 13 - Parameters of cities examined for the analysis of distribution transformer loss-of-life [15]. (Reproduced
courtesy of IEEE)
CHP = combined heat and power; MT = microturbine; PV = photovoltaic; WT = wind turbine
The analysed transformers were rated at 315 kVA (20/0.4 kV) and were equipped with an
oil-natural air-natural cooling system. Their thermal parameters, as shown in Table 14, were
derived from data provided by manufacturer.
| 122
Table 14 - Thermal parameters of transformers analysed in reference [15]. (Reproduced courtesy of IEEE)
The following assumptions were used in the analysis:

• All EG units operate at unity power factor.
• A maximum aggregated capacity limit of 50% is assumed to be applied by the DNSP to the
installed EG units. This means that total installed capacity of EG units under each
proposed scenario should be kept lower than 50% of the DTx-rated capacity.
• All customers in the sample cities are subjected to the same patterns of wind speed and
solar irradiance.
• The analysed transformers are assumed to have a design life of 20.55 years when they are
well-dried, oxygen free and operating continuously at a hot-spot temperature of 110 C.
Table 15 gives the life extension and estimated economic value (US$/year) for each city for each
EG type.
Table 15 - (a) Transformer life extension; (b) estimated economic value (US$) of distributed generation technologies
[15]. (Reproduced courtesy of IEEE)
a) b)
CCN = city code number; CHP = combined heat and power; MT = microturbine; PV = photovoltaic; WT = wind turbine
Reference [17] calculates the reduction in LOL in a DTx due to EG for a combination of
technologies. The method was applied to an actual distribution system in Iran (an urban section
in the city of Sirjan) with 10 km of 20-kV feeder and 44 DTx’s supplying 52 km of LV (400/230 V)
feeders. Three typical DTx’s were selected for analysis. For an EG penetration level of 10%, the
LOL rate is presented for a typical winter and summer weekday. Penetration level is the ratio of
EG to DTx rating. Eight scenarios were tested: no EG, PV, wind, CHP, MT, half PV/wind, PV/CHP
and wind/MT. The results in Figure 37 show that the LOL rate decreases for all scenarios. In
summer, PV provides the most reduction, followed by PV/wind. In winter, CHP provides the most
reduction, followed by PV/wind, MT, PV/CHP and wind/MT, which all provide similar reductions.
| 123
b)
a)
c) d)
Figure 37 - Reduction in loss-of-life (LOL) acceleration rate in (a, b) summer and (c,d) winter for different
combinations of photovoltaic (PV), wind turbine (WT) combined heat and power (CHP) and microturbine (MT) sources
of embedded generation (EG) compared to reference scenario without EG [17]
Figure 38 provides an insight into the reduction in LOL rate over the life of the DTx. The annual
LOL rate for each technology and combination of technologies is plotted against penetration
level. Relative to the reference LOL rate, the best result is achieved through the MT, at 15%
penetration. The LOL rate is reduced from 1.2 to 0.15 (87.5% reduction).
| 124
a) b)
Figure 38 - Annual loss-of-life (LOL) reduction for different combinations of photovoltaic (PV), wind turbine (WT)
combined heat and power (CHP) and microturbine (MT) sources of embedded generation (EG). (a) CHP, PV and WT
compared to reference scenario with no EG; (b) PV/CHP, PV/WT, WT/MT and MT [17]
The impact of EG on distribution networks’ investment deferral in the long-term is assessed in

[16]. Due to the randomness of the variables that affect this (e.g. load demand patterns, EG
hourly energy production, EG availability), a probabilistic approach using a Monte Carlo
simulation was adopted. Several scenarios characterised by different EG penetration and
concentration levels and EG technology mixes were analysed. The study aimed to estimate the
admissible load growth in distribution networks with and without EG. Inhibitors to load growth,
which are incorporated into the modelling process, include: power flow capacity of transformers
and cabling, maximum voltage drop and ohmic losses. EG penetration is defined as the ratio of
EG capacity and peak-load demand.
The model was tested using a semi-rural MV network (Figure 39). The maximum transfer
capacity of the network (the maximum power that can be supplied through the distribution
network in normal operating conditions) is approximately 7 MW.
Figure 39 - Semi-rural network used for the probabilistic model tested in reference [16]
| 125
The technologies used in the model were CHP, wind and PV. CHP generation data was based on
actual historical data. For wind, at each hour a random number was generated with a Rayleigh’s
probability density function, with mean equal to the average speed of the wind farm’s location.
The power produced was then estimated using manufacturers’ power vs speed wind curves. PV
output profiles were generated using theoretical irradiance values, and production then
calculated using data from the manufacturers. Variability due to cloud was then incorporated.
The availability of the EG units is incorporated into the model. Table 16 specifies the
failures/year and the mean time to repair.
Table 16 - Embedded generation units failure rates [16]
Embedded generation type Failures/year Mean time to repair (days)
Wind turbines 5 3
Photovoltaic 0.2 1
Combined heat and power 1 3
Each load node in [16] is considered to have a specific combination of consumer types. Each type
of consumer is characterised by a load pattern using real historical hourly data. To include short-
term demand uncertainty, a ±10% uniformly distributed band noise was added to the hourly
demand of each node. Each test scenario was characterised by different EG penetration levels,
EG concentration levels and EG mix parameters. The percentage chance of overload occurring
with and without EG was then calculated using a Monte Carlo method (Figure 40). The overload
probability is the likelihood that the power flow exceeds the maximum transfer capacity.
| 126
Figure 40 - Monte Carlo algorithm flow chart used in reference [16]. DG = distributed generation
The calculated increase in admissible load growth is given in Figure 41. Figure 41(a) is for CHP at
node 12, (b) is for wind at node 12, and (c) is for PV at node 12. CHP provides the biggest
increase in admissible load growth: more than 275% at 100% penetration. Wind is the worst, at
just more than 225% at 100% penetration. Figure 41d) shows the increase in admissible load
growth when either wind, PV, CHP or CHP/wind is installed at nodes 11, 16 and 21. Again, wind
is the worst performer, with the other technologies performing equally well to provide an
increase in admissible load growth of nearly 250% at 100% penetration.
The financial benefits of EG are not discussed in reference [16]. However, an investment
deferral time could be calculated as the increase in admissible load growth over an assumed
load growth percentage. The savings from investment deferral could then be calculated as in
[34].
| 127
a) b)
c) d)
Figure 41 - Increase in admissible load growth for (a) combined heat and power (CHP); (b) wind turbine (WT); (c)
photovoltaic (PV); and (d) load growth v penetration for WT, CHP, PV and CHP/WT [16]
Key Finding: EG can contribute to peak load reduction. This alleviates congestion,
reduces line losses and results in deferral of equipment capacity upgrades where
capacity is constrained.
Key Finding: Evidence from the literature and distribution network service providers
suggests that the reduction in cumulative net load due to EG can extend the life of
both substation and distribution transformers.
| 128
3.4.9 Cost—benefit analysis
This section discusses studies that present cost—benefit analyses of EG.
3.4.9.1 Savings though reduction in operation costs
The following papers consider the cost benefits of installing EG to reduce network operating
costs.
Reference [6] conducts a cost—benefit analysis of BESS for a PV penetration level of 70%. Figure
42(a) shows that no BESS installation is profitable for the scenarios analysed, and that the
savings from the reduction in OLTC and SVR operation, peak-load shifting and peaking power
generation do not offset the investment cost. Figure 42(a) presents the best identified cost–
benefit scenario; 70% is the highest penetration level simulated, and lower penetrations of PV
give poorer returns. Figure 42(b) lists the inputs used in the analysis. The savings due to reduced
OLTC and SVR operation are calculated as the reduction in operation and maintenance costs.
Savings from peak-load shifting are calculated as the difference in the cost of energy supplied at
peak-load rates compared to off-peak load rates, while savings due to a reduction in peaking
power generation are self-explanatory.
b)
a)
Figure 42 - (a) Results of cost–benefit analysis of battery energy storage systems (BESS); (b) Inputs into cost—benefit
analysis [6]. EPRI = Electric Power Research Institute; LiFePO4 = lithium iron phosphate; NREL = National Renewable
Energy Laboratory; PCS = power conversion system; PJM = Pennsylvania-Jersey-Maryland; SANDIA = Sandia National
Laboratories
| 129
Although this work found that no BESS installation is profitable under the tested scenarios,
varying the inputs to the cost—benefit analysis (Figure 42 (b)) could result in a positive outcome.
The power conversion system cost is set to US$400/kW, taken from a 2003 report [38]. A price of
US$300/kWh is used for the battery cost; reference [39] suggests that $250/kWh is the required
price for battery storage to be cost competitive. Peak load and off-peak load electricity rates
can also differ; peak rates >22 c/kWh and off-peak rates <7.5 c/kWh are certainly possible
(values in US$).
The economic benefits of a distributed STATCOM (DSTATCOM) in combination with a large solar
farm are considered in [40]. The following rules govern the operation of the solar farm and
DSTATCOM:
 If voltage levels are within regulation limits, then the DSTATCOM does not operate, and
generation from the solar farm is uncurtailed.
 If the upper voltage limit is breached, then the DSTATCOM is operated to reduce voltage.
 If voltage cannot be sufficiently reduced by the DSTATCOM, then generation from the
solar farm is curtailed until the voltage is within regulation limits.
Revenue from the installation was calculated as the sum of solar farm generation revenue and
savings from the reduction in conventional generation. The savings from reduced conventional
generation are the product of the reduction in energy demand (due to the solar farm) and
conventional generation operational cost ($/kWh). The installation was located at the end of a
Taipower 11.4-kV distribution feeder, which is 12.1 km long. A typical hourly daily load curve for
the feeder was taken for each season. After the rules of operation were applied, the solar farm
output was calculated to derive the reduction in conventional generation. Insolation
measurements from the Taiwan Weather Bureau were used to estimate hourly solar farm
generation for the year for a range of solar farm capacities. For DSTATCOM sizes of 0, 50, 200,
350 and 500 kVA, the optimum sized solar farm was determined to maximise the net present
value (NPV).
The study found that a 4.5-MWp solar farm in combination with a 500-kVA DSTATCOM is the most
profitable [40]. This set-up gives an NPV of $9.47 million, with a payback period of 9.87 years;
the life of the equipment is 25 years. The study does not take into account potential savings
from the voltage management, such as increased life of other assets (e.g. transformers). It also
assumes no clean energy credits for the clean power generation from the solar farm, and does
not monetise the auxiliary services that such a set-up could provide.
3.4.9.2 Savings through investment deferral
The mechanism by which EG may be considered as an alternative to network investment is

discussed in Section 5.3.3.1. Here, we investigate cost—benefit analyses to help identify the
viability of EG in these non-network solutions. The following papers consider the cost benefits of
installing EG to defer network investment.
| 130
Reference [18] deployed CHP EG at different locations and penetration levels to assess the
deferral in investment for a practical United Kingdom extra-high-voltage (EHV) network (33—
132 kV. The calculation used to evaluate investment deferral is given in equation (2):
𝐴𝑠𝑠𝑒𝑡 𝐴𝑠𝑠𝑒𝑡
∆𝑃𝑉 = ∑𝑀 𝑀 𝑖 𝑖
𝑖=1(𝑃𝑉𝑖 − 𝑃𝑉𝑖𝑛𝑒𝑤 ) = ∑𝑖=1 ((1+𝑑)𝑛𝑖 − (1+𝑑)𝑛𝑖𝑛𝑒𝑤 ) (2)
where:
PVi NPV of future investment without EG installation (current condition)
PVinew new NPV of the future investment with EG installation
△PV change in NPV, which could be regarded as either investment deferral or
acceleration, dependent on the direction of the change
M total number of assets in the network
d discount rate
Asseti modern equivalent assets cost
ni time to reinforce a network asset if no EG is installed
ninew new time to reinforce a network asset if EG is installed.
Investment deferral savings are therefore associated with the amount of time before an asset
needs to be upgraded. The time required before an asset requires upgrading is a function of load
growth. Where EG decreases load growth, more time is required before an upgrade is needed.
Load growth rate, r, is assumed to be 1.6%, the projected load growth rate in the United
Kingdom. The discount rate is set to 6.9%. ni can be calculated through equations (3) and (4):
𝐶𝑖 = 𝐷𝑖 × (1 + 𝑟)𝑛𝑖 (3)
𝑙𝑜𝑔𝐶𝑖 −𝑙𝑜𝑔𝐷𝑖
𝑛𝑖 = (4)
𝑙𝑜𝑔(1+𝑟)
where:
Ci asset load capacity
Di current load
r load growth rate.
In equation (4), a network asset (asseti), such as a circuit, has a capacity of Ci, and supports a
power flow of Di. The time to reinforce the network asset is the number of years that it takes for
the circuit to grow from Di to Ci for a given load growth rate r. Reference [18] assumes that
reinforcement occurs when the circuit is fully loaded. The test system is a 33-kV network with a
peak demand of 81 MW. EG units are 1.2-kW CHP gensets.
Table 17 shows the reduction in investment as a percentage of the required investment if no EG
units were installed. Savings are calculated for EG evenly distributed and allocated in proportion
| 131
to load for four penetration levels, while penetration percentage is the installed EG capacity as
a percentage of the total load [18].
Table 17 - Investment savings presented as a percentage of the required investment if no embedded generation (EG)
unit were installed [18]
Penetration level 1.5% 11% 28% 32%
Reduction in investment (%)
EGs evenly distributed 4.75 32 65 70
EGs allocated in proportion to load 7 42 77 82
Table 17 reveals that large savings can be made in investment deferral when EG is installed
throughout the network, especially when EG units are allocated according to load demand: up to
an 82% reduction for a 32% penetration level. Despite the investment savings calculation in
equation (2) not taking the cost of the EG unit into account, the formulation is at least sufficient
to underline the potential benefits of EG.
In reference [34], a test distribution network was broken up into separate feeder groups to test
the reduction in current due to EG (Figure 43). At each bus, EG was installed with a gradually
increasing capacity from 10 to 100 kVA. For each incremental increase, load flow and feeder
currents were calculated. A load growth of 3% was assumed for all buses. The time taken for
currents to reach levels where a cable upgrade is required was then calculated from when the
EG is installed. The savings were calculated by NPV, equivalent to the interest on the investment
cost of the cable upgrade for the deferral time. The NPV calculation does not take into account
the investment cost of the EG, assuming instead that it is owned by a third party, not the DNSP.
| 132
Figure 43 - Test distribution network used in reference [34]
The cost to upgrade each feeder group is given in Figure 44(a). Figure 44(b) gives the financial
benefit for 100-kVA EG installed at their respective buses, and the equivalent current reduction
due to an EG installation of 100 kVA at bus 830. At bus 858, its deferral benefit of nearly
$1200/kVA equates to more than 15 years’ upgrade deferral.
a) b)
Figure 44 - Feeder upgrade costs used in reference [34]
| 133
Reference [20] calculates the return on investment for DNSPs installing EG technologies (gas and
diesel generators). The return on investment was calculated for three different operating
systems that run only at:
i. maximum, normal, and medium load (5820 hours per year)
ii. maximum and normal load (weekdays from 8:00 am to 8:00 pm, 2880 hours per year)
iii. maximum load (80 hours per year).
The load bands (maximum, normal, medium) are given in Table 18(a). The calculated PV for the
investment is given by equation (5):
𝑅𝑦 +𝐿𝐼𝑦 −𝑂𝑀𝑦
𝑓1 = ∑𝑁
𝑦=1 (1+𝑑)𝑦
+ 𝑁𝐷 − 𝐶𝐷𝐺 (5)
where d is the discount rate, ND is the network deferral benefit, Ry is the annual sales in energy,
OMy is the annual operations cost, CDG is the capital cost of the EG and LIy is an annual loss
incentive payment for reduction in losses.
A case study using an 83-bus, 11.4-kV radial distribution network was used to examine the
potential economic benefits (Figure 45).
Table 18 - (a) Load bands; (b) group conductors, amp capacity and upgrade cost used in reference [20]
(a) (b)
Load band Duration Active Reactive Losses Existing wires Wires for Upgrade
(h) power power upgrades cost (£)
(MW) (MVAR)
Minimum 2940 5.670 4.140 0.0199 AWG Amps AWG Amps —
Medium 2940 8.505 6.210 0.0450 2 60 1/0 94 64,913.31
Normal 2800 11.340 8.280 0.0807 2 94 1/0 150 142,047.75
Maximum 80 14.175 10.350 0.1272 1/0 150 2/0 190 59,188.56
AWG = American wire gauge
Figure 45 – The 83-bus, 11.4-kV radial distribution network used in reference [20]
| 134
EG was installed at buses 7, 12, 20, 32, 52, 72 and 80, and network deferral benefit was
calculated assuming load growth of 1%, 2% and 3% [20]. EG units were assumed to operate at a
fixed power factor of 0.9 leading. The study focuses on deferment of cable upgrades; the cost of
this was derived from the average marginal investments on feeders per growth in system peak
calculated for 124 utilities in the United States. The values used are given in Table 18(b).
The study showed that gas engines have an investment cost of £600/kW, with operation costs at
£28/MWh. The diesel generators have low installation costs of around £100/kW, but much higher
operational costs of around £150/MWh. The price of energy was assumed to be £60/MWh.
Results of the analysis for 3% load growth and 20-year planning horizon are given in Table 19.
Table 19 - Return on investment at 3% load growth and 20 year planning horizon for distributed generation (DG) [20]
DG location Gas Diesel
DG operating scenario
A B C A B C
7 4.3340 4.5901 — — — —
12 3.5590 3.8523 — — — —
20 2.3766 2.6006 0.5016 0.5016 0.5016 0.5016

Capacity (MW)
32 3.8787 4.2642 0.5906 0.5906 0.5906 0.5906
52 3.5263 3.7904 — — — —
72 2.4286 2.6646 0.3380 0.3380 0.3380 0.3380
80 3.7559 4.0955 0.5400 0.5400 0.5400 0.5400
Total DG 23.8591 25.8576 1.9702 1.9702 1.9702 1.9702
Objective function breakdown
ND 597 597 597 597 597 597
LI –243 -94 2 77 45 2
R 103,830 55,684 118 8,574 4,234 118

£ 000s
OM 48,454 25,986 55 21,435 10,607 295
CDG Pg 14,315 15,515 1,182 197 197 197
Objective 41,415 14,686 -520 -12,385 -5,919 224

function
CDG Pg = Capital ost of DG; LI = Loss Incentive; ND = Network Deferral benefit; OM = annual Operation and
Maintenance costs; R = annual energy sales
It is not specified in [20] how the sizing of the EG is selected. The results show that the
proposed gas installations are profitable for the first two operating schedules, but not the third;
the gas generators do not run long enough to generate the revenue to offset their high capital
| 135
cost. The opposite is true for diesel, but the low capital cost and high running costs fit well
when operating during peak demand periods only.
3.4.9.3 Embedded generation as an alternative option
This section presents work that uses EG to deliver outcomes in power quality and network
services.
Reference [10] performs an NPV calculation for adding STATCOM/BESS to a PV system,
calculating a 13% return on investment for the scenario considered. The services/benefits
provided by a combined PV/STATCOM/BESS, such as MW and MVAR support, ancillary services,
voltage regulation and investment deferral provide a revenue stream and payback on
investment.
Reference [8] examined the use of PV and storage to manage voltage profiles on a rural feeder,
comparing the cost in storage to that of purchasing a DSTATCOM. The required capacity of the
DSTATCOM provides the same voltage profile results as the proposed PV storage control system.
The study found that 60 kVAR of DSTATCOM is required, costing US$3600, while the cost of the
batteries (ZnBr) was US$9600. They note that the DSTATCOM cannot provide peak shaving: only
reactive power support.
3.4.10 Stability and reliability
This section discusses the findings of reference [19], which calculates the increase in network
reliability delivered by the installation of EG in rural Brazil.
The rural network consists of conventional generation with EG distributed among subsystems.
The study investigates how effectively EG can maintain subsystem voltage and frequency after
either a:
i. change in state of the EG within the subsystem
ii. load state transition
iii. protection device event (termed islanding).
| 136
The effectiveness of the EG system was measured by comparing the System Average Interruption
Duration Index (SAIDI) and System Average Interruption Frequency Index (SAIFI), with EG and
without.
The test feeder covers 166 km2, providing electricity to 1865 customers with a total peak load of
1.03 +j0.34 MVA. The rural feeder is weak and characterised by poor supply at certain times of
the year. The rural network consists of 10 sectionalisers (and their 10 respective reclosers) and
41 fuses. Load is stochastically assigned as a percentage of the peak; EG is stochastically
assigned a capacity. If islanding occurs, it is assumed that the EG is able to perform load
following perfectly, up to its capacity. EG systems are individually ramped up or down following
a merit order. The state of the system is determined every hour over a period of 593 simulated
years. In the event of subsystem islanding, the capability of the EG within the subsystem to
meet load demand is calculated and reliability metrics recorded.
The study found that EG does improve reliability within the rural network, as illustrated by
Figure 46. The frequency and duration of inadequate services for the whole system dropped
from 3.6 occurrences per year and 2 hours/year to 0.18 occurrences per year and 0.0978
hours/year, respectively. There were an average of 1.4 islanding events a year, of which 92%
were successful.
b)
a)
Figure 46 - Impact of embedded generation on (a) System Average Interruption Frequency Index (SAIFI); (b) System
Average Interruption Duration Index (SAIDI) [19]
3.4.11 Stocktake
Reference [41] provides detailed information on 176 projects, including 116 studies, trials and
demonstration projects from across Australia and 60 key projects from Europe, the United
States, Korea and Japan, among others). The projects discussed in this section examine the
benefits EG provides to the electrical networks. The majority of the relevant projects in [41] are
either ongoing, with no results to date, or otherwise commercial in confidence; those covered
here are complete with results or findings that are publicly available.
| 137
3.4.11.1 Breaking the solar gridlock – potential benefits of installing concentrating
solar thermal power at constrained locations in the National Electricity Market
Reference [1] examines the potential for concentrating solar power (CSP) to defer network
augmentation.
The methodology of this project consisted of four tasks (see Figure 47):
 Task 1 — identify avoidable network investment
 Task 2 — identify the likelihood of CSP being able to generate during peak-load periods at
the identified locations
 Task 3 — identify the locations where CSP can provide cost-effective network support,
and also determine appropriate plant capacity and configuration
 Task 4 — undertake case studies at constrained locations in Queensland, New South
Wales, Victoria and South Australia.
CSP is defined as being able to meet a network constraint when the indicative firm capacity at
the location for the time and season is above 80%, and a CSP plant of capacity equal to the
maximum projected network constraint could be physically connected at the appropriate
connection point.
The cost effectiveness of CSP replacing network augmentation was assessed by comparing the
CSP plant’s levelised cost of electricity to potential revenue, including a calculated network
support payment. Different CSP plant configurations were assessed, ranging from the minimum
size plant to alleviate the constraint to the maximum size able to be connected without
requiring network augmentation to export energy. The configurations included the assessment of
varying amounts of thermal energy storage.
A reduction of 4% per year was included in the modelling for CSP capital costs to allow for the
projected CSP learning curve: a midrange among estimates for likely cost reduction. The
proposed network investment was reduced by 20% prior to calculating the network support
payment, reflecting the fact that electricity generation (of any type) cannot replicate the
certainty offered by wires and poles. This also means the total societal cost of meeting network
constraints is reduced by 20%. However, the comparison of CSP installation to other non-network
solutions is not considered in this study [1].
Figure 47 - Methodology for identifying concentrating solar power (CSP) potential to defer network augmentation [1]
A total of 92 constraints, or constrained areas, were identified in non-metropolitan areas in the

NEM [1]. Approximately $0.8 billion of potentially avoidable network augmentation was
identified across the NEM in areas with suitable solar irradiance for installation of CSP (defined
here as average direct normal irradiance, which is more than 21 MJ/m2/day); Figure 48 breaks
| 138
this down by time period and state. A further $0.5 billion of potentially avoidable network
expenditure was identified in areas with direct normal irradiance below 21 MJ/m2/day. Most of
the investment occurs from 2016 onwards. This reflects the fact that maximum demand
forecasts were reduced significantly during 2012, deferring many cases of proposed growth-
related augmentation.
Figure 48 - Potential network augmentation projects ($) arranged by state [1]. NSW = New South Wales; QLD =
Queensland; SA = South Australia; VIC = Victoria
Table 20 gives the proportion of grid-constrained locations where CSP could indicatively avoid
the need for further network augmentation. Table 21 summarises the cost benefit of CSP
installed at grid-constrained locations, while Table 22 provides an overview of the results for
each case study.
Table 20 - Proportion of grid-constrained locations where concentrating solar power (CSP) could indicatively avoid the
need for further network augmentation [1]
Queensland New South Victoria South All states

Wales Australia
Number of locations where CSP could indicatively 20 7 17 4 48

avoid the need for network augmentation
Proportion of all locations (%) 87 88 53 100 72
Proportion of location with DNI > 21 MJ/m2/day (%) 90 100 100 100 94
DNI = direct normal irradiance
Table 21 - Cost benefit of concentrating solar power (CSP) installed at grid-constrained locations [1]
Queensland New South Victoria South All states

Wales Australia
Proportion of cost-effective sites (%) 30 0 14 67 25
Proportion of sites with cost–benefit > -$20/MWh 45 17 14 67 39

(%)
| 139
Table 22 - Overview of the results for each case study [1]
Location Network Optimum Proposed Proposed Network Net

operator plant augmentation augmentation payment benefit
(MW/TES) year cost ($/MWh) ($/MWh)
($ million)
The Riverland, South ElectraNet 40 MW, 5 hrs 2022 226 110 144
Australia (line
replacement)
The Riverland, South ElectraNet 130 MW, 5 hrs 2022 10 1 60

Australia (line upgrade)
Charleville, Queensland Ergon 20 MW, 5 hrs 2022 70 6 16
Wemen, Victoria Powercor 77 MW, 5 hrs 2021 12 3 23
Gunnedah supply, New Transgrid 50 MW, 5 hrs 2019 24 9 –13

South Wales (CSP at
Moree)
Millchester, Queensland Ergon 40 MW, 15 hrs 2017 46 16 –29
Gunnedah supply, New Transgrid 50 MW, 5 hrs 2019 30 13 –39

South Wales (CSP at
Gunnedah)
CSP = concentrating solar power; TES = thermal energy storage
3.4.11.2 Effect of small solar photovoltaic systems on network peak demand
Ausgrid examined the impact of distributed PV on peak demand [24]. The estimated peak
reduction was between 0.3 and 1.2%, while the capacity of the installed PV contributing towards
peak reduction ranged between 16 and 86% (averaging approximately 32%). The orientation of
the PV panels is not stated in the report, but it is likely that the majority are north-facing. If
panels were oriented towards the west, then the impact on peak reduction would be larger.
Figure 49 shows the potential afternoon gain in generation if panels were oriented towards the
west. The figure compares typical generation of a north-facing system and a tracking system,
but the generation of a tracking system and a west-facing system will be closer towards the end
of the day. The observed peak period here was around 5:00 pm; at this time, generation of a
west-facing system could potentially be double that of a north-facing system.
| 140
Figure 49 - Generation comparison between dual-axis tracking and fixed tilt [42]
3.4.11.3 Maximising the benefit of photovoltaic utility scale storage at PNM

Prosperity project
The PNM Prosperity Energy Storage Project [43] is a smart grid demonstration project in New
Mexico, United States. The 500-kW PV system has two advanced lead-acid batteries: one is
500 kW and provides fast response for power smoothing, and the other is 990 kWh for peak-load
shifting. The system control algorithm’s optimisation objectives are peak-load reduction,
avoided generation and arbitrage. Figure 50 shows the measured output for the system for a
single day and reveals the effectiveness of the algorithm at shifting output.
Figure 50 - PNM Prosperity Energy Storage system output for a particular day [43]
3.4.12 Summary
While significant focus has been given to the costs associated with the rise of EG, little public
attention has been drawn to the potential for networks to derive significant benefit, both from
existing EG and from targeted incentives or deployment of new systems. The uptake of rooftop
solar PV over the last five years — combined with incentivising appropriate EG storage systems,
and strategic deployment of targeted EG assets, such as network-owned batteries — has
demonstrated potential for significant gains, both economically and in terms of network power
quality and reliability.
Some of these benefits rely on an increasing peak load to enable network investment deferral,
which will make EG economically viable. In several networks, the general trend in recent years
is towards a reduction in peak load. However, this is by no means universal. The above
| 141
techniques are often applicable at numerous network levels; even on an unconstrained feeder,
existing distribution substations could benefit from these techniques.
Several of the methods discussed in this section have benefits that go beyond those for the
transmission NSP or DNSP. Reduced line losses, lower costs of operating generation assets, such
as peaking plants, and increasing renewable generation are all of benefit to disparate network
entities as well as consumers.
| 142
References
1. Rutovitz, J., et al., 2013, Breaking the solar gridlock. Potential benefits of
installing concentrating solar thermal power at constrained locations in the NEM.
Prepared for the Australian Solar Thermal Energy Association (AUSTELA) by the
Institute for Sustainable Futures, UTS.
2. Energex distribution annual planning report 2014/2015 to 2018/2019. Energex:
https://www.energex.com.au/__data/assets/pdf_file/0003/219720/DAPR-2014-
15-to-2018-19-Volume-1.pdf.
3. IEEE Guide for Loading Mineral-Oil-Immersed Power Transformers up to and
Including 100 Mva With 55 Degrees C Or 65 Degrees C Average Winding Rise.
ANSI/IEEE Std C57.92-1981, 1981: p. 0_1.
4. Ralph, M., et al., 2011, Transmission and distribution deferment using PV and
energy storage. In Photovoltaic Specialists Conference (PVSC), 37th IEEE. Seattle,
United States, 19-24 June 2011.
5. Alam, M., K. Muttaqi, and D. Sutanto, 2013, Mitigation of rooftop solar PV
impacts and evening peak support by managing available capacity of distributed
energy storage systems. IEEE Transactions on Power Systems 28(4): 3874-3884.
6. Yang, Y., et al., 2014, Sizing strategy of distributed battery storage system with
high penetration of photovoltaic for voltage regulation and peak load shaving.
IEEE Transactions on Smart Grid 5(2): 982-991.
7. Hill, C.A., et al., 2012, Battery energy storage for enabling integration of
distributed solar power generation. IEEE Transactions on Smart Grid 3(2): 850-
857.
8. Kabir, M., et al., 2014, Coordinated Control of Grid-Connected Photovoltaic
Reactive Power and Battery Energy Storage Systems to Improve the Voltage
Profile of a Residential Distribution Feeder. IEEE Transactions on Industrial
Informatics 10(2): 967-977.
9. Craciun, B., et al., 2014, Frequency support functions in Large PV Power Plants
with Active Power Reserves. IEEE Journal of Emerging and Selected Topics in
Power Electronics 2(4): 849-858.
10. Enslin, J.H.R., 2010, Network impacts of high penetration of photovoltaic solar
power systems. In Power and Energy Society General Meeting, IEEE. Minneapolis,
United States, 25-29 July 2010.
11. Wang, G., M. Ciobotaru, and V.G. Agelidis, 2014, Power smoothing of large solar
PV plant using hybrid energy storage. IEEE Transactions on Sustainable Energy
5(3): 834-842.
12. Teleke, S., et al., 2010, Rule-based control of battery energy storage for
dispatching intermittent renewable sources. IEEE Transactions on Sustainable
Energy 1(3): 117-124.
13. Pezeshki, H., P.J. Wolfs, and G. Ledwich, 2014, Impact of high PV penetration on
distribution transformer insulation life. IEEE Transactions on Power Delivery
29(3): 1212-1220.
14. Nourai, A., R. Sastry, and T. Walker., 2010, A vision & strategy for deployment of
energy storage in electric utilities. In Power and Energy Society General Meeting,
IEEE. Minneapolis, United States, 25-29 July 2010.
15. Agah, S.M. and H. Askarian Abyaneh, 2011, Quantification of the distribution
transformer life extension value of distributed generation. IEEE Transactions on
Power Delivery 26(3): 1820-1828.
| 143
16. Mendez, V., et al., 2006, Impact of distributed generation on distribution
investment deferral. International Journal of Electrical Power & Energy Systems
28(4): 244-252.
17. Agah, S.M. and H.A. Abyaneh, 2011, Distribution transformer loss-of-life
reduction by increasing penetration of distributed generation. IEEE Transactions
on Power Delivery 26(2): 1128-1136.
18. Zhang, Y., C. Gu, and F. Li., 2010, Evaluation of investment deferral resulting
from microgeneration for EHV distribution networks. In Power and Energy Society
General Meeting, IEEE. Minneapolis, United States, 25-29 July 2010.
19. Issicaba, D., J.A.P. Lopes, and M.A. da Rosa, 2012, Adequacy and security
evaluation of distribution systems with distributed generation. IEEE Transactions
on Power Systems 27(3): 1681-1689.
20. Piccolo, A. and P. Siano, 2009, Evaluating the impact of network investment
deferral on distributed generation expansion. IEEE Transactions on Power
Systems 24(3): 1559-1567.
21. Rydh, C.J. and B.A. Sandén, 2005, Energy analysis of batteries in photovoltaic
systems. Part II: Energy return factors and overall battery efficiencies. Energy
conversion and management 46(11): 1980-2000.
22. AEMO, 2014. National Electricity Market (NEM) Rooftop PV forecast. Available
from: http://www.aemo.com.au/Electricity/Planning/Forecasting/NEFR-
Archive/National-Electricity-Forecasting-Report-
2014/~/media/Files/Other/planning/NEFR/2014/2014%20Supplementary/2014_NE
FR_Methodology_PV_annual_energy_and_installed_capacity_v2.ashx.
23. Western Power, 2012, Photovoltaic Forecast Public Version 2012. Available from:
http://www.westernpower.com.au/documents/Photovoltaic_Forecast_Public_Ver
sion_2012.pdf.
24. Ausgrid, Effect of small solar Photovoltaic (PV) systems on network peak demand.
Available from:
http://www.ausgrid.com.au/~/media/Files/About%20Us/Newsroom/Discussions/S
olar%20PV%20Research%20Paper.pdf.
25. Ausgrid. Solar home electricity data. Available from:
http://www.ausgrid.com.au/Common/About-us/Corporate-information/Data-to-
share/Data-to-share/Solar-household-data.aspx#.VYe-ydKUeE4.
26. Noone, B., PV Integration on Australian distribution networks: Literature review.
Centre for Energy and Environmental Markets, University of NSW. Available from:
http://apvi.org.au/wp-content/uploads/2013/12/APVA-PV-and-DNSP-Literature-
review-September-2013.pdf.
27. Liu, Y., et al., 2008, Distribution system voltage performance analysis for high-
penetration PV. In Energy 2030 Conference, ENERGY 2008, IEEE. Atlanta, United
States, 17-18 Nov 2008.
28. Varma, R.K., V. Khadkikar, and R. Seethapathy, 2009, Nighttime application of PV
solar farm as STATCOM to regulate grid voltage. IEEE Transactions on Energy
Conversion 24(4): 983-985.
29. Pudjianto, D., C. Ramsay, and G. Strbac, 2007, Virtual power plant and system
integration of distributed energy resources. Renewable power generation, IET
1(1): 10-16.
30. AEP. AEP Tech Centre. Available from: www.aeptechcenter.com/ces.
31. Morjaria, M., et al., 2014, A grid-friendly plant: the role of utility-scale
photovoltaic plants in grid stability and reliability. Power and Energy Magazine,
IEEE 12(3): 87-95.
32. Enslin, J.H., 2014, Integration of photovoltaic solar power-the quest towards
dispatchability. Instrumentation & Measurement Magazine, IEEE 17(2): 21-26.
| 144
33. Commission, A.E.M., National electricity rules [R/OL]. 2008, E2008-10-213.
Available at: http://www.aemc.gov.au/.
34. Gil, H.A. and G. Joos, 2006, On the quantification of the network capacity
deferral value of distributed generation. IEEE Transactions on Power Systems
21(4): 1592-1599.
35. Power Transformers, Part 7: Loading guide for oil-immersed power transformers.
IEC standard, 2005: pp. 60076-7.
36. Australian Standard AS2374 Distribution Transformers Overview and Test
Procedures. Australian Standard AS2374, 2007.
37. Willis, H.L., 2004, Power Distribution Planning Reference Book. CRC press.
38. Schoenung, S.M. and W.V. Hassenzahl, 2003, Long-vs. short-term energy storage
technologies analysis: A life-cycle cost study. A study for the DOE Energy Storage
Systems Program. Sandia National Laboratories.
39. Sandia, 2013, Grid Energy Storage. Available at:
http://www.sandia.gov/ess/docs/other/Grid_Energy_Storage_Dec_2013.pdf.
40. Hsieh, S.-C., 2015, Economic Evaluation of the Hybrid Enhancing Scheme with
DSTATCOM and Active Power Curtailment for PV Penetration in Taipower
Distribution Systems. IEEE Transactions on Industry Applications 51(3): 1953-1961.
41. MHC Consulting, Integrating Renewables in the Grid - Stocktake. Available at:
http://arena.gov.au/initiatives-and-programmes/integrating-renewables-in-the-
grid/.
42. Heslop, S. and I. MacGill, 2014, Comparative analysis of the variability of fixed
and tracking photovoltaic systems. Solar Energy 107: 351-364.
43. Cheng, F., et al., 2013, Maximizing the benefits of PV utility scale storage at
PNM prosperity project. In Energytech, 2013 IEEE.Cleveland, United States, 21-23
May 2013.
| 145
3.5 Australian Standards gap analysis
Considerable work has been done to address the technical challenges associated with
integrating large volumes of EG in distribution networks. Although there are solutions to these
challenges are available, it is unclear whether current technical standards and regulations
can safely and efficiently support them. In addition, new load types and electricity usage
patterns are changing the requirements for the electricity system. Recent consultations with
industry stakeholders by MHC reveal concerns regarding current standards and connection
processes for EG [1].
The aim of this analysis is to establish whether current Australian standards and regulations
reflect and can facilitate the abovementioned changes. Key gaps are identified and
recommendations made for areas where new or revised Australian standards may be
considered. The scope is limited to gaps of a technical nature in the context of EG, with
priority given to standards for the following areas:
• grid-connection requirements, including standards to resolve power quality and
voltage issues caused by high-penetration solar PV
• remote management, as both a load and source (particularly with energy storage and
EVs), including integration of V2G-enabled EVs
• remote DR, including upstream communications and shifting essential loads to match
peak generation from EG (particularly solar PV)
• safety considerations, including battery installations, off-grid systems and protection
systems (fault ride-through and stability).
3.5.1 Summary of standards
This section provides an overview of the Australian standards that are referred to in this
report. The standards outlined in Table 23 are considered relevant in the context of EG. They
primarily cover inverter grid connection, DR enabling equipment, and safety of storage and
EVs.
Table 23 – Summary of standards relevant to embedded generation
Standard Description Published Reference
AS 4777.1 Grid connections of energy systems via inverters Parts 1—3: [2]–[5]
20/05/2005
AS 4777.2  Part 1: Installation requirements
AS 4777.3  Part 2: Inverter requirements
 Part 3: Grid protection requirements
Part 2 draft:
DR AS/NZS
 Part 2 (DRAFT) : Inverter requirements 31/03/2015
4777.2
Framework for demand response capabilities and Parts 3.1— [6]–[10]

supporting technologies for electrical products 3.3:
20/11/2014
AS/NZS  Part 3.1: Interaction of demand response
4755.3.1 enabling devices and electrical products —
Operational instructions and connections for air
conditioners
Operational instructions and connections for
devices controlling swimming pool pump units
Operational instructions and connections for
electric storage and electric-boosted storage Part 3.4
water heaters draft:
18/07/2013
DR AS/NZS  Part 3.4 (DRAFT): Interaction of demand

4755.3.4 response enabling devices and electrical Part 3.5
products — Operational instructions and draft:
connections for grid-connected 30/04/2015
charge/discharge controllers for electric
vehicles
DR AS/NZS
4755.3.5  Part 3.5 (DRAFT): Interaction of demand
response enabling devices and electrical
products — Operational instructions and
connections for grid-connected electrical energy
storage (EES) Systems
AS 4086 Secondary batteries for use with stand-alone power 05/10/1997 [11]
systems
| Page 147
AS/NZS Stand-alone power systems — Safety and installation 20/12/2009 [12]
4509.1
AS ISO Electrically propelled road vehicles — safety 30/06/2014 [13]

6469.1 specifications
 Part 1: On-board rechargeable energy storage
system (RESS)
AS/NZS Australian/New Zealand Wiring Rules 12/11/2007 [14]
3000
AS/NZS Electrical installations — generating sets 27/06/2005 [15]

3010
3.5.2 Gaps in standards
3.5.2.1 Network connection requirements for embedded generation
This section discusses gaps in standards relating to grid-connection requirements for EG.
These connection requirements cover:
 protection and control
 power quality
 voltage
 fault current
 remote monitoring.
The key gaps identified in the analysis include:

1. A lack of technical standards for the grid connection of EG less than 5 MW and not
connected at the LV level.
2. A uniform approach needed to address voltage rise, which should be reflected in
AS/NZS 4777.
3. Standards are required for protection relays for inverter energy systems.
Embedded generation less than 5 MW

The current version of AS 4777 specifies the electrical installation requirements for inverters
and grid protection devices. It currently applies only to ‘micro’ EG: single-phase units with a
generating capacity less than 10 kVA and three-phase units less than 30 kVA. Residential
systems (e.g. rooftop solar PV) would generally fall into this category.
The grid-connection requirements for EG rated between 30 kVA and 5 MW are currently
determined by DNSPs. These requirements are not clear in general and can vary significantly
between jurisdictions. Connection requirements for systems above 5 MW are well covered
| Page 148
under chapters 5 and 5A of the NER [16], [17]. Recent revisions to chapters 5 and 5A of the
NER (effective 1 March 2015) provide connection processes for EG below 5 MW. However,
these processes are largely non-technical and include items such as connection timeframes
and contract negotiation. Additionally, Chapters 5 and 5A do not currently apply in
Queensland or Victoria. The Queensland and Victorian governments announced plans to
implement the National Energy Customer Framework (NECF), which will take effect on 1 July
2015 (Queensland) and 31 Dec 2015 (Victoria). Chapter 5A and its recent amendments, along
with the implications of these, are explored in further detail in Section 4.2.1.2.
Revisions to AS/NZS 4777.2 (inverter requirements) are currently being drafted and will
supersede AS 4777.2 and AS 4777.3 (grid protection requirements) [9]. The draft proposes
extending the coverage of the standard to include all inverters connected at an LV level. This
excludes MV-connected EG; consequently, there may still be a desire for technical connection
standards to be extended for EG that is less than 5 MW and not connected at the LV level.
Gap 1
A lack of technical standards for the grid connection of EG less than 5 MW and not
connected at the LV level.
Restrictions due to high-penetration solar photovoltaics

Increasing solar PV uptake has caused concerns among DNSPs about the cumulative effect
that grid-connected systems may have on distribution networks. Perhaps the most commonly
reported issue regarding high volumes of PV is the reversal of power flow, causing voltage
rise. This happens when the power generated from PV is higher than the load demand
(typically in the middle of the day), which can drive node voltages above specified limits.
To address the issues surrounding increased PV uptake, several DNSPs have introduced PV
system capacity limits and other solutions in constrained areas. For example, Ergon Energy
and Energex require that all newly installed AS 4777 compliant inverters must be able to limit
power exports (either built-in or by separate relay) [18]. Since 2012, Horizon Power require
some solar PV installations (depending on tariff, system size and network hosting capacity) to
have generation management (storage) to control ramp rates [19]. This requirement was
complicated by a lack of compatible inverters, and it took approximately 12 months for the
market to respond.
MHC recently consulted with industry stakeholders, including representatives from the AEMC,
AEMO and Australian Energy Regulator (AER) as well as various utilities. The stakeholders
proposed that, for example, instead of DNSP-imposed generation management at the
customer level (as with Horizon Power), feed-in tariffs (FiTs) could be reduced and the funds
used to implement centralised storage at the network level [1].
The approaches taken by DNSPs responding to the technical challenges of PV are diverse
across Australia. Typical measures to limit the impacts of PV include:
 OLTC (at MV/LV transformer) to reduce voltage levels at times of high PV generation
| Page 149
 power-limiting relays
 SVCs
 network reconfiguration
 energy storage at the customer level – prosumer storage
 reactive power control by PV inverters
 DR through price signalling to shift loads to times of high PV generation
 direct load control by DNSPs (remote demand management).
Developing a more consistent approach that allows for diverse network characteristics may be
beneficial. The technical means to achieve this could be standardised and handled under
inverter connection standards such as AS 4777; some of this is being addressed in the ongoing
review of AS/NZS 4777 Part 2. A collaborative approach with DNSPs to develop standards
would be appropriate to ensure the solutions are viable.
Gap 2
DNSP’s diverse solutions to voltage rise and other EG issues are inconsistent, and
complicate development and manufacturing of inverter systems. A more uniform
approach, which is reflected in AS 4777, should be explored.
3.5.2.2 Protection
AS 4777 and most DNSPs require backup or central protection; however, no technical
standards exist for protection relays. These standards are needed because locally developed
limited export products are entering the market with additional protection functions.
Gap 3
No current standards exist or are under development in Australia governing the
functionality of protection relays for inverter energy systems.
3.5.2.3 Remote management of embedded generation (energy storage and electric

vehicles)
This section discusses gaps in standards relating to the remote management of EG that can be
both a load and source (particularly energy storage and EVs). The key gaps identified in the
analysis include:
4. A lack of standards in general governing the integration of EVs, including V2G-enabled
EVs.
5. A lack of a standardised, well-defined communications protocol for smart meters,
whether a backend meter protocol or a shared-market protocol managed by the
‘gate keeper’, as proposed by the AEMC.
| Page 150
Electric vehicles
EVs with V2G capability allow excess energy stored in batteries to be exported to the grid.
Standards and regulations must facilitate the integration of such vehicles safely and
efficiently. EG in the form of EVs is not well covered in general by Australian standards.
There is an opportunity for AS 4777 and AS/NZS 4755 to be extended to include EVs in the
following areas:
 DR — operational instructions and connections for grid-connected EVs (a draft for this
was published in 2013, but is currently on hold)
 connection requirements for V2G power feedback
 EV supply equipment and associated bidirectional communications protocols — e.g.
charging stations, plugs, other fittings.
Some DNSPs, as well as the Standards working group, have noted that the lack of standards is
unlikely to be an issue in the short term, given the reasonably slow uptake of EVs in Australia.
However, the gap remains, and a proactive approach would ensure the industry is well
positioned in case EVs are adopted faster than expected. Several relevant international
standards in this area could be adopted or modified for adoption in the Australian market.
Gap 4
There is a lack of standards in general governing the integration of EVs, including V2G-
enabled EVs.
Metering energy storage and electric vehicles

This section reviews metering infrastructure in the context of EG. Specifically, it looks at
issues surrounding the metering of energy storage and EV imports/exports that may be
hindering EG uptake.
First, net load metering can be problematic if we consider a customer that, for example, may
be importing on one phase and exporting on another. How should customers be billed in these
scenarios? How is the asymmetry between import and export price models dealt with? These
questions are complicated by smart meter and interval meter recordings being taken only at
half-hourly intervals. These complex issues are not currently reflected in Australian
standards. We recognise, however, that this may be more of a market rules issue than a gap
in standards.
Second, there is a lack of consistency between DNSPs across Australia relating to upstream
communications. Smart meters communicate interval data back to DNSPs over secure, private
‘mesh radio’ networks. A backend protocol system (defining communications from the smart
meter to the DNSP) is currently not specified in Australian standards. Additionally, different
jurisdictions may have different communications requirements. In Victoria, mesh radio
networks provided by SilverSpring or Wimax are used, as well as the well-established, special-
purpose wireless ‘Zigbee’ protocol for home area networks [20].
| Page 151
The Australian Energy Market Commission (AEMC) has published support for a shared-market
protocol, rather than a common backend protocol between smart meters [21] (Figure 51).
Their reason for rejecting a common smart-meter protocol is that allowing proprietary
protocols would reduce initial investment costs and increase the flexibility of new services,
allowing them to be introduced more quickly. They make the point that if a shared-market
protocol were specified, there would be no need for a standardised meter protocol, because
all authorised parties (such as electricity retailers) would communicate via the market
protocol. They recommend that the market protocol be built by extending the current
arrangements in place for business-to-business communications managed by AEMO.
Figure 51 – Australian Energy Market Commission vision of shared-market protocol [21]
The AEMC suggest that shared meter protocol would specify a protocol for different functions
known as ‘agreed metering services’ while maintaining provision for alternative functions and
new services through alternative market protocols (Figure 52). An example of an agreed
service might be the remote control of an appliance that is facilitated by the smart meter.
The alternative market protocols would provide a means for accessing agreed metering
services as well as other services. This would allow the benefits of alternative
communications to be captured. The AEMC recommends further work to define the scope of
the agreed metering services [21].
| Page 152
Figure 52 – Access to agreed market services and other services, as suggested by the Australian Energy Market
Commission [21]
A standardised protocol of some kind (whether meter or market shared) would therefore be
beneficial. It would need to be developed with collaborative input from DNSPs to ensure
existing protocols are incorporated smoothly.
Gap 5
A lack of a standardised and well-defined communications protocol for smart meters,
whether a backend meter protocol or a shared-market protocol managed by the ‘gate
keeper’, as proposed by the AEMC.
3.5.2.4 Remote demand management
Remote demand management refers to the ability of DNSPs to shift certain customer loads to
reduce peak demand, thereby increasing grid efficiency and infrastructure savings. Loads can
be brought forward to match periods of peak renewable generation, such as the middle of the
day for solar PV. This section reviews the standards for remote demand management and the
role it plays in facilitating EG.
AS/NZ 4755 specifies a framework for DR capabilities and supporting technologies for
electrical products including air conditioners, swimming pool pump units and electric water
heaters. The aim of the standard is to ensure that these types of appliances are
manufactured with interfaces and communications protocols that allow them to be controlled
remotely.
6. Demand-response mode (DRM) 4 is not defined in current demand-response
frameworks for air conditioners.
| Page 153
7. A lack of standards for a feature-rich, bidirectional communications protocol to
facilitate DR.
Air conditioners
Air-conditioning systems are one of the main drivers of summer peak demand in Australia
[22]. These systems have the potential to facilitate EG through DR. Research by CSIRO and
others have demonstrated the advantage of pre-cooling residential houses during the middle
of the day when there is a surplus of EG (primarily solar PV), to offset evening peak demand.
While there are arguments for and against this type of DR, the idea is gaining more
recognition. There are gaps in the treatment of air conditioners under AS/NZS 4755.3.1 that
preclude this capability from the DR market.
AS/NZS 4755 defines the following DRMs:
 DRM 1 — turn appliance off or operate at minimal load (this is the only mode required
for an appliance to be AS/NZS 4755 compliant)
 DRM 2 — operate appliance at no more than 50% load
 DRM 3 — operate appliance at no more than 75% load
 DRM 4 — turn on appliance by bringing the load forward (but not extending daily
running hours).
DRM 4 is currently actively excluded from the air-conditioner interface in AS/NZS 4755.3.1,
effectively prohibiting the shifting of air-conditioning loads for pre-cooling. Reasons for not
wanting air-conditioner control are valid, including concerns about how to charge customers
for the usage; however, with appropriate controls, these issues could be addressed. The gap
in the current AS/NZS 4755.3.1 is that it specifically precludes the option for manufacturers
to implement DRM 4 and for customers to enter into demand-management agreements (which
may well have the appropriate controls to respond to the issues raised).
While the ability to turn on load using DRM 4 in AS/NZS 4755 does exist for other appliances,
including pool pumps and water heaters, uptake of this standard suite by manufacturers of
these appliance types has been extremely limited. For this reason, it would be beneficial to
have DRM 4 defined, because it would encourage the manufacture of DRM 4-compatible air
conditioners and take advantage of pre-cooling.
Gap 6
There is currently no standardised method of bringing forward air-conditioning load (pre-
cooling) to manage local issues caused by excess EG.
Upstream communications protocols

DR communications protocols in Australia are currently relay based and one directional. The
DR standards AS/NZS 4755 series may benefit from the inclusion of a standard for a more
advanced protocol that supports:
 detection of the type of appliance (e.g. air conditioner vs EV)
| Page 154
 validation that the DR instruction has been received
 communication about the operating status of the appliance and the types of DRMs that
it supports.
AS/NZ 4755 describes a DRED interface to which a common upstream communication protocol
can be built. No standardised upstream communication protocol is in place for DREDs. As
such, there is no means for manufacturers to standardise a common protocol.
Bidirectional communications protocols have been implemented overseas with success. An
example is the OpenADR specification that has been trialled in the United States [23]. The
OpenADR specifications describe a standards-based model for communications between DNSPs
and end consumers. In 2014, the IEC approved the OpenADR 2.0 specification as a publically
available specification (IEC/PAS 62746-10-1) and announced that it would form the basis for
the commission of a new standard to be developed. OpenADR has since been adopted in other
countries, such as India and parts of Asia.
These international resources could be adapted to develop more comprehensive standards
that are appropriate in an Australian setting.
Gap 7
A lack of standards for a feature-rich, bidirectional communications protocol to facilitate
demand response.
3.5.2.5 Safety
This section reviews standards for safety in the context of EG. It covers battery installations,
protection systems (fault ride-through) and off-grid systems.
8. A lack of standards in general for energy storage, particularly with regards to safety.
9. A lack of standards governing fault ride-through for non-inverter EG.
10. A lack of standards for off-grid EG, particularly with regards to safety.
Energy storage
As a result of increasing EG penetration and decreasing energy storage costs, the uptake of
emerging energy storage technologies in the future is expected to be widespread. Standards
need to be sufficient to facilitate the growth in this emerging industry, particularly with
regards to safety.
Australian standards for energy storage are currently fairly limited, in particular with regards
to safety and network connection requirements. This deficiency poses safety risks and may
delay the uptake of emerging energy storage technologies in Australia. The standards that do
exist are largely focused on mature battery technologies and do not extend to the whole
energy storage system connection to the grid. For example, AS 4086 deals with secondary
batteries for use with stand-alone power systems, which are not designed to inject power to
| Page 155
the grid. AS/NZS 4509 covers the safety and installation requirements for these. Standards for
the grid connection of energy systems via inverters also exist in the AS 4777 series, and the
upcoming AS/NZS 5139 addresses some safety aspects of energy storage systems with inverter
energy systems. However, these standards are limited at a commercial scale, as discussed in
Section 3.5.2.1. Some EV standards cover safety requirements for EV batteries (e.g. electrical
isolation testing in AS ISO 6469.1), but not in great detail. AS/NZS 3000 and AS/NZS 3010
include earthing and neutral connection requirements.
Additionally, while there is generally an obligation for customers to advise DNSPs when they
have installed systems such as energy storage, this is not always adhered to. There is a
concern that this could have implications for safety. It must also be recognised within
standards that energy storage can be both a load and an EG (via an inverter).
Standards for energy storage are being developed internationally. A review of relevant
battery standards is available from http://batterystandards.vito.be/. It would be beneficial
to review how these standards may be relevant and applied in an Australian setting.
There is consequently a need for safety standards to support the growth of emerging energy
storage technologies. They should comprehensively cover a range of new technologies (e.g.
lithium and flow batteries, flywheels). The Clean Energy Council is currently reviewing this
area.
Gap 8
A lack of standards in general for energy storage, particularly with regards to safety,
although work is currently ongoing in this area.
Protection systems (fault ride-through and stability)

With high penetrations of solar PV, disturbances such as sudden cloud cover or sudden load
coming online can significantly increase net load, resulting in large power swings (real and
reactive power oscillations) and voltage dips in the network. These issues are of particular
concern on high-impedance networks, as noted in Section 3.2.5.1.
Because these oscillations are often low in magnitude and well-damped, a system will usually
return to a steady state quite quickly. It may be preferable for EG devices to ‘ride through’
the LV event to avoid a possible chain reaction of generators going offline. There is potential
for EG to stay operational during a fault or other LV event and support the grid with reactive
power, if necessary.
Section 5.2.5.8 of the NER specifies fault ride-through requirements for large-scale
generators. It may be beneficial to outline similar requirements relating to LV ride-through
for EG, particularly in the future with respect to DNSP-operated microgrids. LV ride-through is
currently not dealt with in AS 4777, and systems must disconnect from the grid in the event
that these conditions are met. The current revision of AS/NZS 4777.2 includes a trip delay of
one second for low-level under/over-voltage and under-frequency events to improve fault and
LV ride-through capability. However, it will take some time for this to filter through to
| Page 156
deployed inverters. Note that this restriction applies only to LV-connected inverters, and not
to other inverters or non-inverter generation.
Gap 9
A lack of standards governing LV ride-through for non-inverter EG.
Off-grid systems
AS 4777 is currently not mandatory, although most DNSPs and renewable energy certificates
require it. As such, there is no obligation for off-grid systems to comply with the standard.
Parts of the standard are relevant to safety whether grid-connected or not, such as the
requirement for isolation switches, which isolate the inverter energy system for people
working on other parts of the electrical installation. Inverter signage and labelling
requirements may also present a safety concern for off-grid systems that are not obligated to
comply.
Since off-grid systems are not covered by other standards, there is a recognised gap here with
regards to safety. International standards such as IEC 62109-2 Safety of power converters for
use in photovoltaic power systems – Part 2: Particular requirements for inverters may be
adopted for standalone PV systems.
Gap 10
A lack of standards for off-grid EG, particularly with regards to safety.
| Page 157
References
[1] MHC, 2015, ‘FPDI Review of Policies and Incentives and Advice on Policy Responses’.
[2] Standards Australia, 2005, ‘AS 4777.1’.
[4] Standards Australia, 2005, ‘AS 4777.3'.
[5] Standards Australia and Standards New Zealand, 2015, ‘DR AS NZS 4777.2’.
[6] Standards Australia and Standards New Zealand, 2014, ‘AS/NZS 4755.3.1’.
[9] Standards Australia and Standards New Zealand, 2013, ‘DR AS/NZS 4755.3.4’. [Online]. Available:
http://infostore.saiglobal.com/store/Details.aspx?ProductID=1639036. [Accessed: 28-Apr-2015].
[10] Standards Australia and Standards New Zealand, 2015, ‘DR AS/NZS 4755.3.5’.
[12] Standards Australia and Standards New Zealand, 2009, ‘AS/NZS 4509.1’.
[13] Standards Australia and International Organization for Standardization, 2014, ‘AS ISO 6469.1’.
[14] Standards Australia and Standards New Zealand, 2007, ‘AS/NZS 3000’.
[15] Standards Australia and Standards New Zealand, 2005, ‘AS/NZS 3010’.
[16] AEMC, ‘National Electricity Rules Chapter 5: Network connection, planning and expansion’.
[17] AEMC, ‘National Electricity Rules Chapter 5A: Electricity connection for retail customers’.
[18] Energex and Ergon Energy, ‘Small Scale Parallel Inverter Energy Systems up to 30 kVA’.
[19] Horizon Power, ‘Generation management’. [Online]. Available:

http://www.horizonpower.com.au/renewable_energy_generation_management.html.
[Accessed: 28-Apr-2015].
[20] Victorian State Govt., ‘Lockstep DPI AMI PIA Report: Part 1’. [Online]. Available:
http://www.smartmeters.vic.gov.au/about-smart-meters/reports-and-consultations/lockstep-
dpi-ami-pia-report/scope. [Accessed: 27-May-2015].
| Page 158
[21] AEMC, 2015, ‘Final report: framework for open access and common communication standards’.
[Online]. Available: http://www.aemc.gov.au/getattachment/a46fc6c4-993f-486f-810e-
e306300562ab/1-Final-Report.aspx. [Accessed: 12-Jun-2015].
[22] George Wilkenfeld and Associates Pty Ltd, 2004, ‘A national demand management strategy for
small airconditioners’. [Online]. Available: http://www.energyrating.gov.au/wp-
content/uploads/Energy_Rating_Documents/Library/Cooling/Air_Conditioners/200422-ac-
demandmanagement.pdf. [Accessed: 28-Apr-2015].
[23] National Institute of Standards and Technology, United States Department of Commerce,
‘Building Integration with Smart Grid Project.’ [Online]. Available:
http://www.nist.gov/el/building_environment/mechsys/ibasug.cfm. [Accessed: 28-Apr-2015].
| Page 159
4 Regulatory impact assessment
Note on authorship
This section was written by Marchment Hill Consulting (MHC) and reviewed by CSIRO. It
highlights links to the technical impacts relating to the findings of this chapter.
4.1 Introduction
This section provides an overview of the current regulatory framework as it relates to
connecting embedded generation (EG) in Australia, and considers its suitability in light of the
rise in EG connections over the past five years. It also considers the cost of facilitating
connection and how these costs should be recovered.
Document map
Regulatory
Chapter
impact key Network level Network type Impact Timing
section
findings
EG connection
4.2 and 4.3
regulations
Costs of connection 4.4
4.2 The National Electricity Customer Framework

The National Electricity Customer Framework (NECF) is a set of national laws, rules and
regulations that govern the sale and supply of energy to consumers. The NECF has been
adopted in the Australian Capital Territory, New South Wales, South Australia and Tasmania.
Queensland is set to introduce the framework on 1 July 20159, with Victoria to follow suit
from 31 December 2015.
The NECF includes the National Energy Retail Law, National Energy Retail Rules, and
amendments to existing national regulation, including the introduction of Chapter 5A into the
National Electricity Rules (NER).
Under the NER, network service providers (NSPs) are obligated to facilitate connection
services for EG, which is regulated under chapters 5 and 5A:
9The Queensland Government announced in April 2015 that it would hold off on implementing part of the NECF
known as the ‘market monitoring regime’, although this is not expected to affect Chapter 5 or 5A.
Regulatory impact assessment | Page 160

 Chapter 5 regulates registered market participants; i.e., connections in excess of
5 MW.
 Chapter 5A regulates connections not exceeding 5 MW, including small-scale solar
photovoltaics (PV).
Chapter 5A sets out the connection arrangements for non-registered participants and micro-
embedded generators (connections considered under Australian Standard 4777). It includes
the processes for:
 basic connection services
o connections related to a certain defined customer type
o e.g. residential solar PV and micro-EG (less than 30 kW)
 standard connection services
o connection services that are not basic (i.e. 30 kW — 5 MW) but where the
Australian Energy Regulator (AER) has approved a model standing offer (as
detailed under NER 5A)
 negotiated connection services
o connections that do not fall under either of the above categories
o participants may elect to exercise this option even if they are entitled to a
basic or standard connection.
Basic and standard connection arrangements have been put in place to allow for a more cost-
effective, streamlined connection process. However, given the complexities involved and
varying network characteristics, distribution network service providers (DNSPs) may
determine whether the connecting party must opt for a negotiated connection on a case-by-
case basis.
Negotiated connections provide a framework to ensure that parties have sufficient
information to negotiate effectively and in good faith.
4.2.1 Reforms
Recent reforms have been undertaken to amend chapters 5 and 5A of the NER to clarify and
simplify the EG connection process. These are described below.
4.2.1.1 Chapter 5
Chapter 5 was significantly amended in 201410 to simplify the connection process and ensure
transparency about connection costs and technical requirements. Key provisions of the final
rule include:
10Australian Energy Market Commission, Rule Determination [ERC0147] – National Electricity Amendment
(Connecting Embedded Generators) Rule 2014, 17 April 2014.

 publication of a connection process information pack by the DNSP, including
application requirements and cost estimates
 a new two-stage connection enquiry process and application process, including clearly
stated information requirements
 a register of all EG connected to the DNSP’s network over the past five years
 a dispute resolution process that allows parties to use the Wholesale Energy Markets
Dispute Resolution Adviser in accordance with Chapter 8 of the NER
 allowing DNSPs to charge a fee to recover the costs of preparing detailed enquiry
responses.
This rule change came into effect on 1 October 2014.
4.2.1.2 Chapter 5A
The key change to Chapter 5A of the NER11 involves allowing EG proponents to choose their
preferred connection process. It applies to EG less than 5 MW that is not provided with a
standard connection offer and that is not a micro-embedded generator.
EG proponents can now choose between the connection process set out in Chapter 5 or the
negotiated connection service detailed in 5A.12 This added flexibility allows them to choose
the process most suitable to their specific circumstances, which in turn should result in a
more efficient connection process.
This rule change came into force on 1 March 2015.
4.3 Non-National Electricity Customer Framework states

In addition to the NER, Australian states have certain state-specific legislation relating to
connecting EG. This is particularly important in states that have not yet adopted the NECF,
where the local service rules guide the connection process (e.g. the Victorian Essential
Service Guideline 14 and 15 in Victoria. Further non-National Electricity Market (NEM)
jurisdictional arrangements are detailed below.
4.3.1 South West Interconnected System

Western Australia operates its own network system: the South West Interconnected System
(SWIS). This is not governed by the same rules and regulations that apply to the NEM; instead,
network access is regulated and guided by the DNSP’s determination process. Three separate
arrangements are available, based on the capacity of the connection:
11Australian Energy Market Commission, Rule Determination [ERC0158] - National Electricity Amendment
(Connecting Embedded Generators Under Chapter 5A) Rule 2014, 13 November 2014.
12
The adoption of appropriate standards should also be considered in this context. Please see Section 3.5.2 for
more information.

 Small-scale solar PV (<30 kVA) follows a three-step process in which an application is
first approved by the retailer before being submitted to the relevant DNSP.
 Connections with a capacity of 30—150 kVA are approved under a two-step process in
which an enquiry is submitted to the DNSP, followed by an enquiry consultation. The
application can then be submitted.
 Connections with a capacity >150 kVA follow a two-step process similar to the above.
Unlike the NEM, EG owners must also pay an additional ongoing distribution use of system
(DUOS) charge.
Stakeholders in the SWIS have reported issues relating to the connection process for 150 kVA
and larger EG. Applications may be bundled into a competing access group, in which the cost
of augmenting the network to facilitate connection of all applicants is spread across each
applicant based on their required connection capacity. These costs are then offered to all
members of the group. If any members withdraw their application, the costs are then
recalculated and offered to the remaining members, unless the total required generation
connection capacity still exceeds the capacity being offered as the group’s (network
augmentation) solution. In this case, the offers will be made based on the order in which the
applications were received.
Anecdotal evidence collected by MHC suggests this is a time-consuming process that may
cause uncertainty for investors regarding the costs and timeframes for connection.
4.3.2 Horizon Power

Horizon Power supplies electricity to rural and remote customers in Western Australia outside
the SWIS. The connection process on Horizon’s network follows a three-step process of
application, assessment and installation.
As part of the technical assessment of a proposed solar PV connection, Horizon may insist that
storage is included as part of the installation to provide smoothing and prevent network
instability. This sets their connection process apart from the rest of Australia’s DNSPs.
Horizon’s potential requirement for storage applies to all installations >5 kW and to
installations of <5 kW where the hosting capacity of the customer’s town has been reached.
4.3.3 Northern Territory

Power and Water introduced changes to its connection process in 2014. The customer and its
installer must still submit the required documentation, including a connection agreement,
wiring diagram and certificate of compliance. However, residential installations (not
exceeding 4.5 kW on single-phase and 6 kW on three-phase) are automatically pre-approved
without the need for additional inspections or an engineering assessment.

4.3.4 Reforms
In addition to Victoria and Queensland potentially joining NECF over the coming year,
Western Australia is currently undertaking a significant review of its regulatory framework.
Launched in March 2014, the Electricity Market Review covers the structures of the electricity
generation, wholesale and retail markets within the SWIS. It also investigates incentive
options that would help industry participants to improve their investment decisions, and
thereby realise efficiencies and minimise costs.The current connection and access
arrangements are being reviewed under the Network Regulation work stream. The reforms
will transition the current arrangements to chapters 5 and 5A of the NER. They are currently
in the scoping stage, with a decision on the reforms to be implemented expected to occur
progressively over 2015—17.
The Northern Territory government is also introducing the National Electricity Law and NER in
its jurisdiction, and plans to complete this by 1 July 2015. Once completed, this would
introduce the relevant chapters of the NER to guide EG connections and transfer the
regulatory oversight obligations to the AER.13
4.3.5 Survey responses

The survey indicated that in general, NSPs find the current regulatory framework adequate in
meeting both NSP and customer needs to facilitate safe, effective and efficient connection of
EG.14 However, further refinements are likely be required when different types of EG become
prevalent in the market and new value streams and business models emerge.
Concerns were raised about the yet unknown effect of the recently introduced reforms on
NSP operations and costs. This should be closely monitored to gain a better understanding of
the potential impacts.
Key Finding: Although recent reforms to Chapter 5 and 5A of the National Energy Rules (NER)
have improved the connection process related to EG, the lack of a consistent national
A high-level assessment of the characteristics of this finding against the document map key is
shown below.
13
Department of Treasury and Finance, Northern Territory Electricity Market Reform – Information Paper, February 2014
14
For more details regarding the survey, please see Appendix A

Note: the impact is largely dependent upon the jurisdiction. Medium impact refers primarily to jurisdictions that
have not adopted reforms relating to EG connection.
4.4 Cost of connection

The price charged to EG owners for connection services is based on the costs incurred by the
DNSP in facilitating the connection. It takes into consideration the DNSP’s obligation to
maintain network reliability, safety and quality of supply.
The connection charges are regulated by the AER15, which sets out a guideline for DNSPs to
determine the allowable connection cost. DNSPs must further submit a ‘model standing offer’
for basic and standard connection services to the AER for approval. The offer sets out the
connection process and contractual arrangements between the DNSP and the connection
applicant.
The fees and charges payable to the DNSP under the model standing offer are generally made
up of:
 network charges
o as determined by the AER, payable by the customer’s retailer
 site inspection
o the DNSP may determine that a site inspection is required to determine the
nature of the connection, payable by the connection applicant
 ancillary services
o fees related to site establishment (e.g. updating load data, issuing meters,
assigning a National Metering Identifier) and inspection of service work.
Under a negotiated connection service, the DNSP is allowed to charge a fee to cover expenses
directly and reasonably incurred from assessing the application and preparing a connection
offer.
According to NER 5A, DNSPs are not allowed to charge EG owners for network augmentations
if the application is for a basic connection service (as defined in Section 4.2 above), or if a
relevant threshold set in the DNSP’s Connection Policy is not exceeded.
The Connection Policy is guided by the AER’s Connection Charge Guidelines for Electricity
Retail Customers, and sets out various guiding principles to determine an appropriate
15Australian Energy Regulator, Connection charge guidelines for electricity retail customers, Version 1.0, June
2012.

threshold. The AER also provides the following recommended thresholds that fulfil the
guidelines:
 25 kVA on single-wire earth return (SWER) lines
 the maximum capacity of a 100-Ampere three-phase low-voltage supply elsewhere on
the network.
4.4.1 Connection cost recovery

DNSPs are not allowed to charge EG owners of small residential solar PV systems (basic
connections) for augmentation work, as a result of high levels of solar PV penetration on their
network. This has caused issues relating to how these costs may be recovered by the DNSP.
Ergon and Energex determined the costs forecasted to be incurred due to additional
administrative work (e.g. call centre costs, billing) and infrastructure required as a result of
high levels of solar PV on their network16. The estimated infrastructure cost is shown in Figure
53.
$18
$16
$14
$ million, nominal
$12
$10
$8
$6
$4
$2
$-
2010-11 2011-12 2012-13 2013-14 2014-15 2015-16 2016-17 2017-18 2018-19 2019-20
Energex Ergon
Figure 53 – Solar photovoltaic-related administrative and infrastructure costs
Source: Queensland Competition Authority
As not all of these costs could be recovered under the revenue allowance determined by the
AER, the Queensland Competition Authority indicates that Ergon and Energex likely had to
defer other capital and operating expenditure work to fund the necessary investments
resulting from solar PV.
Additionally, DNSPs may charge connection applicants for augmentation from standard and
negotiated connection services, but only up to ratio of the asset being used by the EG in
16Queensland Competition Authority, Estimating a Fair and Reasonable Solar Feed-in Tariff for Queensland, March
2013.

question. This means that the rest of the investment must be recovered from the DNSP’s
customer base.
Key Finding: Where augmentation costs may occur because of high levels of small scale
solar PV penetration on a network (i.e. basic connections), the NER does not allow for
these costs to be recovered directly from the EG owners in a cost-reflective approach.
Instead, these costs are recovered from the entire customer base which introduces an
element of cross-subsidisation between customers.
shown below.
Note: The impact and timing of this finding is largely dependent upon the jurisdiction and associated network
capacity and EG penetration levels. The given ratings are based on evidence suggesting that this finding is of
moderate significance to some jurisdictions in the near term.

5 Commercial impact assessment
Note on authorship
This section was authored by Marchment Hill Consulting (MHC) and reviewed by CSIRO. It
highlights links to the technical impacts relating to the findings of this chapter.
5.1 Introduction
Growing levels of embedded generation (EG) are challenging network service providers (NSPs)
to meet their responsibilities to maintain a safe, secure and reliable power supply. However,
EG can also provide NSPs with new opportunities, such as addressing constraints on the
network and deferring capital investments.
EG introduces new nuances and complexities to the calculation of market benefits for
electricity network planners and managers. Benefits calculation is no longer as simple as
increasing capacity on a single network element. With the introduction of EG, network
planners and strategists must now understand and calculate:
 the potential impacts of an EG asset on adjacent network elements during emergency
conditions
 how an EG asset can reduce energy at risk across the network element on which it is
located as well as upstream elements
 how EG response times affect energy at risk and market benefits.
These and other considerations must be taken into account for electricity networks to
determine the cost—benefit outcome of any EG solution for each network element.
This section puts the commercial impacts and opportunities of EG into context, and broadly
covers three main areas:
i) key commercial impacts (positive and negative) of EG on the network
ii) a framework that places a value on relevant commercial impacts
iii) commercial opportunities that EG presents for NSPs.
Commercial impact assessment | Page 168

Document map
Commercial
Chapter
impact key Network level Network type Impact Timing
section
findings
Network Capacity 5.2.1
Network Charges 5.2.4
Valuation
5.3
Framework
Commercial
Opportunities – price 5.4
signals
Commercial
Opportunities – 5.4
regulatory barriers
Commercial
Opportunities – 5.4
metering reform
Business Models –
5.5
partnerships
Business Models – EG
5.5
integration
5.2 Commercial impacts

The ongoing change in how electricity is produced, delivered and consumed is changing the
role and management of the electricity system. The rise of solar photovoltaics (PV) and the
increased consumer interest in gaining choice and control over their electricity consumption
and expenditure are further changing the energy market.
The impacts of EG can broadly be grouped into the following categories:
 network impacts

o how increasing levels of EG on the network affect NSPs’ ability to provide
reliable and safe supply as economically as possible17
o e.g. voltage regulation, power quality, network reliability
 customer impacts
o the effect on customers investing in EG and the broader customer base
o e.g. EG investment and cross-subsidisation
 generation and retail impacts
o how EG affects centralised generation and the operating model of retailers
o e.g. hedging and arbitrage opportunities
 societal impacts
o the broader effects of EG on overall society
o e.g. environment and public health.
A high-level overview of the commercial impacts is provided in Table 24.
Table 24 – Embedded generation (EG) commercial impacts
Stakeholder Impact Description/hypothesis
Network service Network EG can reduce peak demand, defer capital investments and increase
providers (NSPs) capacity diversity in supply sources
EG may also lead to network congestion or the need to change protection
schemes or settings, leading to additional network augmentation
Network EG could improve network reliability and backup supply reliability to

reliability individual customers, but may also reduce reliability through brownouts
and low-voltage trips
Voltage High levels of EG on networks can cause voltage issues and NSPs must
regulation actively manage it to meet prevailing standards
EG could provide voltage support at the ends of long rural feeders
Power quality High levels of EG may cause power quality issues, which may result in
network losses and appliance and equipment damage
Distribution Network charges collected by the service provider

use of system
(DUOS) and Current (kWh-based) charges result in a cross-subsidy from non-EG
transmission customers to EG customers
use of system
(TUOS)
Network losses More localised EG may reduce overall losses on the network by reducing
peak demand and the transmission and distribution distance of
transported electricity
Safety Costs for NSPs to ensure a safe electricity network
Network customers Installation Cost of system, installation work, network augmentation charges and
connection charges
17 These impacts are further explored from a technical perspective in Section 3.2.

Bill reduction Reduced bill from onsite generation and added incentives for exporting
and incentives energy, i.e. feed-in tariffs
Generators and Fuel costs Renewable EG can reduce fuel hedging costs incurred by centralised
retailers fossil-fuel-based generation
Wholesale EG reduces grid-transported demand, depressing wholesale prices and

prices hence gentailers’ profitability
Hedging costs/ EG can reduce network peaks and flatten load profiles, leading to lower
arbitrage hedging costs and potential market arbitrage
Society Environmental Reduced CO2 emissions from greater penetration of renewable energy and
generation efficiencies
Public health Improved public health and less reliance on fossil fuel imports through
and security greater renewables penetration
This report focuses on the impacts of EG on NSPs and their customers. The following sections
provide additional details on these impacts.
5.2.1 Network capacity
New investments in electricity network infrastructure are required to maintain network

performance over time in the face of peak demand growth. Networks are incentivised to
meet and manage demand at least cost.
EG can act as both a benefit and cost with respect to network capacity. For instance, high
penetration of EG on the network may lead to additional network augmentation. As the level
of EG on certain sections of network increases, the network may become congested (assets
exceeding their nominal rating). An example of the additional costs created by high
penetrations of EG was detailed in Section 4.4.1. However, EG output can also help defer
network expenditure, if it is available at the right time and in sufficient amounts. For
example, AusNet Services deferred $2.9 m of a planned transformer upgrade by contracting a
local, downstream, 10-MW gas-fired power station operator to operate during times of
network constraint18,19.
18 Grattan Institute, Sunrise, Sundown, May 2015.

19 A simple cost-benefit analysis of this implementation is provided in Section 3.4.9.2.

Distribution network service providers (DNSPs) should be in a position to promote the use of
EG in ways that result in the efficient use of the network – supporting them to meet their
network performance standards at the lowest cost. To do this, DNSPs must be able to provide
clear price signals to consumers that reflect the cost or benefits of EG on the network at a
given time and in a given location. The introduction of more cost-reflective network pricing
through recent changes to network pricing rules20 should enable progress towards this (see
sections 4.4.1 and 5.2.4). DNSPs must also be properly incentivised to seek non-traditional,
lower-cost solutions to network constraints.
Several measures in the regulatory framework oblige and incentivise NSPs to adopt new
technologies in the provision of network services, where this represents the most efficient
alternative to traditional network expenditure. This includes the:
 ability of the AER to review and revise forecast expenditure on the basis of the
efficient cost, taking into account the availability of new technologies
 ability of the AER to conduct an ex-post capital expenditure review and remove assets
from the regulatory asset base (RAB) in instances of overspending
 incentives inherent in the Efficiency Benefit Sharing Scheme and the Capital
Expenditure Sharing Scheme, which allow the NSPs to retain a portion of the benefit
derived from using new technologies to reduce costs while still providing the required
service.
Two recent key regulatory changes also incentivise DNSPs to explore the potential of EG to
resolve network constraints, compared with traditional network augmentation:
 the Regulatory Investment Test for Distribution (RIT-D)
 the Demand Management and Embedded Generation Connection Incentive Scheme
(DMEGCIS).
These are discussed further in the following sections.
5.2.1.1 Regulatory Investment Test for Distribution
Under the RIT-D process, DNSPs may consider non-network solutions to meet increasing
demand on their network. For example, DNSPs could potentially use EG to reduce feeder
loads, and hence defer upgrades to the distribution network.
The RIT-D process requires DNSPs to assess the costs — and, where appropriate, the benefits
— of each credible investment option to address a specific network problem. They then need
to identify which option maximises net market benefits (or minimises costs where the
investment is required to meet reliability standards). The RIT-D applies to investments
exceeding $5 million, and does not include investments to meet network refurbishment
requirements.
20Australian Energy Market Commission, National Electricity Amendment (Distribution Network Pricing
Arrangements) Rule 2014 No. 9.

The RIT-D rules have only recently been implemented21, and very few RIT-Ds have actually
been released to the market for competitive tendering.
MHC’s own industry consultation on this topic as part of the Clean Energy Council (CEC)’s
Future Proofing the Electricity Distribution Industry project found anecdotal evidence that
the RIT-D was creating greater focus on the potential opportunities available via non-network
solutions (incorporating EG)22. However, the consistency of standards and approaches to the
RIT-D process across DNSPs varies; hence, this provides an opportunity to develop best-
practice standards for the industry.
5.2.1.2 The Demand Management and Embedded Generation Connection Incentive

Scheme
Chapter 6 of the National Electricity Rules (NER) provides the Australian Energy Regulator
(AER) with the discretion to establish the DMEGCIS, which consists of an allowance for DNSPs
to trial innovative demand management solutions, and a mechanism to allow DNSPs to
recover foregone revenue resulting from the demand management implementation.
The stated purpose of the DMEGCIS is to:
 provide incentives for DNSPs to implement efficient non-network alternatives, or
 manage the expected demand for standard control services in some other way, or
 efficiently connect EG.
The Australian Energy Market Commission (AEMC) initiated a consultation process23 in
February 2015 to address two rule change requests. One was from the Total Environment
Centre, and aimed to simplify the DMEGCIS process. The other, from the Council of Australian
Governments, sought to strengthen the incentives for NSPs to undertake demand-
management initiatives24.
The AEMC have responded with a draft rule determination25, which proposes the DMEGCIS is
replaced with a:
 Demand Management Incentive Scheme (DMIS), which is
o of the AER’s own design
o based on a set of defined principles aimed at providing distribution businesses
with an incentive to undertake efficient expenditure on relevant non-network
options relating to demand management
 Demand Management Innovation Allowance (DMIA), which is
21 The Distribution Network Planning and Expansion Framework came into effect on 1 January 2013
22 At the time of writing, this report had not yet been published by the CEC.
23Australian Energy Market Commission, CONSULTATION PAPER National Electricity Amendment (Demand
Management Incentive Scheme) Rule 2015.
24 The two rule change requests have since been consolidated into one single request.
25 AEMC, National Electricity Amendment (Demand management incentive scheme), 28 May 2015.

o of the AER’s own design
o based on a set of defined principles aimed at providing distribution businesses
with funding for research and development in demand-management projects
that have the potential to reduce long-term network costs.
The draft rule proposes that AER should have until 1 December 2016 to develop and publish
the DMIS and DMIA.
In the United Kingdom, the Office of Gas and Electricity Markets has recently introduced26 a
new regulatory framework, known as RIIO (Regulation = Incentives + Innovation + Outputs). It
includes an Innovation Stimulus Package, which explicitly encourages innovation that could
result in better environmental and financial outcomes for the energy sector. The package has
three main components:
 Network Innovation Allowance27 — an annual allowance targeted at small-scale
innovation or to fund the preparation of submissions to the Network Innovation
Competition
 Network Innovation Competition28 — targeted at large-scale projects, open to non-RIIO
network licensees and awarded through an annual competition, with a maximum
lifetime funding per project of £81 m
 Innovation Roll-out Mechanism29 — aimed at supporting the deployment of fully
developed technologies that provide environmental and customer benefits, but would
not otherwise proceed, because the network
o cannot fund the proposed roll-out using their existing price control allowance
o cannot use an alternative price control mechanism to fund the roll-out
o does not have a financial incentive to implement the roll-out.
Stakeholder submissions to the AEMC’s review of the DMEGCIS indicated that the relative size
of the innovation allowance (a maximum of $1 m per year per DNSP) may have contributed to
its lower uptake30. The final rule for the DMIA states that the level of the allowance ‘should
be reasonable, considering the long-term benefit to retail customers’ 31. There may be
opportunity expand this allowance, in line with international initiatives.
26 The RIIO applies to the revenue collection period commencing 1 April 2015.
27 OFGEM, Electricity Network Innovation Allowance Governance Document, April 2015.
28 OFGEM, Electricity Network Innovation Competition Governance Document, July 2015.
29OFGEM, Assessment of benefits from the rollout of proven innovations through the Innovation Roll-out
Mechanism (IRM), April 2015.
30AEMC, Rule Determination National Electricity Amendment (Demand Management Incentive Scheme), August
2015, p 69.
31 Ibid, p 73.

Key Findings: Within the Demand Management Incentive Scheme (DMIS), Demand
Management Innovation Allowance (DMIA), RIT-D and RIT-T processes networks have
opportunities to consider and implement non-network solutions including EG to resolve
network constraints in the most cost effective manner, however there are areas for
improvement and potential for change as part of current regulatory reform initiatives.
5.2.2 Network reliability

Australian NSPs are required to meet specific reliability standards. They are incentivised to
maintain and improve network reliability through the Service Target Incentive Performance
Scheme and Guaranteed Service Levels under the DNSP’s licence agreements.
Network reliability is commonly measured by the:
 System Average Interruption Frequency Index (SAIFI)
 System Average Interruption Duration Index (SAIDI)
 Customer Average Interruption Duration Index (CAIDI)
 Momentary Average Interruption Frequency Index (MAIFI).
SAIDI, SAIFI and CAIDI relate to faults with a duration exceeding one minute, while MAIFI
relates to faults with a duration of less than one minute.
The key input in determining the benefit of investments aimed at improving network
reliability is the value of customer reliability (VCR). This is also used as an input under the
Service Target Incentive Performance Scheme to calculate incentive rates. VCR is defined as
the value different types of customers (residential, business or direct connection) place on
reliable electricity supply, or conversely, the perceived incurred cost of a power outage in
$/kWh terms.
Several studies have been conducted in Australia to determine the VCR, initiated with a study
by Monash University in Victoria in 199732. Most recently, the Australian Energy Market
Operator (AEMO) updated the VCR numbers in November 2014 following a national survey33.
Networks must ensure the most appropriate VCR is applied in the assessment of any network
augmentation or replacement project. Locational VCRs are therefore essential in allowing
network spend to align with market needs, and ensuring that network assets provide returns
to network operators and avoid investment in underused assets.
High levels of EG on the network can reduce reliability through brownouts and low-voltage
(LV) trips34. However, installing EG at strategic points on the network could improve network
reliability by providing voltage and frequency support, as well as minimise the duration of
32Monash University, The Value of Lost Load: A Study for the Victorian Power Exchange, Centre for Electrical
Power Engineering, 1997
33 Australian Energy Market Operator, Value of Customer Reliability Review – Final Report, November 2014
34 Please see section 3.2.5.1 for more details regarding the technical aspects of this impact

unplanned and planned outages by allowing the switching of loads to adjacent network
assets. This type of network support requires controllable EG (e.g. storage, diesel
generators); hence, intermittent renewable generation is not suitable.
DNSPs could also use storage and other generation sources (e.g. solar PV, diesel generators)
to form microgrids at strategic points on the network. This would allow sections of the
network to island in the event of a fault, improving reliability for both the DNSP and its
customers. However, this is currently just a concept, and not an existing or near-term
solution.
Storage and other generation sources at customers’ premises can also be used as
uninterruptible power supplies (UPS) or standalone power systems in the event of a network
outage. This would improve reliability for such customers, even if network reliability figures
themselves are not improved. Fringe-of-grid customers that experience poor network
reliability may value such backup supplies35.
5.2.3 Voltage regulation and power quality

The NER sets out performance standards to which NSPs must adhere when operating and
managing their network. In relation to voltage regulation and power quality, the key
categories include36:
 power frequency voltage
 the NSP must keep supply voltage within a determined range
 voltage unbalance
o the NSP must ensure that the average voltage unbalance measured at a
connection point does not vary by more than 0.5—3.0% (depending on the
voltage and averaging period)
 voltage fluctuations
o the NSP must maintain voltage fluctuation (flicker) levels in accordance with
applicable standards (AS 2279.4:1991)
 voltage harmonic distortion
o the NSP must keep harmonic distortion below levels determined in AS/NZS
61000.3.6:2001.
High levels of solar PV penetration can result in voltage rise. DNSPs are required to operate
their networks so that observed voltages do not exceed their allowable voltage range (230 V,
+10%/—6%) according to AS 6003837. Additionally, according to AS 4777, solar PV inverters
must shut down within two seconds of voltage or frequency moving outside of tolerance. If
voltage moves outside the range, the inverter will trip, resulting in ‘solar spillage’. There
35We note that AS4777 inverters are required to have anti-islanding protection, which would prevent the use of
any connected EG in the event of a network outage.
36 For more technical details regarding this impact, see Section 3.2.5.
37 Allowable voltage range in Queensland and Western Australia is 240 V (+6%/—6%)

have also been reports of solar PV inverters’ upper voltage limit having been set too high,
resulting in the inverter not tripping and the HV current damaging household equipment.
Voltage fluctuation refers to momentary variations in voltage that can cause household
lighting to flicker. This can be caused by intermittent supply from a power source (e.g. cloud
cover varying the output from a solar PV unit), as well as by high EG penetration on a section
of the network.
Harmonics distortion occurs when the current sinusoidal wavelength moves away from the
ideal range of 50—60 Hz. It can result from nonlinear loads, such as air conditioners or
personal computers, or from certain network equipment. It can also be caused by high levels
of EG on a particular section of the network, because inverter switches may inject additional
harmonics into the network. Harmonic distortion can contribute to network losses and
damage household equipment and appliances.
High penetration of solar PV on a feeder (in excess of 40%, according to recent reports38)
increases the risk of voltage issues. Furthermore, voltage issues are more likely during periods
of light loading, and at locations further away from the substation, such as rural networks.
DNSPs generally control network voltage via volt-volt-amps reactive (VAR) control devices,
such as switched capacitor banks and tap-changing transformers. These devices rely on
localised measurement and tend to have simple control objectives (e.g. to maintain local
voltage within a pre-determined range), and operate independently of similar devices
deployed at other points in the network. This is a relatively static approach, and is not well
suited to the increasingly dynamic network state resulting from increasing levels of EG.
As EG increases on the network, additional investments are required to manage emerging
voltage and power quality issues. Increasing levels of EG will mean more costly and complex
solutions to network issues. Forecasting this expenditure will therefore become increasingly
difficult.
5.2.4 Network charges
Australian NSPs recoup the costs associated with operating the network from their customers
through the distribution use of system (DUOS) and transmission use of system (TUOS) charge.
38 Energex, Power Quality Strategic Plan 2015-20, September 2014.

These are generally structured as a consumption-based component ($/kWh) and a fixed
charge ($/day).
These tariff structures were designed for a network with one-directional power flow, where
all customers consume energy in a similar manner. When customers install solar PV, they
reduce their grid consumption. However, the solar PV generation profile is not well aligned to
offset network peaks, as illustrated by Figure 54.
Source: Energex
Figure 54 – Solar photovoltaic impact on feeder load profile39
This means that while the NSP’s revenue base is reduced, their overall cost structure and
revenue targets remain unchanged. To recoup the cost of operating the network, the NSP will
hence increase the (fixed and variable) network charge. As a result, customers who have not
invested in solar PV — including customers that are excluded from the market, due to
affordability or lack of appropriate installation space — pay more for their electricity because
other customers have invested in EG40.
In the survey undertaken for this report, DNSPs highlighted this cross-subsidisation as the key
impact of EG. Several studies have sought to understand the size of this cross-subsidy, and
the effect on electricity bills for both EG and non-EG owners under different tariff structures.
39 Energex, Distribution Annual Planning Report 2014/15 to 2018/19, September 2014.

40 A similar issue exists for air conditioners.

As part of the national business case for smart grid technologies developed for Smart Grid,
Smart City, the forecasted effect on electricity bills was modelled under an inclining block
tariff and a cost-reflective tariff. The modelling found that by 2034, an annual cross-subsidy
of $420 would occur from non-solar PV to solar PV customers, while under cost-reflective
tariffs it would amount to $47.41
In 2014, the Energy Networks Association (ENA) commissioned a report to further model the
overall community cost of different tariff structures. The modelling used an inclining block
tariff as the base case and compared it to a declining block tariff, seasonal time-of-use and a
maximum demand tariff. Similar to Smart Grid, Smart City, the modelling was forecasted
until 2034, and found that energy-based tariffs (inclining block, declining block and seasonal
time-of-use tariffs) would result in an annual cross-subsidy of $500—700, while the maximum
demand tariff would deliver only a marginal cross-subsidy.42
Most recently, the Grattan Institute analysed the costs and benefits of EG (particularly solar
PV) under different tariffs, technology developments and subsidies. Like others, they found
that a significant cross-subsidy exists, and that it would continue to grow under a business-as-
usual scenario. The aggregate net cross-subsidy from consumers to prosumers would be
$14 billion by 2030, $3.7 billion of which would be due to the current structure of network
tariffs.43
Recent reforms44 to the NER have introduced cost-reflective network pricing in Australia.
Under the new rules, Australian NSPs will be required to design their network charges in a
way that reflects the true cost of providing network services to customers with different
consumption patterns. These tariffs are not specific to the technology choices of customers
(i.e. are neutral). Draft proposals are to be submitted to the AER in late 2015 for new
network prices that will take effect no later than 2017.
The implementation of new cost-reflective tariffs is not straightforward. Issues include:
 the reliance on interval or smart meter technology
o this is particularly the case for demand-based tariffs, which are based on
customer maximum demand in a period
 the impact of demand-based tariffs on customers with existing solar PV systems
o this typically contributes to lowering a customer’s overall usage, but not
necessarily their maximum demand
 overcoming customer resistance to change
o this includes their ability to understand and accept a new pricing regime, even
when it may mean they are better off, or at least not worse off45.
41 Ausgrid, Smart Grid, Smart City: Shaping Australia’s Energy Future, July 2014.
42 Energeia, Network Pricing and Enabling Metering Analysis, November 2014.
43 Grattan Institute, Sundown, Sunrise – How Australia can Finally get Solar Power Right, May 2015.
44Australian Energy Market Commission, National Electricity Amendment (Distribution Network Pricing
Arrangements) Rule 2014 No. 9.
45 CSIRO, Australian Consumers’ Likely Response to Cost-Reflective Electricity Pricing, June 2015.

The new network pricing rules require DNSPs to consult with customers and retailers while
developing their new pricing approaches. However, ongoing effort will be required to manage
implementation of the new tariffs and address existing cross-subsidy impacts to give
customers the best possible outcome.
Key Finding: Current volume-based network pricing structures includes a cross-subsidy from
non-EG owners to EG owners. The recent changes to the National Electricity Rules require
different patterns of consumption. These tariffs are cost reflective and are not specific to
the technology choices of customers (i.e. are ‘technology neutral’). There is opportunity
within these new rules to support the efficient uptake and use of EG that reduces future
costs to customers to below the level they might reach under volume-based pricing
structures.
5.2.5 Network losses

Line losses occur throughout the transmission and distribution network. In Australia, annual
losses across networks between the generator and the customer are equal to roughly 10% of
the electricity sent out46. Losses relate to the energy that heats up the network components.
They are a factor of how close the asset is operating to its thermal capacity, and not strictly
due to the distance between generation and consumption.
Network losses must be accounted for to ensure that enough electricity is generated to meet
demand. The NER requires DNSPs to determine distribution loss factors for all levels of their
network, which must then be approved by the AER. AEMO is tasked with ensuring that these
are properly factored into spot prices.
Embedded generation (DG) installed at strategic points of the network or at customer
premises is closer to the point of load. It can help reduce peak demand, which may reduce
network losses.
The value of reduced network losses associated with EG should first be calculated. If this is
significant, it could be leveraged for emerging commercial opportunities for NSPs arising from
EG. This is discussed further in Section 5.4.
5.2.6 Network safety

NER chapters 5 and 5A stipulate that installations and connections must follow the relevant
jurisdictional legislation as it relates to technical and safety requirements. These vary
depending on the jurisdiction, but generally include:
 service and installation rules
46 For more technical details regarding this impact, see Section 3.2.2.4.

 electrical safety rules
 relevant network standards and connection policies.
The relevant safety and technical requirements must be included in the NSP’s model standing
offer for network connection. These generally spell out the rules to which the EG installer
must adhere when facilitating the connection. However, it is ultimately the NSP’s
responsibility to ensure the safety of the customers connected to their network, and they
must to routinely inspect and audit installation work47. In the survey undertaken for this
report, DNSPs highlighted safety as one of the key considerations when facilitating
connections to the network48.
Australian NSPs are further required to meet specific reliability and quality of standards and
are incentivised to maintain and improve network performance over time. High levels of EG
on the network may impact fault level, asset loading and network stability and may hence
lead to safety and reliability issues.
5.3 Valuation framework
The above assessment highlights that the commercial impacts from EG may be significant. To
enable the design of appropriate pricing and incentive mechanisms to promote efficient
uptake of EG, it is becoming increasingly important to understand its true impacts on the
network.
The following sections provides an overview of key frameworks to value these impacts, put
forward both internationally and in Australia, which have been used as the foundation for the
commercial framework proposed in this report.
5.3.1 International context

To understand the value of EG and prepare utilities for the changing market state, several
valuation frameworks have been developed internationally over the last decade. Some of
these are described in more detail below.
47AS 5577, released in 2013, provides nationally consistent requirements for the development of an Electricity
Network Safety Management System and is an important first step towards national harmonisation of energy
technical and safety regulation across Australia.
48 For more details regarding the technical implications of this impact, see Section 3.2.3.

The main drivers for the development of these frameworks have generally been to allow for
integration of high levels of EG on the network, and accurate pricing of incentive
mechanisms, demand-response (DR) initiatives and electricity charges.
5.3.1.1 Advanced Energy Economy Institute
In a report49 prepared for the Advanced Energy Economy Institute, Synapse Energy Economics
put forward a framework to capture the costs and benefits attributed to EG. Table 25 shows
the cost and benefit categories identified in the report.
Table 25 – Advanced Energy Economy Institute embedded generation cost and benefit categories (Synapse Energy
Economics)
Be ne fits Costs
Ca te gory Exa mple s Ca te gory Exa mple s
Avoided energy generation Program marketing,

Load Reduction and Program administration
and line losses, price, administration, evaluation,
Avoided Energy Costs costs
suppression incentives to customers
Avoided transmission,
Integration capital costs,
Demand reduction and distribution and generation
Utility system costs increased ancillary service
avoided capacity costs capacity costs, price
costs
suppression
Avoided renewable energy

Transactional platform
Avoided compliance costs compliance costs, avoided DSP costs
costs
power plant retrofits
Impa c ts
Regulation, reserves
on a ll Ancillary services
energy imbalance
c ustome rs
Reduced financial and

Utility operations accounting costs, lower
customer service costs
Reduction in market power,

Market efficiency market animation, customer
empowerment
Project risk, portfolio risk

Risk
and resiliency
Contribution to measure
Participant non- energy Health and safety, comfort,
Participant direct costs cost, transaction costs,
benefits tax credits
O&M costs
P a rtic ipa n
t Impa c ts Increased heating or
Participant resource Water, sewer, and other cooling costs, value of lost
Other participant impacts
benefits fuel savings services, decreased
comfort
Economic development,
Public benefits Public costs Tax credits
reduced tax burden
S oc ie ta l
Impa c ts
Avoided air emissions and
Emissions and other
Environmental benefits reduced impacts on other Environmental costs
environmental impacts
natural resources
49 Synapse Energy Economics, Benefit-Cost Analysis for Distributed Energy Resources, September 2014

The framework accounts for costs and benefits attributable to the impacts of EG for three
main categories of stakeholders: utility customers, participants in distributed energy
resources (DER) programs (essentially EG owners and operators), and society as a whole.
The report puts forward the following key approaches to assess the value of the impacts of
different EG technologies on the three stakeholder groups:
 Direct monetisation
o the preferred option, because it captures the costs and benefits directly
attributable to EG
 Proxy valuation
o the second best alternative, which involves applying a multiplier to the impact
(e.g. avoided energy or capital)
 Alternative screening benchmark
o a simplistic method that uses cost—benefit ratios for the EG program or
implementation as a whole to determine its cost effectiveness
 Regulatory judgement
o allows the regulator to determine if an implementation is cost-effective, based
on the best available data (where the impacts cannot be monetised)
 Multi-attribute decision analysis
o a method used to internally rank impacts using a decision matrix summarising
the available data, and assigning a weighting to each impact based on its
importance.
5.3.1.2 Princeton Roundtable
The Princeton Roundtable brings together economic and financial academia and professionals
to discuss and develop new investment ideas and approaches for several areas and industries.
The annual expert energy policy roundtable has been held since 2013. Part of the scope of
work is to develop a new valuation framework for EG.
A report50 published in preparation for the roundtable puts forward a proposed valuation
approach based on two core principles. The valuation model must:
 consider both energy and capacity impacts from EG
 capture all costs and benefits attributable to EG.
With these high-level assumptions made, the report suggests the following approach to value
DER:
Net Impact = E + C – Co + Be + Ext
where
50Travis Bradford and Anne Hoskins, Valuing Distributed Energy: Economic and Regulatory Challenges - Working
paper for Princeton Roundtable, April 2013.

E = savings from offsetting wholesale energy purchases
C = savings from generation capacity investments
Co = costs associated with EG
Be = benefits associated with EG
Ext = environmental externalities (e.g. environment, job creation, security)
The first two categories are relatively self-explanatory, while the next two contain several
subcategories:
 costs associated with EG
o cross-subsidisation between EG and non-EG customers under an energy-based
charge
o administrative costs attributable to the NSP (e.g. for interconnection, changed
billing process, system planning, integration, protection)
o firming expenses incurred by the NSP to maintain network reliability as EG
levels increase
o asset lifetime and performance costs incurred by the NSP to maintain
acceptable voltage levels and power quality as EG levels increase
 benefits associated with EG
o transmission NSP and DNSP capacity investment offset — EG aligned with
network peaks may defer otherwise required network augmentation
o line losses and congestion — EG technologies that provide load reduction may
reduce network line losses
o merit order effect — reductions in load reduce energy and capacity clearing
prices
o fuel price hedge — relevant to renewable technologies only (fossil-fuel-based
generation depends on a underlying fuel price volatility, which is borne by the
customer).
The fifth category, environmental externalities, includes impacts that are not easily
quantified but that may still materially affect society as a whole.
5.3.1.3 The Electric Power Research Institute
The Electric Power Research Institute (EPRI) published a report51 in 2015 presenting an
integrated approach for valuing EG on the network. The report focuses on the development of
a universally applicable valuation framework, rather than a specific methodology that may
only be applicable to a certain region or network type.
The EPRI framework uses a best-practice economic and engineering analytical tool to
determine the extent of the impacts of EG. Similar to the frameworks described above, it
provides an impact assessment across the entire energy supply.
51 The Electric Power Research Institute, The Integrated Grid – a Benefit-Cost Framework, February 2015.

The impacts as they relate to the distribution and transmission network and large-scale
generation are shown in Table 26.
Table 26 – Electric Power Research Institute distributed energy resources impact assessment
Impact Benefit Cost
Distribution Loss reduction 
Capacity upgrade deferral 
Reconductoring 
Line regulators/static synchronous compensators


(STATCOMs)
Relaying/protection 
Load tap changers accelerated wear 
Voltage upgrade 
Smart inverters  
Operation and maintenance 
Bulk power system Generation mix changes  

(generation and
transmission) Deferral of transmission upgrades 
Transmission losses 
Operation and maintenance  
Fuel savings 
Congestion 
System operation/uncertainty 
The framework also provides the impact assessment on the customer (which was determined
to be only the EG investment) and society as a whole (e.g. environment, health, security).
EPRI’s proposed framework highlights that geographical characteristics (e.g. voltage,
topology) of the network must be taken into account. As a result, it also suggests that the
impact analysis should be undertaken down to the feeder level.
As part of this analysis, EPRI introduces the concept of hosting capacity. This refers to
determining the amount of EG that can be allowed on a certain feeder without adversely
affecting supply quality and existing infrastructure. The analysis essentially involves
incrementally adding EG to the feeder until a threshold is breached (defined as when services
such as reliability or quality of supply fall below required standards). At this point, a
mitigation strategy is developed and costed to restore the feeder to the required service

level. Additional EG is then added until the next threshold is reached, which in theory would
create a cost-curve for addressing increasing levels of EG on the network.
5.3.2 Draft Clean Energy Council valuation framework

Ernst and Young recently drafted a report52 for the CEC as part of its Future Proofing the
Electricity Distribution Industry program. The report reviews national and international
valuation frameworks and puts forward a recommended, consolidated approach to value EG
on the network.
The proposed framework has been designed to be scalable and applicable to any EG
technology. As such, it has attempted to capture all value streams available from EG and put
a value on the each specific cost and benefit to assess the total net present value (NPV) from
EG on the network.
A high-level overview of the material value drivers and valuation approach put forward in the
CEC report is provided in Table 27.
Table 27 – Clean Energy Council value drivers of embedded generation (EG) and valuation approach
Commercial impact Description Valuation approach
Network capacity EG could reduce peak load on a feeder, A simplified version of the current
upgrades deferring otherwise required network Regulatory Investment Test for
upgrades Distribution (RIT-D) approach with a
half-hourly resolution
High levels of EG penetration may also lead
to network congestion and additional
network augmentation
Network support Altering the generating profile of network A simplified dispatch model using short-
support generators or deferring generation run marginal cost with half-hourly
investments using distributed EG granularity to capture seasonal trends
Parameters to consider include fuel
costs, reliability statistics,
maintenance, ramping capabilities and
demand profiles
Voltage regulation High penetration of EG on the network can Steady-state powerflow analysis to
lead to both over-voltage (during periods quantify the magnitude of the required
of low load and high EG output) and under- reactive or voltage regulation support
voltage (during periods of high load but low
EG output) Potentially in-depth transient and fault
analysis in networks where more
DNSPs must actively manage the voltage on sophisticated voltage management
the network to reduce this impact schemes may be required
Power quality issues Harmonics (inverter switches injecting Transient analysis for different
unwanted harmonics into the network) operational states and switching
behaviour
Flicker (variations in voltage, e.g. from
momentary cloud cover) that change Unbalanced three-phase analysis
lighting levels
52Clean Energy Council, Evaluation Methodology of the Value of Small Scale Embedded Generation and Storage to
Networks, April 2015

Commercial impact Description Valuation approach
Network reliability Network service providers (NSPs) are Statistical analysis (Monte Carlo) to
incentivised to maintain and improve estimate the expected unserved energy
network reliability and number of outages with and
without EG
High levels of distributed energy resources
(DER) penetration may reduce network
reliability, but also improve it and offer
other network support benefits
Reassessment of fault Legacy protection schemes are currently Nothing proposed at this point
level coordination, established (e.g. fault current detection
islanding and protection and protection relays)
schemes
Emerging capabilities (e.g. fault ride-
through and islanding) may change the
value proposition
Dynamic control Distribution NSPs can actively control EG to Proposed to be assessed on a case-by-
manage local network conditions case basis
Expected to be particularly valuable on
rural networks with EG growth, where
network augmentation can be deferred
The CEC’s proposed approach entails calculating the value of each impact on a feeder-by-
feeder basis to reach an NPV for the EG in question. This could also be done a feeder category
basis, using CSIRO’s recent work53 (e.g. urban 11 kV).
5.3.3 Proposed consolidated impact categorisation and valuation

approach
A best-practice approach is starting to emerge from the international and Australian valuation
frameworks assessed above. The approach:
 takes geographical and network specific conditions into consideration
 is applicable to all different types of EG
 is applicable to all stakeholders
o customers, EG owners, NSPs, generators and regulators.
Limiting factors that explain why certain impacts may not be appropriate to include in the
valuation framework include the following:
 The impact cannot be monetised (e.g. environmental, which in the absence of a
carbon pricing mechanism, it is challenging to monetise).
 The impact is not material or not available under the current market state.
Table 28 presents our consolidated list of commercial EG impacts relating to NSPs and their
customers, and the rationale for including an impact from the final proposed valuation
framework.
53 CSIRO, National Feeder Taxonomy, June 2013

Table 28 – Marchment Hill Consulting embedded generation (EG) valuation framework
Included in
Description Rationale
framework?
One of the key potential impacts of EG is the ability to defer

Network capacity  capital investments
Certain types of EG can be used for network support and outage

Network reliability  management
Issues are already emerging on Australian networks, and network

Voltage regulation  service providers must make investments to limit the impact
Similar to voltage regulation

Power quality 
Distribution use of Estimated EG costs and benefits will form an input when
system (DUOS)
 determining DUOS
A split incentive exists to reduce network losses, which could be

considered in a future framework
Network losses  EG can reduce marginal annual peak demand losses, freeing up
additional capacity; it has been included as an impact
consideration under ‘network capacity’ above
Following the review of the available frameworks and inputs from the stakeholder survey and
interview process, we have identified the following key categories:
 network capacity upgrades
o the impact of EG on peak demand (including peak demand line loss reduction)
and associated capital investments
 network reliability
o the impact of EG on the reliability and safety of the electricity network, and
reliability of supply to customers to cover network outages (backup or UPS
capability)
 power quality and voltage regulation
o the impact of EG on the quality and safety of supply
 future considerations
o impacts that may be monetised in the future resulting from technological or
regulatory advances.
The suggested framework design and valuation approaches are described below54.
54Although a brief description of the technical issues for each of these categories is necessarily included, further
details can be found in the technical impacts sections of this report.

5.3.3.1 Network capacity upgrades
Network upgrades are often limited to one type of upgrade at a time (e.g. adding a
transformer or line reconductoring). The process for determining the capital costs and market
benefits of these upgrades is well established in Australia55.
The process for determining whether EG is a viable alternative to network augmentation has
been established under the RIT-D process. It uses a scenario-based approach that requires the
NSP to assess credible options (including DER) to offset the otherwise required investment.
When a network understands and accurately calculates the impact of DER solutions, the RIT-D
is seen as an appropriate approach for larger investments given the risks and capital
contribution involved.
However, RIT-D is only applicable to investments exceeding $5 million, which does not
include the largest volume of projects (e.g. feeder upgrades and smaller substation projects).
Furthermore, the RIT-D involves significant amounts of administration, time and effort on the
part of both NSPs and EG proponents. Consequently, for smaller network projects in which EG
(or equivalent DR technologies) can deliver superior market and financial benefits, a
simplified version may be more appropriate. In such an approach, the valuation and
calculation methodologies are the same as under RIT-D, but the NSP may deal directly with
proponents in line with their internal contracting and governance arrangements.
5.3.3.2 Network reliability
Outage statistics on a feeder-level basis are captured by DNSPs as part of their regulatory
investment notice, which can be used to determine network reliability with and without EG.
This can be done through statistical analysis, either on a deterministic basis with all input
parameters constant, or a probabilistic basis using Monte Carlo analysis to generate faults at
random.
Further analysis could then determine the circumstances under which EG could improve
network reliability (i.e. SAIDI, SAIFI and MAIFI savings). Such circumstances could include:
 applying fault events on different sections of the network to determine the types of
faults for which EG can reduce the number of customers affected
o e.g. by allowing a section of the network to island or providing voltage support
 avoiding faults by using EG to ensure that a network asset does not exceed its thermal
rating
 avoiding outage time (SAIDI) after the event of a fault by allowing the NSP to
undertake switching action
o e.g. where EG is used to reduce load on a feeder adjacent to the faulted
feeder to facilitate its additional load.
55 This is further explored in Section 3.4.9.2.

As mentioned in Section 5.2.2, VCR captures the perceived market benefits of avoiding a
power outage to a customer. As such, it could be used to determine the potential reliability
value available from EG. However, the value of reliability to NSPs (and as a result, to their
customers) is currently captured through the Service Target Incentive Performance Scheme,
which makes it a more appropriate approach to determine the reliability impact of EG.
However, this scheme does not currently incorporate the full value of customer benefits of
outage reduction and mitigation.
For EG to improve network reliability, additional remote or automatic control capabilities
may be required, as well as possible changes to network topology (e.g. increased
interconnection).
5.3.3.3 Voltage regulation and power quality
The key consideration for the valuation approach for voltage regulation and power quality is
determining the potential deferral or addition of network support technologies for different
network types under varying EG penetration levels.
Most NSPs can undertake this analysis (e.g. through steady-state powerflow, transient and
fault analysis), although their ability to perform this analysis on the LV network is more
limited.
5.3.3.4 Safety
It is ultimately the NSP’s responsibility to ensure the safety of the customers connected to
their network, as well as staff and contractors working on the network.
It is challenging to put a value on maintaining a safe network. Safety has been included as a
category largely to highlight that it is the number one priority of NSPs, and that it comes at a
cost.
At a minimum, the costs directly attributed to maintaining network safety when connecting
and operating EG would need to be considered. Further work is needed to understand the
magnitude of these and any additional costs incurred by the network relating to EG and
network safety.

5.3.3.5 Future considerations
The items detailed below are impacts that may be monetised in the future, given
technological or regulatory advances.
Fault level coordination, islanding and protection schemes

Several legacy protection schemes are currently established, but it is not well known how
appropriate they are in the presence of higher levels of EG on the network56. Legacy
protection schemes and equipment include:
• protection relays or reclosers
o detect fault level currents on the network
o order circuit breakers to open in the event of a fault
• anti-islanding protocols
o switch off EG in the event of a fault to protect from continuing to feed the
fault and damaging equipment
o are meant to de-energise the network section so that field staff can work on it,
but cannot (and should not) be relied on alone to ensure the safety of field
staff.
Recent technology advances may require certain elements of these legacy schemes to be
reassessed. Advanced inverters are able to support fault ride-through (i.e. additional
tolerance for short-term deviations in frequency and voltage). This means that they may
remain connected through a contingency and offer grid-stabilising services, such as injecting
reactive power.
Fault ride-through and (intentional) islanding capabilities under higher penetration of EG may
change the value proposition, allowing EG to support the network and improve reliability.
However, this is still reasonably far from reality.
Dynamic control
Advanced inverters could allow NSPs to remotely control the output from EG. This means that
output can be limited during periods of high EG output to limit the stress on the network. It
may be particularly valuable on rural networks with EG growth, where network augmentation
can be deferred. The addition of storage could reduce peaks and flatten the network load
profile, and hence also potentially defer augmentation investments.
Network support
Several network support generator sets are distributed across the network, some of which are
owned and operated by NSPs. Similar to the capacity benefits described above, distributed EG
can alter the generating profile of network support generators, or defer generation
investments or operational costs.
56 This is discussed in detail from a technical perspective in Section 3.2.3.

5.3.3.6 Framework design
This section provides an overview of the proposed framework design to value the impact of
EG. It includes preparatory steps and an overall approach to determining the net impact of a
particular EG. Each individual impact will require its own specific dataset and calculation
approach.
The valuation process will follow three main steps:
i) gathering the relevant data and inputs
ii) determining the network classification
iii) calculating the net impact.
Gathering data and inputs is relatively straightforward, and is largely a function of the
availability of data. We assume that the majority of inputs would be provided by the NSPs.
The following key network classifications should be considered:
 feeder type
o based on reliability categorisation (central business district, urban, short rural
or long rural)
o will affect the level of redundancy and interconnectivity of the network
 geographical location
o affects the generation profile of EG, particularly with solar PV, due to solar
radiation and length of day
 customer mix
o the load profile between residential and different type of commercial
customers will affect the network load profile and hence the relative impact of
EG
 network type
o the loading and relative strength of the network will determine the extent of
the impact of EG, as well as other equipment in the network that helps
mitigate EG impacts (e.g. voltage regulators, automatic tap-changing
transformers, static synchronous compensators)
 additional potential considerations
o legacy issues (e.g. network design and topology)
o health of the network
o specific network assets and existing network at the connection point (e.g.
single or multiphase, single-wire earth return).
Once the network classification has been determined, the costs and benefits of each impact
category would be added together to calculate the net impact of the relevant EG.
An overview of the framework design is provided in Figure 55.

Impacts Valuation Net Impacts
Capacity
NSPs Change in
Net Capital and NSP
Reliability O&M Change Costs
Net
Impacts
Customers
PQ and Voltage Direct
Regulation - Reliability Customer
- Investment Benefits
and O&M
Safety
Figure 55 – Marchment Hill Consulting framework designed to value the net impacts of a specific type of embedded
generation. NSP = network service provider; O&M = operation and maintenance; PQ = power quality
The net benefits to the network would be considered, as would the direct customer net
benefits (e.g. the required EG investment and any reliability and cost improvements).
Key Finding: Currently, there is no universally applicable and agreed regulatory model
to value the impact of EG on networks. Networks require an appropriate valuation
approach if they are to identify the costs and benefits of EG and to signal the efficient
sizing, location and operation of EG. In addition, there is a gap in the current
understanding of the impact of EG on network reliability, safety and quality of supply
and an appropriate valuation approach is also needed to allow NSPs to identify these
impacts.

5.4 Commercial opportunities
As highlighted in the previous section, EG can disrupt electricity networks. However, it also
gives NSPs the opportunity to make use of new technology, benefiting both themselves and
their customers.
This section explores commercial opportunities for NSPs arising from EG, as well as the
related strategic benefits and barriers to EG implementation. It is not intended to be a
comprehensive review of the merits of these commercial opportunities. Rather, it provides a
high-level overview and qualitative assessments to highlight opportunities for further work.
In this section, we focus on the following technologies:
• residential and commercial solar PV
• residential storage
• electric vehicles
• network storage
• microgrids
• demand response.
5.4.1 Strategic benefits and barriers

The strategic benefits of each of the technologies listed above include the ability to:
• improve network utilisation57
o e.g. flattening the load profile and increasing grid-supplied consumption
o this reduces the cost of the network for customers and improves the long-term
viability of network assets
• incentivise customers to remain network customers
o e.g. offering cost-effective services (either grid-connected or off-grid solutions)
o this maintains the network’s revenue base
• provide network support
o e.g. power quality, voltage and outage management
o this reduces the cost of running the network
• facilitate the introduction of cost-reflective network pricing
o e.g. offering products and services that could managing the impact of time-of-
use pricing for EG owners
o this promotes improved use of the network
• use an existing core competency to offer a service in a competitive market
o e.g. maintenance field services, asset management, network operations
o this enable DNSPs to use their existing skills and experience to increase
unregulated revenue.
57Network utilisation is defined as the average electricity flow through the grid over a period of time (e.g.
one year) as a percentage of total capacity.

The commercial opportunities described below focus particularly on options that meet one or
more of these strategic benefits. Opportunities where these benefits are not available will be
provided as well, but in more general terms.
Barriers facing NSPs when pursuing commercial opportunities from EG include:
• competition
o the market is mature, or a future market is expected to include a number of
interested parties
• technology maturity
o the technology solution is not mature enough to realise all available benefits,
or NSPs lack experience offering and operating the technology
• standards and regulatory
o lack of standards or regulatory guidance about the technology and related
commercial opportunities
• data and customer access
o the solution requires access to customer data
o this could require additional customer consent or additional expense
o the processes and costs are still emerging from market reforms.
The relationship between the barriers, strategic benefits and the relevant commercial
opportunities of EG is provided in Table 29.
Table 29 – Embedded generation strategic benefits and barriers identified by Marchment Hill Consulting
Strategic Benefits Barriers & Issues

Maintaining Cost- Data Access
Commercial Opportunities Network Network Leveraging Technology Standards &
Customer reflective Competition & Customer
Utilisation Support Capabilities Maturity Regulatory
Base Pricing Acceptance
Solar PV
Excess Solar PV Strategy      
Residential Energy Storage
With Solar PV       
Without Solar PV        
Network Storage
Network Support     
Storage-as-a-Service        
Electric Vehicles
Installation and Maintenance  
Rollout Strategies  
Strategic Deployment  
Controlled Charging        
Microgrids
Fringe-of-grid / Off-grid    
Embedded. / Urban     
Demand Response
Expanding Current Capabilities   
DR Trading Platform     
DR = demand response; PV = photovoltaic
Table 29 reveals that the greatest strategic benefits of EG are generally accompanied by
barriers (e.g. network storage). The strategic benefits and specific barriers to each
commercial opportunity are discussed in the following sections.

5.4.2 Solar photovoltaics
The Australian solar PV market is relatively mature. As such, it offers limited opportunities
for NSP market entry. However, NSPs can support their potential strategic objectives by
offering products and services relating to the use of excess solar PV generation.
This is a promising commercial opportunity, because solar PV uptake is presumed to remain
strong. Many network customers have installed oversized systems in relation to their in-house
consumption, having been incentivised to do so through previously generous feed-in-tariffs
(FiTs). NSPs could implement a conventional direct load control solution, similar to the water
heating and pool pump schemes widely implemented in Queensland.
Suitable technologies include:
• heating and cooling (air conditioning)58
• water heating
• pool pumps
• electrical energy storage.
To pursue this opportunity, NSPs would need to target customers with solar PV installations,
particularly those with oversized installations. This could be done directly, or via established
retail channels such as electricity retailers or solar PV providers. Customers would need to be
further incentivised to allow the NSP to access their data and to manage the EG and the
demand-management technology installed at the premises. NSPs could implement this as part
of cost-reflective network tariffs, since the use of such technologies would benefit solar PV
customers under these new tariff arrangements.
The service could be marketed directly to existing solar PV customers. However, given the
generous FiTs that many of these customers receive, they would need to be significantly
incentivised to limit their exports during the day by using the excess generation to supply
loads earlier, rather than those loads contributing to the evening network peak. A more
realistic approach would be offer a bundled product (solar PV, storage unit, tariffs) to new
customers. This would require the DNSP to work actively with the multiple parties, including
technology providers, retailers and other service providers. Introduction of well-designed
network and retail tariffs is fundamental to promoting desirable customer behaviour and
protecting the commercial interests of NSPs.
Customers are generally likely to be reluctant to let NSPs turn on customer loads earlier in
the day than they would otherwise choose. However, well-designed tariffs will encourage
customers to respond more in accordance with the aim of the NSP’s direct load control.
Further opportunities may arise from restructuring premium FiTs to allow customers to divert
this subsidy towards a discount on the cost of installing a storage solution.
58 This is explored from a standards perspective in Section 3.5.2.

5.4.2.1 Grid-optimised smart solar
The Rocky Mountain Institute recently published a report analysing the economics of grid-
defection and potential pathways and implications for the United States electricity markets59.
The report highlights that increasing EG on the network means new opportunities for utilities,
and puts forward some key options such as optimising new and existing solar PV on the
network.
The majority of installed solar PV has a ‘dumb’ inverter: i.e. an inverter that cannot provide
services such as voltage management and fault ride-through. These capabilities would benefit
NSPs by enabling more active management of the distribution network. The Rocky Mountain
Institute suggests that NSPs may incentivise the uptake of smarter inverters by reducing the
connection charge or expediting the connection process.
5.4.2.2 Strategic benefits
Making use of excess solar PV generation represents a key strategic opportunity for NSPs. It
would allow NSPs to:
• ease the overall impact of solar PV on the network by limiting exports
• improve network utilisation by increasing load during the day and freeing up network
capacity during peak times through load shifting
• ease the introduction of cost-reflective pricing by offering a solution to limit the bill
impact on EG owners
• incentivise network customers to remain customers by offering them additional
benefits
• use existing capabilities such as direct load control to provide customers with an
additional service.
5.4.2.3 Barriers and issues
Standards
AS/NZ 4755 specifies the framework for DR capabilities and supporting technologies for
electrical products such as air conditioners, swimming pool pump units and electric water
heaters. The current version of this standard does not support turning on air conditioners by
bringing the load forward (e.g. for pre-cooling).
Data and customer access

It is not yet entirely clear under which arrangements NSPs may have full access to customer
data under the new metering rules. Under the proposed rules, customer data would be the
responsibility of the meter data provider. Home area networks-enabled smart meters could
59The Rocky Mountain Institute, The Economics of Grid Defection – How Grid-Connected Solar-Plus-Battery
Systems Will Compete With Traditional Electric Service, Why it Matters, and Possible Paths Forward, April 2015.

be used for this purpose. However, this is currently not included in the proposed minimum
functionality requirements for smart meters. If implemented as a direct load control solution,
NSPs will face a further barrier by having to visit every eligible site to reprogram the demand
response enabling device (DRED).
Customers also lack options when engaging different service providers for different portions
of their load. This would have to be negotiated via the customer’s retailer. The possibility of
customers changing to an alternative retailer adds further risk to the approach. AEMO
recently submitted a rule change request60 to the AEMC to initiate the process of establishing
multiple trading relationships, which would enable different entities to provide different
services at a single customer site.
The new metering reforms in the National Electricity Market (NEM) could provide
opportunities for multiple parties to contract via the metering coordinator to gain access to
meter data and functionality, and thereby provide services to customers. This could include
using the meter’s home area networks function to control devices in the home.
The metering rule also allows DNSPs to take on the roles of metering coordinator, metering
provider and metering data provider. This would allow the DNSP to address the majority of
challenges that may be raised by the rules. This would, however, require the DNSP to acquire
appropriate accreditaion from AEMO, as well as adhere to any ring-fencing requirements.61
If the DNSP is not acting as the metering coordinator, it would be able to negotiate a
commercial arrangement with the relevant metering coordinator for meter access. An
alternative arrangement could also be made for the DNSP to partially fund, in whole or in
part, the retailer’s meter roll-out. In return, they would gain access to the metering services.
5.4.3 Residential storage

Residential storage is an emerging distributed energy service that represents both a technical
and commercial opportunity for DNSPs62. Commercial opportunities include marketing
residential storage solutions to:
• existing or new solar PV owners
o this could include installation and operation and maintenance (O&M) services
• non-PV customers
o this would take advantage of the difference between peak and off-peak prices.
NSPs may be able to market residential storage solutions directly to customers, similar to the
product New Zealand NSP Vector launched in 2013. This involved a leasing arrangement for a
60 Australian Energy Market Operator, Rule Change Request – Multiple Trading Relationships, December 2014.
61Australian Energy Market Commission, Draft Rule Determination National Electricity Amendment – Expanding
Competition in Metering and Related Services, March 2015.
62 Further information on the potential to leverage energy storage to mitigate the technical network impacts of EG
is provided in Section 3.3.2.2.

combined solar PV and battery storage solution.63 The program resulted in the installation of
around 300 storage solutions across Auckland, but has now closed.64
Residential storage products could be structured either as a leasing arrangement or outright
sale. Leasing is likely to be most attractive, allowing customers to avoid the substantial
upfront cost. This would also require a strategic partnership with one or more technology
providers. Storage services would predominantly be marketed to existing solar PV customers,
or offered as a bundled solar PV and storage product to new customers.
As part of this offer, NSPs may also offer installation and O&M services to customers. It could
also be linked to network cost-reflective prices to help manage customer bills.
DNSPs may also have an opportunity to obtain alternative energy seller status (as a separate,
ring-fenced trading entity) for selling bundled solar PV and storage services to customers.
Under such an arrangement, the DNSP would own, manage and control the PV and storage
unit installed at the customer’s premises, and sell energy generated by the solar PV to the
customer as an exempt retailer. The storage unit would enable the customer to maximise the
use of their generated power, and limit feedback into the grid (a common cause of technical
issues). This added control of the storage unit would be especially attractive for DNSPs,
because it would allow them to use it for other purposes such as network support. If such
units were aggregated by, or for, the DNSP, it would potentially also allow NSPs to defer
otherwise required network augmentation by discharging during times of network peak.
This combination of benefits, in which customers can provide support to networks in addition
to meeting their own needs, will ultimately determine the cost effectiveness of distributed
residential storage.
Strategic benefits for NSPs from implementing residential storage with controlled charging
arrangements are relatively significant. It would allow the NSP to:
 use the storage for network support, e.g. for power quality and voltage management
 ease the introduction of cost-reflective pricing by offering a solution to limit the bills
of EG owners
 use existing capabilities and asset management to provide an additional service to
customers.
Additionally, if the storage is not bundled with solar PV, but is simply creating value for
customers by enabling greater use of off-peak rates, network utilisation could be improved by
increasing load during low load periods and freeing up network capacity during peak times
through load shifting. Network customers could also be incentivised to remain so by offering
them additional benefits.
63 Vector, Shareholder Review 2013, August 2013.

64 Vector, 2014 Annual General Meeting, October 2014.

Despite all of the attention being offered to residential storage and the number of solutions
on offer in the market, its widespread uptake faces several barriers.
Competition
The main barrier for NSPs who wish to market storage solutions to customers is the likely
presence of significant competition. The tier-1 retailers have been exploring opportunities in
offering residential storage solutions, and are likely to be followed by specialised retailers
and service providers.
Technology cost
Significant efforts are being made globally to reduce the manufacturing cost of energy
storage technologies, and cost-effective solutions were generally expected to become
available in the next 5—10 years. However, Tesla’s launch of its Powerwall in April 2015
changed market expectations. Retailing at a starting price of US$3,500 for a 10-kWh system
and US$3,000 for a 7-kWh version, Powerwall represents a step change in the price of
residential storage and brings it one step closer to financial viability.
Regulatory barriers
To obtain alternative energy seller status, DNSPs need to apply for a retail exemption from
the AER. As the National Electricity Customer Framework does not currently apply in all
states (most notably Victoria and Queensland), DNSPs would have to apply for separate
exemptions in those states under jurisdictional regulations. This creates a barrier to service
providers competing across state lines.
Additionally, it is currently unclear whether chapters 5 and 5A of the NER sufficiently address
the connection of storage solutions; there is uncertainty as to whether grid-connected energy
storage batteries fall under the definition of ‘generating plant’. This has implications for
whether the requirements for a connection application and connection offer need to be
fulfilled for storage solutions. Chapters 5 and 5A should hence be reviewed and amended (if
applicable) to support the connection of energy storage.
The AEMC and CSIRO are currently undertaking a joint project aimed at understanding the
regulatory impacts of storage solutions in the NEM. The final report, expected to be published
in September 2015, should provide an informed view of any changes to the NER that will be
required to facilitate storage technologies.65
65Available at: <http://www.aemc.gov.au/News-Center/What-s-New/Announcements/AEMC-launches-project-on-

integration-of-electricit>.

The required standards for installation and operation of storage technologies also need to be
developed. The CEC has recently issued draft installation guidelines for storage66 that could
be used as a starting point to develop nationally consistent rules and guidelines.
Additionally, legacy premium FiTs are another potential barrier. Customers on these
arrangements will have little incentive to invest in technologies that limit their electricity
exports. These subsidies will stay in place across Australia for significant amount of time67 and
will be a sizeable expense for governments. As an example, a recent report by the
Queensland Competition Authority forecasted that premium FiT payments made to
Queensland prosumers until 2028 would equal $2.9 billion68.
5.4.4 Electric vehicles

Electric vehicles (EVs) have been available on the international market for some time, but
uptake in Australia has been slow. This is partly due to prohibitive costs, insufficient
infrastructure and customer concerns such as limited driving range.
EVs represent a range of opportunities to NSPs. They will be a potential new source of load on
the network as well as provide a new market in which NSPs can offer additional products and
services.
The EV industry will require the expertise of the incumbent electricity industry players to
safely and effectively develop, manage and maintain EV connection and charging
infrastructure. The development of strategies to uncover these commercial opportunities will
require focusing on customer solutions and empowerment as well as commercial
relationships.
In addition to providing negotiated services relating to new connections (e.g. in car parks or
business precincts) to cater to high levels of EV adoption, several strategic options are
available to grow network business. These include:
 partnering with EV charging infrastructure providers for future roll-out plans, asset
management and maintenance
 working with governments or EV manufacturers to develop and manage public charging
infrastructure
 developing relationships with private companies (such as supermarket and fast food
chains) for the roll-out of privately owned charging infrastructure
 partnering with EV sellers to promote connection and supply packages with innovative
tariff structures and demand-management options to capture EV owners’ charging
needs
66 The Clean Energy Council, Grid-connected Energy Systems with Battery Storage – Draft in Progress, May 2015.
67 In Queensland and South Australia until 2028, and in Victoria until 2024.
68Queensland Competition Authority, Estimating a Fair and Reasonable Solar Feed-in Tariff for Queensland, March
2013.

 making use of vehicle-to-grid (V2G) arrangements in which the storage capacity of the
EV provides network support.
All of the above strategies will require the NSP to operate in a proactive manner and engage
strongly with customers and industry.
Greater EV uptake will present NSPs with several strategic benefits. EVs:
 increase electricity usage
o hence help drive prices down, provided they are charged at the right time
 rely on customers being connected to the grid
o EG/storage sources are unlikely to be sufficient to service residential charging
demands
 make use of existing NSP capabilities
o e.g. engineering and technical expertise, asset management.
EVs therefore present a relatively unique EG value proposition for NSPs, because they
increase grid consumption and promote grid dependence.
Regulatory and policy settings

Apart from the forecasted slow uptake of EVs, they present few barriers to NSPs engaging the
market. Regulatory arrangements would need to be developed to allow for remote-controlled
charging, as well as for appropriate cost recovery from charging infrastructure.
Supporting policies for EVs are currently lacking in Australia. These include direct funding for
charging infrastructure, stringent emission standards and tax rebates. Such policies and
incentives have driven increased uptake of EVs in other markets (e.g. California, Norway).
Recent analysis by Energeia for the Energy Supply Association of Australia suggests that an
additional 620,000 EVs could be added to Australia’s vehicle fleet over 2023—2035 via
cost-effective policy interventions69.
Customer acceptance
As for residential storage, data access and customer acceptance will be required to allow
NSPs to access controlled EV charging benefits.
Customers must also be incentivised (through a properly structured tariff or direct payment)
to shift charging away from peak times themselves, or allow the NSPs to do so. This is critical.
Otherwise, NSPs will have a new peak demand problem to manage with costly network
augmentations: such as with the past boom in air-conditioner takeup, which customers are
now reluctantly paying for through necessary tariff increases. The public, regulators and
69 Energia, Review of Alternative Fuel Vehicle Policy Targets and Settings for Australia, July 2015.

governments will be reluctant to fund augmentations for peak demand that will be poorly
used and economically inefficient.
5.4.5 Network storage

Network embedded storage essentially refers to DNSPs installing storage on strategic points of
their network to offer storage services to their customers and capturing additional value
streams70. Such strategic points could include at or near the zone substation, or near
distribution transformers scattered around subdivisions.
Storage services could include:
 wholesale market arbitrage
o where storage is used to purchase electricity at cheap rates and sell back to
the grid at high price periods
 network support
o e.g. voltage management, peak shaving, islanding, outage management
 customer storage-as-a-service
o e.g. enabling customers to store excess energy generated during the day (from
their solar PV system) and use it at peak evening tariff times.
Network storage technologies are being extensively trialled and installed internationally.
California in particular has establishing itself as the leading energy storage market, supported
by the 1.3 GW of capacity mandated to be installed by 2020. Australian activity in network
storage has been more limited, with AusNet Services’ trial of a 1-MW Samsung lithium-ion
battery on its network being one of the more significant examples.
In selling storage services, the DNSP could sell part of the storage capacity to retailers or
demand aggregators, who in turn sell it on to customers. The storage used for network
support could be part of the RAB, while a separate ring-fenced business could be set up for
storage sold to end customers or demand aggregators: similar to how DNSPs currently sell
pole space to telecommunication companies.
Opportunities for DNSPs in network storage include the following:
• DNSPs hold the potential real estate and know-how to facilitate network storage
installations.
• Network storage could benefit from economies of scale.
• Network storage could be more practical for network purposes, because it involves
fewer control points and subsequent integration issues.
One counter aspect is that the closer the storage is located to the loads, the more the
network will benefit from it. For example, storage located at the zone substation will not
benefit the downstream feeder(s), distribution transformers or LV network in terms of
70 Further information is provided in Section 3.3.2.1.

levelling the load on them, or reducing their required capacity or backup capacity from
adjacent sources.
Establishing storage-as-a-service would require certain foundational pricing arrangements,
including virtual net metering. This allows EG owners to assign exports that otherwise would
have fetched the grid FiT to other customers in the local area. Virtual refers to the metering
arrangements, since electricity is not physically transferred between the sites, with
transactions settled based on metered data. Networks would be ideally placed to develop
these pricing arrangements as part of a storage and service product.
Depending on the approach of the NSPs, network storage could offer the following strategic
benefits:
• Customers can use the grid to access storage capabilities, encouraging them to retain
their relationship with their NSP.
• Network storage can provide network support services, such as power quality and
voltage management.
• Storage capacity can provide peak support, flattening the network load profile and
freeing up capacity.
• Offering a solution that will limit the impact of cost-reflective pricing on EG owners
may make such pricing arrangements less contentious.
• The installation, maintenance and operation of network storage can make use of
existing NSP competencies such as technical and engineering capabilities and asset
management.
Barriers to network storage are described below.
Technology cost and maturity

As mentioned previously, storage technologies are currently generally too expensive to be
financially viable at the residential scale, except where one installation can provide multiple
benefits. However, recent installations by Australian DNSPs indicate that network storage
could become a viable investment under certain circumstances.
Technology maturity is another current barrier. The majority of storage technologies have not
been adequately tested to determine whether they can perform reliably for network
purposes. Australian DNSPs are also in the early stages of trialling storage technologies, and
are still familiarising themselves with its operational risks and capabilities.
The additional revenue streams described above under the ‘storage-as-a-service’ option make
network storage an attractive option. However, this option would require the establishment

of virtual net-metering services. As described above, these arrangements would require
multiparty strategies, and they have not yet been sufficiently trialled in Australia71.
Regulatory barriers
As noted above, the deployment of storage solutions may face regulatory issues. However,
AEMC and CSIRO’s joint project is expected to provide recommendations on any amendments
required to the regulatory framework to facilitate storage technologies.72
There are also some regulatory restrictions on the ability of DNSPs to buy and sell electricity.
AusNet Services has established a fit-for-purpose energy purchase agreement with
EnergyAustralia. This is likely to be the most straightforward solution for similar future
deployments, at least in the near term.
Ring-fencing arrangements would also need to be developed to facilitate a portion of the
storage unit to form part of the NSP’s RAB, and a separate portion to be used for commercial
purposes.
5.4.6 Microgrids
Although several microgrids have been established in Australia, they are largely an emerging
opportunity for most NSPs. The most viable option in the near term is likely to be at fringe-of-
grid locations. Embedded microgrids installed at strategic points of the network, or for
greenfield developments, may become attractive at a later stage.
Embedded microgrids, which are defined in this context as microgrids established in more
urban areas of the network, hold a slightly different value proposition to fringe-of-grid
microgrids. Because urban networks are designed with a higher degree of redundancy,
network reliability is less of a driver to form a microgrid.
Urban microgrids represent an opportunity for DNSPs to offer premium reliability to targeted
customers (e.g. data centres, hospitals, emergency services). Embedded microgrids could also
be an option for greenfield developments, where they could establish a secondary revenue
stream for the DNSP through O&M management fees. The embedded microgrid could also be
established as a trading entity, which provides market access and subsequent arbitrage
opportunities.
Fringe-of-grid microgrids involve DNSPs offering microgrid solutions where the cost to serve,
on a per-customer basis, is greatest (i.e. in remote and rural communities). To ensure that
customers across the state pays the same amount for electricity, state governments currently
subsidise rural and remote customers through a payment to the local DNSP. This is called a
71MHC notes the trial of local network charges and virtual net metering currently being undertaken by the
Institute for Sustainable Futures and ARENA. The project involves trials in five different locations and is an
important step towards better understanding of the impact and necessary technology and regulatory arrangements
required to facilitate such a solution.
72Available at: <http://www.aemc.gov.au/News-Center/What-s-New/Announcements/AEMC-launches-project-on-
integration-of-electricit>.

community service obligation (CSO) payment. The annual CSO payments made to DNSPs are
significant; for example, Ergon Energy received close to $600 million in CSO payments from
the Queensland government in 2013, up 20% from 201273.
Instead of maintaining connections to fringe-of-grid communities, DNSPs with significant
amounts of long rural feeders (e.g. Ergon, which has 65,000 km of SWER lines) could use CSO
payments to install DG (e.g. diesel, solar PV) and storage to establish off-grid microgrids in
these rural communities.
The advantages this could bring to DNSPs include:
 enabling the use of energy storage where it is most economical
 establishing more sustainable electricity services to remote communities
 reducing maintenance and fault management costs
 improving reliability of supply to customers where network reliability is poor.
In the near term, the most attractive option for NSPs seems to be installing battery storage at
strategic points on SWER lines to provide peak support and islanding capabilities, rather than
establishing multi-EG microgrids. Ergon are currently trialling 20 batteries for these purposes
on its network, with a view to expand the program over time.74
As technology options mature, DNSPs could take entire (particularly costly to serve) sections
of the network completely off-grid, while maintaining ownership and control of the microgrid
to generate revenue.
The strategic benefits available to NSPs from edge-of-grid microgrids include:

 providing network support
o primarily voltage management, but also potentially increasing reliability
 reducing cost to serve
o as such, incentivising rural customers to remain grid-connected (microgrid or
main grid)
 using existing competencies to offer additional services
o e.g. technical and engineering capabilities, asset management.
Barriers to cost-effective microgrids include technology maturity level of energy storage and
control solutions, as well as regulations and standards required for safe operation in island
mode. These are described further below.
73 Ergon Energy, Annual Financial Statement, 2013

74Available at: <https://www.ergon.com.au/about-us/news-hub/media-releases/regions/general/battery-
technology-on-electricity-network-and-australian-first>.

Safety
A key concern of network operators is the potential implications of widespread, uncontrolled
islanding on their grid. Uncontrolled islanding could lead to a safety issue for field staff, who
may not know whether a line is energised; or a network stability issue, where significant
ad-hoc synchronisation occurs.
The safety and stability implications of microgrids could be managed by developing and
implementing appropriate procedures, management and staff training. This includes proper
planning and design of microgrids.
Technology cost and maturity

Since storage is an important component of microgrid solutions, their deployment faces
similar barriers to those of energy storage. Solar PV and generators (diesel or gas) are both
mature technologies, while bespoke microgrid control solutions are also available in the
market.
Regulations
The NER does not provide a regulatory mechanism for networks to consider the creation of
microgrids or standalone systems as an option for the most efficient approach to energy
supply. However, EG technology advancements may make such options the most cost-
effective in the future.
In addition, rather than providing a payment directly to the DNSP to serve remote locations,
the CSO regulations could be amended to allow DNSPs to pursue alternative options to service
the rural network. The regulatory framework could be reformed to ensure DNSPs can receive
dedicated funding, as long as they can prove that a microgrid is a more cost-effective option
than the CSO payments. However, reforming the CSO payments would involve a great deal of
political risk.
Standards
At the moment, there are no Australian industry standard protocols for a network operator to
control a connection to allow it to island. The norm (both in Australian and internationally)
has been to force EG, such as rooftop solar PV systems, to have some form of anti-islanding
protection installed — i.e. they are required to disconnect from the network when the
connected network loses power. This approach specifically prevents systems capable of
operating in island mode from islanding while still connected to unintended part(s) of the
(non-microgrid) network.
An international standard, IEEE guideline 1547.4, was approved in 2011 to specifically guide
the design, operation and integration of microgrid systems with electric power systems. It
includes current best practices for implementing the various ways in which microgrids can
separate from a part of the main grid, including planned (or intentional) islanding, and
reconnect while providing power to the islanded grid.

5.4.7 Demand response
Demand management, including DR, allows DNSPs to meet customer demand by shifting or
reducing demand through non-network alternatives, rather than increasing supply through
network augmentation75. Non-network alternatives can involve controlling certain customer
loads (e.g. pumps, cooling, heating), or dispatching EG to offset demand or obtain capacity
credits in the Wholesale Electricity Market76.
These services are generally performed by some DNSPs (for residential customers), as well as
by demand aggregators (for commercial and industrial customers). Demand aggregators
contract capacity across the network to access arbitrage opportunities and support network
augmentation deferral.
For NSPs, DR could provide the following commercial opportunities:
 expanding DR to increase network capacity and reliability
 providing a standardised commercial offer to potential DR proponents
 establishing and operating a trading platform for DR.
The first two points above represent what would be considered business as usual, and could
be fairly easily implemented if the NSP sees sufficient value.
The trading platform is an emerging opportunity that could be owned and operated by the
NSPs, allowing them to control the process and capture potential transaction fees. It would
essentially allow:
25. NSPs to post their specific DR requirements on the platform
o (e.g. xMW on a certain feeder)
26. DR proponents and demand aggregators to bid in to the market
o (e.g. x$/MW) where they have capacity available or contracted
27. swift and transparent market clearing.
The strategic benefits available to NSPs from DR include improved network utilisation and
capacity, and improved reliability through load shedding or support to enable switching in the
event of a fault.
The most significant barrier for the DR opportunities outlined above relates to the trading
platform, because it has not been trialled and tested in Australia. The required rules and
75Further information on the potential of demand management to mitigate network impacts is provided in
Section 3.3.4.2.
76In the Wholesale Electricity Market, which has a peak demand of around 4000 MW, more than 500 MW of DR is
already signed up by aggregators with commercial and industrial customers. It is an alternative to peaking
generation capacity for the Independent Market Operator’s capacity market reserve capacity mechanism.

regulations would need to be established, as well as the technical requirements and platform
design.
Valuation approaches to capture the capacity benefits of demand management are well
established. However, approaches to capture the reliability impact of demand-management
initiatives are less developed. The framework put forward in this report could be used as a
starting point, but further work would be required to formalise such an approach.
5.4.8 Key findings

The key findings relating to all of the above commercial opportunities are listed below.
Key finding: Ring-fencing requirements for NSPs currently differ by jurisdiction and
are generally seen as inadequate for the purposes of emerging markets and
technologies. In particular it is unclear how storage technologies are to be treated
under the current ring-fencing requirements and how NSPs might support the efficient
creation of microgrids or standalone systems as an option for the most efficient
approach to energy supply.
shown below.
Key finding: Several of the potential commercial opportunities for networks to offer
new products and services based on EG depend on the networks having cost-effective
access to customer data and/or a direct channel to the customer. (e.g. in home EG and
demand optimisation, residential storage). The proposed new metering rules present
for metering and related services but they also create risks associated with potential
new costs for data access and the loss of a traditional DNSP service.
shown below.

5.5 Implications for the network service provider business model

Under the current market structure, customers essentially make unilateral decisions regarding
the type, size and location of their EG; third parties undertake sales and installation; and the
NSP oversees the interconnection process to the network.
As EG technologies continue to decline in cost and different types of EG become prevalent in
the market, new value streams will emerge. The business model of the NSP will likely have to
evolve in turn.
This section discusses insights into how NSPs may need to change their business models and
practices to capture the benefits and manage the challenges of increased levels of EG on the
network.
5.5.1 Accenture — network business model evolution

Accenture recently published a report for the ENA exploring how NSPs’ business models may
adapt to a changing operating environment77.
The report puts forward six main business model alternatives that are available to NSPs:
1. Information Services
2. Intelligent Grid Operator
3. Beyond the Meter Services
4. Distributed Platform Integrator
5. Distributed Platform Integrator and Trader
6. Distributed Energy Production Services.
Each proposed business model is described briefly below.
Information Services
This involves the NSP investing in advanced metering infrastructure to provide their
customers with greater information about their energy usage. It would provide the foundation
for improving data availability to establish a truly intelligent network.
Intelligent Grid Operator
77Accenture, Network Business Model Evolution – An Investigation of the Current Trends on DNSP Business Model
Planning, January 2015.

This model would see the network operator invest in additional monitoring and
communications technology (in addition to advanced metering infrastructure) to control both
customer-sited EG and network equipment. This would enable the NSP to take further steps
towards intelligent, dynamic operation of the network.
Beyond-the-Meter Services
In this model, the NSP would offer additional services to the customer extending beyond the
metering point. These would include DR, home energy management systems, and installation
and maintenance of equipment. The rationale behind this business model is that NSPs are
relatively well positioned to provide these improved energy services to customers.
Distributed Platform Integrator
Similar to the better-known term of smart grids, the Distributed Platform Integrator business
model would involve NSPs optimising grid services by improving their monitoring, control and
automation capabilities. This involves real-time decision making, based on load and network
state.
Distributed Platform Integrator & Trader
Similar to a virtual net-metering solution, this approach would involve NSPs facilitating
transactions between distributed prosumers. It would allow customers to optimise their
energy usage based on prevailing pricing and availability of EG.
Distributed Energy Production Services
This would involve the NSP marketing EG solutions to customers through leasing or technology
partnership arrangements.
5.5.1.1 Strategic considerations
For all of the potential business models outlined above, Accenture suggests that a successful
NSP would have to develop the following strategic capabilities:
1. intelligent network operations
o invest in additional information and operational technology capabilities to
increase the level of information available and improve subsequent decision
making
2. industry and partnership management
o identify suitable partners to provide capabilities, products and services lacking
within the NSP and develop an appropriate regulatory strategy
3. market and commercial intelligence
o gain a better understanding of customer choice to accurately price products
and services, and improve market and competitor intelligence capabilities
4. performance management
o develop and maintain performance metrics to quantify the impact of business
decisions
5. customer interaction and marketing

o develop marketing, branding and communication capabilities to attract and
retain customers.
5.5.2 Key findings

Although not all of the above business models are directly related to the impact of EG on the
network, they do hold some common themes that are important for NSPs to consider in
developing a strategy to capture the benefits and tackle the challenges emerging from EG.
Key Finding: Active involvement in the EG and energy services markets on the part of
NSPs or their related entities as either a market participant or to simply incentivise
the efficient use of EG for the benefit of all customers, will require developing
partnerships with other stakeholders including technology service providers and
retailers.
shown below.
Key Finding: The NSP business model needs to evolve to facilitate the integration of
multiple distributed energy resources, including EG. This could provide greater network
capacity and energy diversity to optimise grid performance for both supply and
demand.
shown below.

6 Policy and regulatory options
Note on authorship
This section was written by Marchment Hill Consulting (MHC) with contributions from CSIRO
for the key technical findings. All policy and regulatory options in this section were developed
by MHC.
6.1 Introduction
This report has reviewed the technical, commercial and regulatory impacts of embedded
generation (EG) on the electricity network. Through this process, we have found that
increasing levels of EG on the network present both challenges and opportunities for NSPs.
However, given appropriate conditions and incentives, EG can deliver great benefits to NSPs
and their customers alike. The challenge for the industry is to ensure that the policy,
regulatory and commercial environment supports the efficient level of deployment of EG to
serve the long-term interests of consumers.
The sections below present the key technical, regulatory and commercial findings identified
in this report. It also gives policy and regulatory options to address these findings:
 A policy option relates to approaches that involve a change to existing government
policy, or a change to the general operational policies or approaches adopted by NSPs
as a whole.
 A regulatory option relates to potential approaches that require a change to existing
industry regulations in order to be implemented.
Note that the policy and regulatory options identified in this section reflect some available
options to pursue, rather than a recommended course of action, and many of them are not
ultimately recommended to be pursued further. The policy and regulatory options that are
carried forward for further consideration in support of the recommendations in Section 7 are
highlighted in blue.
6.2 Key technical findings

Key findings from the technical review are described in the following sections.
6.2.1 Technical constraints and network impacts
1 Increasing penetration of EG, including greater uptake of rooftop solar and other EG, is
changing load profiles on Australian electricity networks and making network load
prediction more challenging. Appropriate planning and management of network load
demand across different timeframes is essential to ensure both adequate power quality
and a reliable supply of electricity of the network.
Policy and regulatory options | Page 213

 High penetration of EG increases the gap between the maximum and minimum
network load. In addition to peak demand forecasts, accurate minimum demand and
energy consumption predictions are required. These will assist in network planning
relating to managing voltage and potential reverse power flow issues.
 Accurate prediction of network load demand is essential for effective utilisation and
management of EG sources on the network, and to establish sustainable load
management systems for the smart grid.
2 Reverse power flow has been observed at the medium-voltage and high-voltage network
levels. In radial networks designed for unidirectional power flow, this can affect the
operation of protection and voltage regulation devices. In islanded networks, such as
some remote communities, it can also have a detrimental effect on generator control and
stability.
3 EG has the potential to contribute current to any fault in the vicinity. Protection system
design and settings must be considered in rural and off-grid systems to ensure faults are
cleared quickly. In urban areas, the added fault level (the maximum current that can
flow in the network segment as a result of a fault) may restrict the amount of EG that
can be connected due to system design fault levels.
Policy and regulatory options Comment
Working closely with a range of industry This will help distribution network service
stakeholders (including EG proponents, the providers to perform a cost–benefit analysis
Australian Energy Market Operator and the on mitigation strategies, as well as provide an
Australian Energy Regulator) to develop an input to any future cost-reflective pricing
appropriate valuation framework that can be regimes relating to EG.
leveraged to develop pricing arrangements to
Efficient EG uptake refers to deployment and
promote efficient uptake of EG.
use of EG at times, in locations and at scales
that reflect the value of EG to all customers.
6.2.2 Options for mitigating network impacts and risks

4 Grid-scale storage has the potential to provide a range of network benefits relating to
EG, including voltage management, energy balancing, and improving network stability.
These benefits can increase the flexibility, reliability and efficiency of power delivery to
consumers.
 Currently, grid-scale storage is limited due to a combination of system cost and
market maturity, and as such is generally only deployed in specific applications and
locations as noted in Section 3.3.2.1. However, considering the market for energy
storage devices has been estimated to reach 3000 MW by 2030, as well as expected
improvements in related technologies, the deployment of grid-scale energy storage is
predicted to see increasing uptake in the coming years.
5 Distributed storage can provide numerous benefits to networks, including improved
management of voltage and power flows, peak load and generation management, and

reactive power support. Customer-owned storage has the capacity to provide these
benefits alongside direct benefits to the customer, despite not being network-owned.
 For customer-owned storage, appropriate dispatch management is key. Unlocking its
network benefits requires appropriate incentives to ensure that it is dispatched at
times that are beneficial to networks. If incentives are not well-designed, customers
may utilise storage in ways that are detrimental to the network. If direct control of
non-network storage is desired, the deployment of operating systems (e.g. through
demand response systems such as AS/NZS 4755 Part 3.5) and relevant communications
platforms will enable controlled dispatch.
Create appropriate pricing mechanisms to Storage holds significant potential for

incentivise the deployment of energy storage network support and should be implemented
both on the network and behind the meter wherever it is cost effective.
where cost-effective. This will help manage
power flow and voltage issues, as well as
make use of the capabilities of power
electronics.
6 Power electronics solutions, such as STATCOMs and smart inverters, have the capability
to mitigate power quality issues relating to EG, including managing voltage ramp-rates
and excursions. In some cases, they can also reduce harmonic content.
 Smart inverters in particular can use a pre-existing generation resource to self-
manage their own generation. As this functionality is not currently mandated, the
penetration of these devices in networks is low. Retrofitting existing fleets of
inverters would require a large capital investment.
 The current revision of AS/NZS 4777 part 2 includes some smart inverter functions to
self-manage voltage issues where possible. However, this functionality is not
currently mandatory within the standard, which may continue to limit market
penetration even once the updated standard is published.
Using the valuation framework to manage the Efficient deployment of technologies could be
voltage regulation, power quality, reliability promoted by determining the value of power
and safety impacts of EG on the network. This quality support on different parts of the
could, for example, support the strategic network.
deployment of medium-scale power
electronics solutions, such as STATCOMS, at
locations where they may be the least-cost
solution to help manage power quality issues.

Create incentives for customers to retrofit An EG valuation framework should be able to
existing distributed inverters (e.g. on rooftop determine the benefit of retrofitting existing
solar photovoltaics) to unlock STATCOM-like EG. This could be translated into an incentive
capabilities, such as managing local voltage for the customer to take action, possibly via
issues and harmonics. reduced connection charges or cost-reflective
network tariffs.
7 Improved prediction of renewable generation, informed through dedicated metering, can

improve EG integration by increasing the utilisation of assets such as storage and solar
photovoltaic (PV)-controlled air conditioning. It can also inform decisions on power
system unit commitment and network planning. These benefits can be broadened beyond
EG to medium and large-scale generation.
8 Improved network load prediction techniques that incorporate increased penetration
levels of EG are necessary to retain the current benefits of short, medium and long-term
prediction of net load in a high-penetration EG environment. The prediction techniques
also need to incorporate any other factors that may be contributing to changing
utilisation patterns of electricity networks.
Implement appropriate mechanisms that Those options need to explicitly consider

standardise the collection and exposure of pathways for guaranteeing consumer privacy,
household load and generation data. while still delivering the data necessary for
fine-grained residential models and
forecasts. Gross metering will better enable
the disaggregation of load and generation.
6.2.3 Options for maximising network benefits

9 Utilising EG for power and voltage profile levelling reduces the operation of expensive
peaking plants. It can also ease voltage management requirements, improve power
system reliability and provide frequency support.
10 Insecure, non-dispatchable energy source types of EG (predominantly PV and wind) alone
provide only limited power and voltage levelling and frequency support. Including storage
offsets the insecure, intermittent nature of PV and wind, allowing EG to provide these
services effectively.
11 EG can contribute to peak load reduction. This alleviates congestion, reduces line losses
and results in deferral of equipment capacity upgrades where capacity is constrained.
12 Evidence from the literature and distribution network service providers suggests that the
reduction in cumulative net load due to EG can extend the life of both substation and
distribution transformers.

Working closely with a range of industry While some EG is already bringing a net-
stakeholders (including EG proponents, the positive cost benefit, attention should be
Australian Energy Market Operator and the paid to technologies that are close to net-
Australian Energy Regulator) to develop an positive, such as energy storage, with the
appropriate valuation framework that can be expectation that their position may improve
leveraged to develop pricing arrangements to soon.
The framework should take into account the
critical network points where EG can provide
a direct economic benefit by reducing
operating costs and deferring network
investment.
6.2.4 Australian Standards gap analysis

13 Australian standards do not well cover the functionality and performance requirements of
protection relays used to prevent islanding and reverse power flow relating to embedded
generators. This reduces the consistency with which DNSPs can implement these devices,
with significant flow-on effects for the market and other industry stakeholders. These
effects include increased cost of development, implementation and commissioning;
increased cost and reduced consistency of type-testing; and potentially reduced
performance and reliability in-situ.
Supporting the development of a new Standardising the performance of protection

Australian Standard to manage the relays will reduce the risk of the application
functionality of protection relays for EG. of these devices by utilities. It will also
provide guidance to manufacturers about the
core and additional functions of these
systems.
14 There is a significant lack of standards relating to the integration and operation of

electric vehicles (EVs), including vehicle-to-grid-enabled EVs. This may increase the risks
and reduce the potential benefits associated with EVs by limiting the uptake of EV
demand response and vehicle-to-grid systems, as well as reducing consistency of
approach and increasing implementation costs.
Revisiting the development of standards such Current industry projections indicate that
as AS/NZS 4755 Part 3.4, or facilitating the significant EV uptake is unlikely for some
local acceptance of a similar international time. However, near-term development of
standard for managing EV charging and Australian standards or acceptance of
discharging. international standards will ensure that local
network operators will have tools ready to
manage any issues that arise if needed.

15 Grid-connected stationary energy storage systems utilising emerging technologies
including lithium-chemistry and zinc-bromine flow batteries, are not well standardised in
Australia particularly in the areas of safe installation, operation and disposal. This will
likely have an impact on consistency of approach between jurisdictions, with significant
implications for the market and industry; it may also increase the likelihood of incidents
relating to the improper use of these systems.
Supporting the development of standards Some standards relating to energy storage,

relating to the safety of small-scale and such as AS/NZS 5139, are beginning to be
large/grid-scale energy storage systems. developed. However, there is a significant
Focus should be directed to those addressing risk that such systems will be deployed
safe installation/operation and fire en masse before standards are in place to
protection. In addition, lithium chemistry ensure systems can be deployed and
may require particular attention, because its operated safely.
standards are immature.
6.3 Key regulatory and commercial findings

This report has reviewed the current regulatory framework relating to connecting EG to the
network; the commercial impacts of EG; the development of a commercial framework to
value these impacts; and the commercial opportunities available to network service providers
(NSPs) from EG.
The key findings of the review have been categorised below as either regulatory or
commercial.
6.3.1 Key regulatory findings

The below findings include impacts that require a regulatory focus in order to be addressed.
16 Although recent reforms to Chapter 5 and 5A of the National Energy Rules (NER) have
improved the connection process related to EG, the lack of a consistent national
Ensure the adoption of the National Energy Adoption of National Energy Customer
Rules chapters 5 and 5A (or equivalent) Framework in Victoria and Queensland would
reforms in Victoria, Queensland, Western have this effect, because Chapter 5A reforms
Australia and Northern Territory to introduce are included in the Retail Law. However,
a consolidated regulatory framework across alternate state-based derogations would also
Australia. suffice.
A uniform regulatory approach across
Australia would provide EG proponents with
greater certainty and ability to compete
nationally.

17 Where augmentation costs may occur because of high levels of small scale solar PV
penetration on a network (i.e. basic connections), the NER does not allow for these costs
to be recovered directly from the EG owners in a cost-reflective approach. Instead,
these costs are recovered from the entire customer base which introduces an element of
cross-subsidisation between customers.
Introduce a cost-reflective connection This would ensure that EG proponents pay

charge, reflecting the fact that connections for required augmentation. However, it
at certain sections of the network are more would charge the last customer, and as such
costly due to required augmentation work. would not be equitable and create a barrier
to future efficient uptake of EG. It is also
inconsistent with the new requirement for
networks to establish prices based on the
long-run marginal cost to the network.
Charge even small-scale EG owners directly Similar to the above point, this would create
for required network augmentation. a barrier for future efficient uptake.
Forecast expected annual network This may be the most straightforward option
investments required due to EG on the and would ensure EG-related costs are
network that can currently not be recovered recovered from EG customers only. An
from the relevant EG owner. appropriately designed EG valuation
framework (see below) would address these
Forecast annual uptake of EG and add the
impacts, as well as capture potential
expected required investment to the
benefits.
connection charge for EG customers. This
could also be expanded to connections for
heating and cooling devices that also affect
network expenditure.
Work closely with a range of industry A valuation framework would capture the
stakeholders, including EG Working closely augmentation cost as well as the benefits of
with a range of industry stakeholders EG to the network. The cost of augmentation
(including EG proponents, the Australian could then be reflected in the connection
Energy Market Operator and the Australian charge or other pricing mechanism to the
Energy Regulator) to develop an appropriate customer.
valuation framework that can be leveraged
to develop pricing arrangements to promote
efficient uptake of EG.

18 Within the Demand Management Incentive Scheme (DMIS), Demand Management
Innovation Allowance (DMIA), RIT-D and RIT-T processes networks have opportunities to
consider and implement non-network solutions including EG to resolve network
constraints in the most cost effective manner, however there are areas for improvement
and potential for change as part of current regulatory reform initiatives.
Working with the industry and regulators to The Australian Energy Market Commission
identify where there may still be a recently published a determination on the
disadvantage for network providers to utilise DMIS, which includes provisions that the
EG for network purposes compared with Australian Energy Regulator (AER), when
network upgrades. NSPs could then build on designing the future DMIS scheme, should
the recent reforms to the Demand take into consideration the full value stream
Management Incentive Scheme to ensure that from demand-management initiatives when
any disadvantages for NSPs using EG in this determining appropriate incentives.
way is addressed.
The determination makes no specific
mention of increasing access to capital
expenditure benefits, other than noting the
final design of the incentives will be decided
by the AER.
Develop a set of best-practice principles and This could be developed in conjunction with
a model process for planning network demand response, demand management and
responses to constraints that distribution EG technology proponents.
network service providers (DNSPs) could use
to improve the quality and consistency of
their approach.
Develop an incentive mechanism for DNSPs to Benchmarking may not necessarily reflect
use non-network solutions (which may the efficient deployment of EG. Care would
include EG) based on appropriately be needed to take relevant local network
benchmarked expenditure levels across the and consumer characteristics into account
DNSPs (and relevant international (e.g. availability of load transfer
comparators). opportunities).
A best-practice expectation could be set by It is unclear whether this would support a
the regulator for the use of non-network goal of driving the efficient deployment of
solutions, which the DNSPs could then be EG (i.e. at a time, location and scale that
incentivised to outperform. supports required network service delivery at
lowest cost).

19 Currently, there is no universally applicable and agreed regulatory model to value the
impact of EG on networks. Networks require an appropriate valuation approach if they
are to identify the costs and benefits of EG and to signal the efficient sizing, location and
operation of EG. In addition, there is a gap in the current understanding of the impact of
EG on network reliability, safety and quality of supply and an appropriate valuation
approach is also needed to allow NSPs to identify these impacts.
Working closely with a range of industry A valuation framework to determine the

stakeholders (including EG proponents, the impacts of different types of EG would allow
Australian Energy Market Operator and the customers to base investment decisions on
Australian Energy Regulator) to develop an the full value of the EG. The framework
appropriate valuation framework that can be should consider geography, customer mix,
leveraged to develop pricing arrangements to and network and feeder type.
20 Ring-fencing requirements for NSPs currently differ by jurisdiction and are generally seen
as inadequate for the purposes of emerging markets and technologies. In particular it is
unclear how storage technologies are to be treated under the current ring-fencing
requirements and how NSPs might support the efficient creation of microgrids78 or
standalone systems as an option for the most efficient approach to energy supply.
Supporting the removal of state-based rules Network service providers (NSPs) should be
barring or limiting network providers’ ability free to invest in EG when it is cost-efficient
to own and operate EG (for the purpose of to do so.
network support).
Supporting changes to current state-based It will be difficult to define when a network

ring-fencing rules to create a national would be best placed to provide the most
framework with the overarching principle to efficient solution, or who would decide this.
ensure no market participant is unnecessarily
constrained in a competitive environment
with regard to providing services to
customers.
78A microgrid is defined as ‘a group of interconnected loads and distributed energy resources within clearly
defined electrical boundaries that acts as a single controllable entity with respect to the grid and that connects
and disconnects from such grid to enable it to operate in both grid-connected or “island” mode.’ For the purposes
of this report, this definition is extended to also include permanent island networks.

Ensuring that they are free to enter into This would further enable NSPs to pursue
direct commercially based trading commercial opportunities relating to
relationships with customers and that NSP’s customer-sited EG.
related entities should not be restricted from
providing energy selling services to
customers, where they meet appropriate
ring-fencing requirements.
Supporting a review of relevant rules to This could be supported by reforms to

ensure no barriers exist to the creation, existing subsidies (e.g. community service
operation and maintenance of microgrids or obligation payments) where this opportunity
standalone systems by networks as an option proves to be a more efficient use of funds to
for the most efficient approach to energy serve remote customers.
supply.
6.3.2 Key commercial findings

The below findings include impacts that require a commercial focus in order to be addressed.
21 Current volume-based network pricing structures includes a cross-subsidy from non-EG
owners to EG owners. The recent changes to the National Electricity Rules require
different patterns of consumption. These tariffs are cost reflective and are not specific
to the technology choices of customers (i.e. are ‘technology neutral’). There is
opportunity within these new rules to support the efficient uptake and use of EG that
reduces future costs to customers to below the level they might reach under volume-
based pricing structures.
Introduce a cost-reflective demand charge A demand charge would limit the cross-
for all customers. subsidy and signal costs of future expansions
in network capacity.
Introduce a distribution use of system charge This would probably be more contentious
or negative demand charge for energy than a demand charge, because it would only
exported to the grid. target embedded generation (EG) owners.
Increase the connection charge or introduce A demand charge would reduce the cross-
an additional fixed charge for EG owners. subsidy and ensure that all customers receive
the same price signal, rather than
specifically targeting EG owners.

Introduce cost-reflective network charges for Reduced network charges could be applied to
EG output that take into account the relative virtual net metering to reflect the reduced
value of EG as a generation source (e.g. use of network infrastructure (and reduced
reduced line losses, network support). line losses), as well as the potential value for
network support (depending on the time and
location of the EG output).
This would support EG exports at times and
in locations of most value to an efficient
network, and support emerging business
models (e.g. virtual net metering,
community energy, storage-as-a-service) that
promote efficient, ongoing use of the
network.
Supporting a national approach to cost- The impact on consumers is a critical

reflective pricing that removes inconsistent consideration for the implementation of
jurisdictional obligations on tariff structures more cost-reflective network pricing.
and assignment.
Technology is available that can minimise
Supporting the implementation of a balanced this impact (e.g. automated demand-
framework for the uptake of smart meters management devices, adding storage to
for the fastest economic roll out to benefit solar, enabling network control of inverters)
all customers. and simplify pricing messages for consumers.
However, effective product bundling will
Encouraging adoption of technologies that
require coordinated stakeholder input.
support efficient pricing and customers’
ability to access data and manage their
demand (e.g. in-home displays, apps, data
portals).
Working closely with retailers, technology
proponents and metering service providers to
bundle products and tariffs to optimise value
for customers and simplify product offerings
while supporting efficient network outcomes.
Developing appropriate price signals to A valuation framework should be able to

encourage efficient uptake of EG, reducing determine the benefit to the network from
future network costs to below the level they EG.
might reach in the absence of EG uptake and
This could then be translated into a customer
thereby benefiting all customers. This could
incentive (e.g. cost-reflective network tariff
include:
arrangement) that would benefit the
 discounting connection charges, network.
network charges or providing direct
payment, reflecting the costs and
benefits available to the NSP, or
 implementing price signals or

incentives for customers to
undertake efficient investment in
storage, retrofit distributed inverters
and to be appropriately rewarded for
demand management.

Identifying and supporting opportunities to Historical subsidies have resulted in cross-
restructure current incentives and subsidies subsidies between different customers.
so that they can support technologies and These subsidies are committed and long
services that result in more efficient use of lasting (e.g. many premium FiTs do not
the network for the benefit of all customers expire until 2028). However, they do not
(for example, restructuring premium feed-in- promote uptake of new technologies, which
tariffs to enable customer to choose to use can improve the efficiency of the network by
these subsidies to install storage solutions reducing:
which support lower peak demand and
reduced network expenditure).  the technical impacts of high
penetration of solar photovoltaics
 peak demand, and hence the capital
investment required to augment and
maintain the network
the need to augment and maintain long
network connections to fringe-of-grid
communities.
Working with the industry and regulators to Innovation allowances allow network service
build on the principles of the Demand providers to trial new technologies that will
Management Innovation Allowance to help shape the network of the future.
enhance the industry’s ability to support
A key international scheme is the United
research and development activities that
Kingdom’s RIIO (Regulation = Incentives +
support innovation and integration of new
Innovation + Output) program.
technologies, and ultimately create a more
efficient network for the benefit of Note that the Australian Energy Market
customers. Commission recently published a rule change
to reform the Demand Management Incentive
Scheme (DMIS), including the DMIA, which
includes an innovation allowance for demand
management and the connection of EG.
Investigating opportunities for the This would necessarily include more detailed
development of cost-effective policies and cost —benefit analysis. It and would also rely
incentives (such as model availability on the availability of time -of -use pricing or
requirements and fuel standards) supporting similar demand- management incentives to
EV uptake, where it promotes efficient use ensure EV charging contributesd to efficient
of the network for the benefit of customers. network usage.

22 Several of the potential commercial opportunities for networks to offer new products and
services based on EG depend on the networks having cost-effective access to customer
data and/or a direct channel to the customer. (e.g. in home EG and demand
optimisation, residential storage). The proposed new metering rules present
for metering and related services but they also create risks associated with potential new
costs for data access and the loss of a traditional DNSP service.
Allow network service providers (NSPs) This would allow NSPs to use smart meters
access to a certain amount of customer data for certain network purposes and minimise
required for network services for free, and the commercial impacts on the meter data
access to additional data for network provider.
purposes at a reasonable cost, e.g. daily
voltage alarms.
Safeguard NSPs’ access to the customer data For NSPs to proactively and dynamically
required to safely manage the network. respond to challenges from increasing levels
of EG, increasingly granular real-time data
would likely be required.
Under the proposed new metering rules,
customer data will be accessed via the
metering coordinator. As such, NSPs may
need to pay to gain access to some data.
NSPs must work closely with stakeholders to
determine the ideal level of customer data
and how this access may be obtained.
In the metering rule change process and
Energy Market Reform Working Group review
of new products and services, the details of
how customer de-energisation and re-
energisation and management of load
switching to ensure customer and employee
safety and network security remains under
consideration.
23 Active involvement in the EG and energy services markets on the part of NSPs or their
related entities as either a market participant or to simply incentivise the efficient use of
EG for the benefit of all customers, will require developing partnerships with other
stakeholders including technology service providers and retailers.
Removing knowledge and organisational Partnerships and bundled products will be

barriers to cooperation between NSPs, crucial to ensure that customers are
technology providers and retailers to offer presented with persuasive offers that
bundled products. promote mutually beneficial EG outcomes.

24 The NSP business model needs to evolve to facilitate the integration of multiple
distributed energy resources, including EG. This could provide greater network capacity
and energy diversity to optimise grid performance for both supply and demand.
Propose that in future market arrangements, Distribution network service providers will be
the NSP can coordinate the day-to-day ideally placed to provide these services.
operation of multiple sources and types of
EG and other distributed energy resources,
where this provides the most cost-effective
option for the customer.
Reviewing the current regulatory framework Leading the market transition to a network
to determine its suitability for a more with increasing amounts of EG will require
customer-oriented electricity market. This NSPs to identify international best practice.
should include assessing its flexibility to
adapt to an increasingly complex market as a
result of the adoption of new technologies
and the changing priorities of customers.

7 Recommendations
Note on authorship
This section was written by Marchment Hill Consulting.
7.1 Introduction
The following recommendations have been prepared with the overarching principle of the
National Electricity Objective in mind. That is, investment in and operation of the network
should maximise the efficiency of the electricity supply chain, allowing consumers to enjoy
the quality, safety, reliability and security of supply that reflects their long-term interests at
the lowest cost. This should result in the efficient deployment and use of embedded
generation (EG) in locations and at times and scales that reflect the value of EG to the
customer, community and the network.
Increasing levels of EG on the network present challenges for network service providers
(NSPs). However, they can also benefit customers through the delivery of new energy services
and support more efficient use of existing network infrastructure. The challenge for
networks, retailers, regulators, technology proponents and customer representatives alike is
how they might work together to develop future business models and pricing structures that
fairly apportion value and costs to all stakeholders and turns the challenges into
opportunities.
These recommendations also attempt to recognise the often diverse responsibilities and
interests of stakeholders, while aligning their actions to benefit customers.
The recommendations have been grouped into three key main categories:
 Enabling Technologies and Pricing
o What are the key technologies to deliver future value?
o How should they be incentivised?
 Enabling Business Models
o How will the network of the future operate?
o What regulatory changes would promote this transition?
 Enabling Partnerships
o What partnerships are required for NSPs to actively participate in the EG
market?
Policy and regulatory options that support the recommendations are also provided, where
appropriate. Policy options relate to approaches that involve a change to existing government
policy, or a change to the general operational policies or approaches adopted by NSPs as a
whole.
Recommendations | Page 227

7.2 Enabling technologies and pricing
An appropriate valuation approach to capture all the costs and benefits of EG to
networks should be developed and implemented in close consultation with stakeholders
to enable effective signalling of the network value of EG to customers.
Our review of the technical, regulatory and commercial impacts of EG has found that EG
technologies — and storage technologies, in particular — may hold significant potential for
NSPs, retailers and customers alike.
The challenge for the future consists of designing appropriate incentives and pricing
structures to promote uptake and operation of EG technologies in a way that makes the
network more efficient – as this is fundamentally in the best interests of customers.
As a first step in this process, the actual impact of different EG technologies on the network
needs to be determined. This should include determining the capacity, reliability, safety and
quality of supply impact of different types of EG, taking into consideration specific network
topologies. The valuation framework developed for this report may be used as a starting
point, with the ultimate objective of ensuring that customers’ investment decisions reflect
their impact on the network.
Within in this context, networks could also incentivise customers to allow NSPs access to their
EG (e.g. their energy storage system) where network control of system can unlock additional
benefits. The incentive could take the form of a discount on customer connection charges,
tariffs or bills, or a direct payment reflecting the net benefits available to the network.
Policy and regulatory options and further work that support this recommendation are detailed
in Table 30.
Table 30 - Policy and regulatory options
Network P&R option type
NTR
Transmission
Organisation
Government
Distribution
Regulatory
Policy and regulatory option Further work scope of Priority

work
Working closely with a range of Review the National Feeder Taxonomy to

industry stakeholders (including determine its suitability for use in a WP6 –
EG proponents, the Australian     High
valuation framework for EG and seek to WPO 3/4
Energy Market Operator and the update it where required.
Australian Energy Regulator) to
develop an appropriate
valuation framework that can
Create a working group to finalise a
be leveraged to develop pricing WP6 –
    valuation framework and begin testing High
arrangements to promote WPO 3/4
with real NSP cost and benefit data.
efficient uptake of EG.
Developing appropriate price Based on the proposed valuation

signals to encourage efficient framework, undertake further modelling
WP6 –
uptake of EG, reducing future     to develop appropriate incentive High
WPO 3/4
network costs to below the mechanisms that promote efficient
level they might reach in the uptake of EG.

absence of EG uptake and
thereby benefiting all
customers. Review and consider the implications of
the recent rule change proposal regarding
This could include: Local Generation Network Credits79,
keeping in mind the overarching principle WP6 –
- discounting connection High
that the impacts of bidirectional power WPO 3/4
charges, network charges or flow and costs and benefits of EG should
providing direct payment, be taken into account when valuing the
reflecting the costs and benefits impact of EG.
available to the NSP, or
- implementing price signals or
incentives for customers to Review and consider the current trial of
undertake efficient investment virtual net-metering arrangements being
in storage, retrofit distributed funded by ARENA
inverters and to be (http://arena.gov.au/project/investigati
appropriately rewarded for ng-local-network-charges-and-virtual-net- WP6 –
Low
demand management. metering/), and investigate which lessons WPO 3/4
from this exercise can be applied by
DNSPs to explore emerging business
models that support efficient and
sustained network usage.
A comprehensive analysis should be

undertaken to identify the impacts of
reverse power flow at all stages of
WP6 –
transmission and distribution networks, Medium
WPO 3/4
including the effect on network assets
designed for one-way power flow and
Using the valuation framework other associated protection issues.
to manage the voltage
regulation, power quality,
reliability and safety impacts of Capture high-fidelity (temporal and
EG on the network. This could, geographical) data on the intermittency
for example, support the and related impacts of high-penetration
   intermittency distributed generation, in WP6 –
strategic deployment of Medium
medium-scale power electronics particular rooftop solar photovoltaics, to WPO 3/4
solutions, such as STATCOMS, at help identify optimal mitigation
locations where they may be strategies and associated control
the least-cost solution to help algorithms.
manage power quality issues.
A coherent framework should be sought
to help DNSPs identify mechanisms for
dealing with high-ramp situations caused WP8 –
Medium
by high-penetration, intermittent EG. WPO 4
AEMO = Australian Energy Market Operator; AER = Australian Energy Regulator; ARENA = Australian Renewable
Energy Agency; DNSP = distribution network service provider; NSP = network service provider; NTR = Network
Transformation Roadmap; P&R = policy and regulatory; STATCOMs = static synchronous compensators; WP = Work
Program; WPO = Work Program Objective
Support the implementation of cost-reflective network price signals that promote

uptake of EG in ways that are beneficial to all customers.
Current energy-based tariffs have been found to create a cross-subsidy from non-EG owners
to EG owners. Implementing appropriately designed demand tariffs would go some way
towards addressing this impact and allow NSPs to accurately signal the costs to the customer
of future expansions in network capacity. This is a crucial part of promoting uptake and
operation of EG that contributes to the long-term objective of making the network more
efficient.
79Oakley Greenwood, Local Generation Network Credit Rule Change Proposal – Proposed by City of Sydney, Total
Environment Centre and Property Council of Australia, July 2015.

Cost reflectivity should be an overarching principle for pricing across the network, capturing
the impacts of EG depending on its location and the network type. Further initiatives for cost-
reflective pricing could include cost-reflective feed-in tariffs (FiTs) or connection charges for
EG owners.
in Table 31.
Transmission NTR
Organisation
Government
Distribution
Regulatory
work
Supporting a national approach Model the impact of EG uptake (type and

to cost-reflective pricing that capacity) under different demand tariffs.
removes inconsistent This could be further expanded to model WP8 –
    High
jurisdictional obligations on the impact of introducing a variable feed- WPO 5
tariff structures and in tariff for EG owners that reflects
assignment. network value.
Supporting the implementation Review the proposed metering reforms to

of a balanced framework for highlight potential risks and opportunities
the uptake of smart meters for for network service providers, and
the fastest economic roll out to develop recommendations to manage WP8 –
    Medium
benefit all customers. risks and leverage opportunities. WPO 3
Highlight issues (where possible) through
the remaining stakeholder consultation
process for the reforms.
Encouraging adoption of
technologies that support
Undertake a review to determine the
efficient pricing and customers’
technologies that hold the most promise WP8 –
ability to access data and    Medium
for customers to effectively manage their WPO 3
manage their demand (e.g. in-
demand.
home displays, apps, data
portals).
Working closely with retailers,

technology proponents and
metering service providers to Undertake modelling to understand how
bundle products and tariffs to offers might be structured (e.g. storage + WP8 –
  Medium
optimise value for customers demand tariff offer for solar photovoltaic WPO 5
and simplify product offerings owners) to be attractive to customers.
while supporting efficient
network outcomes.
NTR = Network Transformation Roadmap; P&R = policy and regulatory; WP = Work Program; WPO = Work Program
Objective
Support incentives and policies that promote innovation and commercialisation in

relation to energy storage technologies in ways that enhance the efficient use of the
network for the benefit of all customers.
Historical incentives for EG, including premium solar photovoltaic FiTs, have resulted in
commercial imbalances such as cross-subsidisation, and present a major diluting impact to
the intentions of cost-reflective network pricing.

Any future incentives should be designed to support efficient use of the network. This
includes winding back or restructuring current, inefficient incentives and subsidies. One
option may be for all customers to transfer all or part of the value of the premium FiT subsidy
(which in some cases remains until 2028) towards a demand management solution that
unlocks additional benefits to the networks and their customers. Subsidies currently provided
to rural communities, such as the community service obligation payment in Queensland, could
also be redirected to install microgrids in order to reduce reliance on costly long distance grid
connections.
in Table 32.
NTR
Transmission
Organisation
Government
Distribution
Regulatory
work
Review the arrangement for community

service obligation and other similar
subsidies for opportunities to divert part
Identifying and supporting of this subsidy to the installation of EG,
opportunities to restructure where this provides the most cost- WP6-
High
current incentives and subsidies effective solution to the customer. WPO 5
so that they can support
technologies and services that Further model the cost—benefit impact to
result in more efficient use of determine where on the network such EG
the network for the benefit of installations are cost effective.
   
all customers (for example,
restructuring premium feed-in- Model the impacts of transferring the
tariffs to enable customer to value of premium feed-in tariffs to
choose to use these subsidies to demand-management technologies to
install storage solutions which understand how such a solution would be
WP6-
support lower peak demand and structured to be attractive to customers, High
WPO 5
reduced network expenditure). networks and retailers, and avoid
unintended consequences (e.g. impacts
on environmental charges or wholesale
energy prices).
Working with the industry and

regulators to build on the
principles of the Demand
Management Innovation
Review international innovation schemes
Allowance to enhance the
(e.g. the United Kingdom’s RIIO scheme)
industry’s ability to support
to identify learnings and best practice. WP6-
research and development    Medium
Present the findings of the review to the WPO 5
activities that support
Australian Energy Regulator to inform
innovation and integration of
their development of the new DMIA.
new technologies, and
ultimately create a more
efficient network for the
benefit of customers.
Investigating opportunities for

the development of cost-
effective policies and incentives Build on the work completed by Energia
(such as model availability for the ESAA on cost-effective policy
WP6-
requirements and fuel     options to support the uptake of electric Medium
WPO 5
standards) supporting EV vehicles and assess the appropriate next
uptake, where it promotes steps.
efficient use of the network for
the benefit of customers.

DMIA = Demand Management Innovation Allowance; EG = embedded generation; ESAA = Energy Supply Association
of Australia; NTR = Network Transformation Roadmap; P&R = policy and regulatory; RIIO = Regulation = Incentives
+ Innovation + Output; Work Program; WPO = Work Program Objective
Support the development of appropriate standards that facilitate the safe integration
of EG on the network.
The standards gap analysis undertaken for this report found that standards need to be
developed to support the integration of EG technologies and ensure safe operation of the
network.
Energy storage is likely to be the next EG technology to be widely adopted by customers. The
development of a standard to provide guidelines for the safe installation and operation of
small-scale energy storage systems is therefore of particular importance.
in Table 33.
NTR
Transmission
Organisation
Government
Distribution
Regulatory

work
Consider proposing (or petitioning

Supporting the development of
Standards Australia to commence)
a new Australian Standard to
development of a standard to set out WP8 –
manage the functionality of   High
specific functionality and requirements WPO 4
protection relays for inverter
for protection relays for inverter energy
energy systems.
systems.
Revisiting the development of

standards such as AS/NZS 4755
Part 3.4, or facilitating the Petition or propose to Standards Australia
WP8 –
local acceptance of a similar   to recommence the development of High
WPO 4
international standard for AS/NZS 4755.3.4 standards.
managing EV charging and
discharging.
Supporting the development of

standards relating to the safety
of small-scale and large/grid- Encourage the development of Australian
scale energy storage systems. standards in energy storage that relate to
Focus should be directed to the safe installation, operation and
control of small-scale energy storage WP8 –
those addressing safe    High
systems. WPO 4
installation/operation and fire
protection. In addition, lithium Investigate the progress of the Clean
chemistry may require Energy Council and build on these efforts.
particular attention, because its
standards are immature.
NTR = Network Transformation Roadmap; P&R = policy and regulatory; Work Program; WPO = Work Program
Objective

7.3 Enabling business models
There should be no inherent advantage or disadvantage for network providers in
competing to provide services and pursuing commercial opportunities related to EG.
This report has found a number of barriers that currently exist to network service providers
engaging in the EG market, including restrictions on owning and operating generation units,
cumbersome national ring-fencing regulations, and potential data access restrictions under
the proposed new metering rules.
The regulatory framework should be reformed with the overarching principle that NSPs should
be able to pursue opportunities and capture the full benefits available to them from investing
in EG, within the reasonable limitations of the competitive environment.
in Table 34.
Table 34 – Policy and regulatory options
NTR
Transmission
Organisation
Government
Distribution
Regulatory

work
Supporting changes to current

state-based ring-fencing rules
to create a national framework
with the overarching principle
to ensure no market participant    
is unnecessarily constrained in a
competitive environment with
regard to providing services to Review the current ring-fencing rules and
customers. other state-based rules to identify key WP6 –
Medium
shortcomings and prepare a submission to WPO 4
the Australian Energy Market Commission.
Supporting the removal of
state-based rules barring or
limiting network providers’
   
ability to own and operate EG
(for the purpose of network
support).
Ensuring that NSPs are free to

enter into direct commercially
based trading relationships with
Further investigate the potential for
customers and that NSP’s
customers to enter into direct,
related entities should not be WP6-
    commercially based trading relationships Medium
restricted from providing WPO 5
with NSPs, or for NSPs to offer services as
energy selling services to
an alternative energy seller.
customers, where they meet
appropriate ring-fencing
requirements.
Supporting a review of relevant

rules to ensure no barriers exist
to the creation, operation and Review the current regulatory framework
maintenance of microgrids or to identify barriers to NSPs establishing WP6-
  Medium
standalone systems by networks and operating microgrids where it is the WPO 5
as an option for the most most cost-efficient supply option.
efficient approach to energy
supply

NTR = Network Transformation Roadmap; P&R = policy and regulatory; RAB = regulatory asset base; WP = Work
Program; WPO = Work Program Objective
There should be no inherent advantage or disadvantage for network providers to utilise

EG for network purposes compared with network upgrades.
Current regulations for NSPs to use EG for network purposes do not provide sufficient
incentives. NSPs are incentivised to increase capital expenditure through the regulated return
they earn on the regulatory asset base. A similar incentive does not exist for non-network
solutions, which means that NSPs would need to capture additional market benefits from the
demand management initiative to make it worth their while.
Recent proposed reforms to the Demand Management Incentive Scheme have partly addressed
this by stating several core principles for the Australian Energy Regulator to consider when
designing a future scheme. This includes consideration of the full value stream from demand
management initiatives. Overall, these principles are sound. However, more clarity will be
required regarding the design of the scheme to ensure that NSPs can capture enough benefits
to offset the forgone return on capital.
in Table 35.
NTR
Transmission
Organisation
Government
Distribution
Regulatory

Work
Working with the industry and Undertake modelling to understand the

regulators to identify where level of benefits required to make
there may still be a demand-management initiatives more WP6-
Medium
disadvantage for network persuasive for network service providers WPO 5
providers to utilise EG for to support the development of the
network purposes compared revised DMIS.
with network upgrades. NSPs   
could then build on the recent Further research is needed to improve
reforms to the Demand net load forecasting methodologies
Management Incentive Scheme incorporating EG and demand response. WP6-
to ensure that any Medium
This will assist in network planning and WPO 5
disadvantages for NSPs using EG scheduling of generation and storage
in this way is addressed. assets.
DMIS = Demand Management Incentive Scheme; NTR = Network Transformation Roadmap; P&R = policy and
regulatory; WP = Work Program; WPO = Work Program Objective
Support the evolution of the network provider business model to include that of the grid
integrator — efficiently and effectively integrating multiple distributed energy
resources, including EG.
As the levels of EG on the network increases, it will become more challenging to effectively
manage the network. However, proactively managing the network under these conditions
could provide NSPs with additional value. NSPs should therefore move towards becoming the

provider of EG integration services, with the ultimate goal of reducing costs and improving
grid performance.
This could be further expanded to facilitate a marketplace in which energy services, demand
and supply from distributed and centralised sources may be traded, providing NSPs with an
additional revenue stream. Under this scenario, it will be crucial to understand the EG value
streams available to different stakeholders to enable efficient pricing and delivery of
services.
in Table 36.
NTR
Transmission
Organisation
Government
Distribution
Regulatory
work
Propose that in future market Assess the impact of proposed metering

WP6-
arrangements, the NSP can reforms and other existing regulations on Medium
WPO 5
coordinate the day-to-day the ability of NSPs to perform this role.
operation of multiple sources
and types of EG and other     Review (including international case
distributed energy resources, studies) how this role might evolve for
where this provides the most WP6-
DNSPs and the steps needed in the near Medium
cost-effective option for the WPO 5
term to support DNSPs’ contribution to
customer. customer value.
DNSP = distribution network service provider; NTR = Network Transformation Roadmap; P&R = policy and
regulatory; WP = Work Program; WPO = Work Program Objective
7.4 Enabling partnerships

Network providers should work towards developing partnerships with relevant energy
service providers to promote sustainable and efficient uptake and operation of EG.
The active involvement of NSPs in the EG and energy services market relies on developing
partnerships with technology and service providers and retailers.
Partnerships will be crucial to ensure that customer are presented with persuasive product
offers that promote mutually beneficial EG outcomes. Bundling of product and pricing offers
to minimise the impact on customers from the introduction of new tariff structures (e.g.
maximum demand tariffs), while at the same time supporting efficient operation of EG to
maximise its benefits, will be crucial under a future market state with high levels of EG.
in Table 37.

NTR
Transmission
Organisation
Government
Distribution
Regulatory
work
Undertake modelling to understand how

offers might be structured (e.g. storage + WP8 –
Medium
Removing knowledge and demand tariff offer for solar PV owners) WPO 5
organisational barriers to to be attractive to customers.
cooperation between NSPs,
   
technology providers and Enable the Energy Networks Association
retailers to offer bundled to facilitate understanding and removal
products. WP8 –
of current barriers to cooperation (e.g. Medium
WPO 5
organising workshops with technology
providers and retailers).
In response to a future electricity market that is likely to be much more customer-

oriented, NSPs should optimise existing services around customer needs and expand
organisational capacity to enable the provision of services that future electricity
customers will value.
NSPs will require an increasingly commercial and customer-oriented mindset to actively
engage in the emerging market for energy services. In addition to relevant marketing and
communications abilities, NSPs may also need to improve performance management and
customer insights capabilities.
This would represent a step change in the culture of many NSPs. Close cooperation with
regulators will be key to ensure that his market transition is introduced in a way that provides
NSPs with sufficient time to implement the required changes.
in Table 38.
NTR
Transmission
Organisation
Government
Distribution
Regulatory

work
Reviewing the current

regulatory framework to
determine its suitability for a
more customer-oriented
electricity market. This should Undertake a review to determine WP8 –
Medium
include assessing its flexibility      international best-practice regulatory and WPO 5
to adapt to an increasingly policy approaches.
complex market as a result of
the adoption of new
technologies and the changing
priorities of customers.

Appendix A – Survey responses
As part of this project, we surveyed members of the Energy Networks Association to
understand their views on the impacts of embedded generation (EG) on the network. Nine
responses were received. Figure 56 shows the number of respondents by state.
1
2
1
VIC
QLD
NSW
SA
1 WA
NT
2
Figure 56 - Embedded generation survey responses by state
The survey questions are reproduced on the following pages.
Appendix A – Survey responses | Page 237

Impacts of Embedded Generation on Distribution Networks –
Survey
The survey covers Technical, Regulatory and Commercial impacts for Embedded Generation
(EG). For the purposes of this survey, please consider EG as a unit that generates electricity
at a customer’s premises and is connected to a distribution system.
General Information
Organisation:
Sections /
Contact details
Respondents Name Role / Title questions
(email/phone)
completed
#1
#2
#3
#4
Technical Impacts
1. Please indicate what issues you are starting to see in your networks because of
embedded generation:
☐ Voltage Rise Power Quality80 ☐ Protection Problems ☐ Thermal

☐
Loading
☐ Stability / High Neutral ☐ Other (please

☐
Unbalance Current specify)
Comment:
80
Please indicate which qualities are of concern

2. In which part(s) of your networks are you seeing issues due to embedded
generation?
☐ Low-voltage Medium-voltage81 ☐ High-voltage ☐ Other (please

☐
specify)
Comment:
3. What do you believe are the primary causes of the issues you are seeing on your
networks due to embedded generation?
Comment:
4. What are the penetration-levels on your networks of the following EG

technologies?
Technology MW Number %
Solar PV (at-scale or distributed)
Wind
Battery storage
Diesel / Gas Generators
Other (please specify)
Comment:
81
Specify voltage levels if relevant

5. Which type(s) of embedded generation are causing issues in your networks?
Comment:
Comment:
6. What do you think are the barriers for higher penetration levels of embedded
generation in your networks? Please also state the type(s) of network if applicable
(e.g. low source impedance, conductor impedance and length, adequate
protection).
7. What of the following mitigation approaches have you taken or are you considering
to manage the impacts of embedded generation?
☐ Ramp-rate restrictions
☐ Size restrictions / disallowing connection of EG
☐ VAr compensation (e.g. statcoms, capacitor banks, smart inverters)
☐ Balancing network
☐ Changing transformer taps
☐ Voltage regulation devices
☐ Other (please specify)
Comment:

8. Where do you see the greatest value for your networks for embedded generation?
(e.g. reduced network capital expenditure)
Comment:
9. What would be the key drivers for enabling and encouraging higher penetration
levels of embedded generation in your networks? Please also state the type(s) of
network which might benefit from increased levels of embedded generation (e.g.
reduced ADMD and reduced network requirements)
Comment:
Comment:
10. Does embedded generation play a meaningful part in network planning and in
deferral of network augmentation?
11. What protection mechanisms are being deployed, upgraded or refined in light of
new embedded generation and (particularly) reverse power flow? (e.g. changed
settings on Regulators, adjustment to Line Drop Compensation)
Comment:

12. What embedded generation do you own/operate, where is it typically deployed and
what role does it generally play?
Comment:
13. What are the emerging trends for embedded generation uptake for your customer
base and over the next fifteen years how do you believe you will need to respond?
Technology Comment
Distributed Solar
Behind-the-meter
Storage
Off-grid EG / Back-
up Power
Community /
Commercial EG
(e.g. medium
scale)
Other (please
specify)

14. What mechanisms do you have to support the managed growth of embedded
generation?
Comment:
15. Are there any standards that you believe need improvement to properly address
the impacts of embedded generation? (e.g. AS4777 needs to keep pace with IES
technology)
Comment:
16. Are there any areas of embedded generation that you believe could benefit from
standards-development in those spaces?
Comment:
17. Are you aware of overseas standards that Australia should incorporate to better
address the impacts of embedded generation?
Comment:

Regulatory Context
18. Please rate on a scale of 1-10 the adequacy of current regulatory instruments in
meeting DNSP and customer needs to facilitate safe, effective and efficient
connection of EG (either network or customer connected) [1-2 = entirely
inadequate, 3-4 = inadequate, 5-6 = neutral, 7-8 = adequate , 9-10 = perfectly
adequate]
Rating (1-
Regulatory Instrument
10)
NER Chapter 5
NER Chapter 5A
AER Connection Charge Guidelines for Electricity Retail Customers
Service and installation rules (State-specific)
Comment:
19. Where you have found current regulatory instruments to be inadequate, please
provide a brief description of your main concerns.
Comment:

20. If possible, please also provide any proposed solutions where regulatory
instruments have been found inadequate.
Comment:
21. Please rate on a scale of 1-10 the primary cost components (or resource effort)
related to connecting EG (either network or customer connected) to your network?
[1-2 = insignificant cost or effort, 3-4 = minor cost or effort, 5-6 = moderate cost
or effort, 7-8 = major cost or effort, 9-10 = entire cost or effort]
Rating (1-
Components
10)
Administrative
Hardware (meters)
Site inspections
Installation work inspections
Network studies
Augmentation
Comment:
22. Please rate on a scale of 1-10 the primary regulatory cost drivers related to
connecting EG (either network or customer connected) to your network? [1-2 =
insignificant driver, 3-4 = minor driver, 5-6 moderate driver, 7-8 major driver, 9-
10 sole driver]
Rating (1-
Cost Drivers
10)
Safety

Reliability
Quality of supply
Service and installation rules
Comment:
23. Are the regulatory cost drivers detailed above sufficient or should they be
strengthened or relaxed in some way? Please explain.
Comment:
24. Has your network required augmentation as a result of high penetration of

embedded generation (either network or customer connected) which could not be
fully recovered through the current regulatory framework?
☐ Yes ☐ No
If Yes, what was the specific regulatory barrier?
Comment:
25. If No, is there a concern that this will be a future issue due to continued
embedded generation penetration?
Comment:

26. Do you have any suggestions for how the current regulatory framework may be
reformed to allow for more efficient recovery of costs associated with connecting
embedded generation (either network or customer connected)?
Comment:

Commercial impacts
27. Please rate on a level of 1-10 the (direct or indirect) commercial impacts on
network businesses of ever increasing penetration levels of EG into the distribution
network.
[1-2 = insignificant impact, 3-4 = minor impact, 5-6 = moderate impact, 7-8 =
major impact 9-10 = extreme impact]
Rating (1-
Commercial Impacts
10)
Costs of physical connection services
Under-recovery of network value from EG customers (e.g. cross-subsidy

between solar PV customers and non-solar PV customers)
Inability to recover augmentation costs resulting from high embedded

generation penetration
Installation of additional power quality monitoring and control equipment
Costs of related services (load, power quality surveys etc.)
28. Do you have any suggestions for regulatory or policy options to address these
impacts? (excluding cost reflective network pricing)
Comment:
29. Do you see potential for a mechanism that allows DNSPs to pay owners of
embedded generators to supply peak demand support?
☐ Yes ☐ No

30. How should such a mechanism be structured:
☐ Direct payment? (based on actual contribution)
☐ Discounted bill? (annual average estimate)
☐ Discounted tariff? (annual average estimate)
☐ Other (please specify)
Comment:
31. Should it be within or outside of the RIT-D process? Please provide justification
Comment:
If you have any additional information that you believe is relevant to the impacts and
potential benefits for embedded generation, please detail them below:
Comment:

Appendix B – Analysis of embedded
generation/distributed energy resources uptake
and network impact forecasting techniques
Introduction
Electricity utilities in Australia are potentially entering into a time of declining network
utilisation at some locations, which places upward pressure on network prices and hence
retail electricity prices. A range of factors, including economic volatility, improving
household energy efficiency, and in particular, the increasing prevalence of embedded
generation (EG), have contributed to reduced electricity consumption and deteriorating
network utilisation. This is especially the case with residential solar photovoltaics (PV). At the
same time, peak electricity demand is growing due to a wide use of air conditioners. Future
large-scale adoption of electric vehicles (EVs) will significantly increase cumulative load on
the electricity grid, but also present opportunities to manage peak demand (Paevere et al.
2014). Of particular current interest to utilities is the distortionary effect that large uptake of
EVs and solar PV (including off-grid systems) is having on household electricity consumption
patterns and its cumulative effect on the electricity distribution network.
This report reviews the past research, technologies and reports that address these questions.
By highlighting the strengths and weaknesses of the different approaches, this report
identifies priorities for improving medium to long-term uptake forecasting of EG. Since these
questions are complex, past methodologies have only partly addressed the questions or
considered a limited number of variables involved.
Review of methodologies
As part of providing valuable information addressing the medium to long-term forecasting
questions for utilities, methodologies in research papers and reports tended to predominantly
look at forecasting adoption of technologies that will have a major influence on cumulative
and peak electricity demand from the electricity grid. These technology-forecasting
methodologies can be broadly grouped into techniques focusing on residential and non-
residential uptake sectors.
Technology forecasting — residential sector
Forecasting technology uptake across building stock allows utilities to estimate the
incremental change to load, given assumptions of electricity consumption and usage of the
technology. The most common approach to forecasting technology uptake is based on
mathematical diffusion and choice modelling methods.
Modelling approaches to diffusion commenced in the 1960s, particularly with the introduction
of the well-known Bass model (1969) and the logistics model Mansfield (1961). These models
basically mimic the Rogers (1962) adoption curve over time (innovators, early adopters, early
majority, late majority, laggards), which is of a sigmoidal form. Over the past 50 years, these
Appendix B – Analysis of embedded generation/distributed energy resources uptake and network impact forecasting techniques |
Page 250
methods have been applied to a wide range of appliances to estimate the percentage of
consumers adopting over time. The early Bass and logit models only considered basic
variables representing innovation and imitation. They were limited to appliances that already
had adoption beyond the innovator cycle, to allow calibration to the adoption cycle.
In later years, these basic diffusion models were extended to incorporate features that
influence adoption. For example, a Generalised Bass Model (Bass et al. 1994) was produced
that can accommodate pricing and advertising variables based on observed data. Other
attempts included Horsky (1990), who incorporated a basic utility function of wage,
amusement factor (or non-financial benefits), price and savings. This was extended by Higgins
et al. (2011) to accommodate ceilings of adoption and interactions between intervention
options. The literature expanded rapidly with diffusion models (Greene et al. 2005; Horne et
al. 2005; Jaccard and Dennis 2006; Soderholm and Klaassenn 2007) that incorporate a utility
function based on regression to represent the different features of the technology (e.g. price,
going costs, appeal) that the consumer would consider for adoption.
Technologies that are considered to potentially have the largest medium to long-term impacts
on the electricity grid are solar PVs, water heaters and batteries connected to PV and EVs.
The diffusion of PVs and EVs has been extensively studied in Australia and overseas.
Application of diffusion models to each of these technologies requires identifying the
variables that impact adoption, the strength of each variable, and a way to calibrate a
mathematical diffusion model using these variables.
Choice experiments have been an effective approach to identifying key features important to
the adoption of these types of technologies. This involves modelling consumer preferences for
these technologies to derive the effect of key attributes, such as installation cost, energy
efficiency and out-of-pocket expenses after rebates.
Islam and Meade (2013) conducted a discrete choice experiment in Ontario, Canada, to
estimate key attributes of consumer adoption intention of PVs over time. The authors use
these attributes in a discrete time survival mixture analysis to estimate future probabilities of
adoption for different consumers. Yamaguchi et al. (2013) used consumer preferences to
estimate the marketing effort component of a Bass diffusion model.
In Australia, discrete choice data gathered from stated preference experiments have been
used to understand the drivers and interaction effects of uptake of different types of water
heaters. Bartels et al. (2004 and 2006) surveyed 129 plumbers and 312 consumers in Sydney
who purchased a water heater. The stated choice experiment allowed the respondent to
indicate important features to the purchase of their gas or electric water heaters. In the case
of solar PVs, the Australian Renewable Energy Agency (ARENA 2013) quantified the effect of
different demographic and economic drivers to PV adoption across different regions of
Australia. They used actual PV uptake data by postcode and applied a regression to a range of
variables in Australian Bureau of Statistics (ABS) demographic data. While these methods help
understand drivers of consumer behaviour to adoption, their application to forecast future
uptake needs to be linked with diffusion models.
Once methods such as choice experiments have identified key variables, they can be used in
mathematical-based diffusion and discrete choice models as part of producing future
forecasts in the residential sector. Discrete choice modelling has been extensively used for
technology diffusion (Jun and Kim 2011; Kim et al. 2005) due to the ability to consider the
Page 251
sensitivity of multiple product options available to consumers, which is an important
consideration for EVs and PV systems. There have also been applications for the diffusion of
EV options (Lin and Greene 2010; Higgins et al. 2012) and energy-efficient appliances (Murphy
et al. 2007; Jaccard and Dennis 2006). Linear utility functions, which can be derived from
choice experiments, have been extensively incorporated into a choice model to represent the
different values of the product features (e.g. price, rebate, annual costs) to the consumer
(Horsky 1990; Revelt and Train 1998; Greene et al. 2005; Soderholn and Klassen 2007).
When considering a medium-to-long forecasting horizon, there is the need to incorporate
product replacement along with first-time purchasers into a choice or diffusion model (Revel
and Train 1998; Olson and Choi 1985; Jun and Kim 2011). Initially, consumers are mostly first-
time purchasers of PV systems, though they can replace or upgrade to a larger system. In the
case of EVs, a consumer will have the choice to replace the internal combustion engine
vehicle with a similar vehicle or an EV.
There have been some diffusion model extensions that incorporate repeat purchases either
due to an upgrade or upon product failure (Revelt and Train, 1998; Olson and Choi, 1985). A
multiproduct choice model by Jun and Kim (2011) considers replacement timing to be
dependent on a survivability (or failure) probability function. It assumes consumers will
replace a PV system or water heater upon failure of the existing one. Higgins et al. (2014)
considered the case where the timing of product replacements is a function of failure
distribution, along with a utility of other attributes associated with replacement. When
implemented, it accommodates the socio-demographic differences of consumers who are
early versus late replacers of their existing appliance, as well as the effects of incentives to
reduce the time to replacement. This is important for understanding the timing of uptake of
different types of EVs into the Australian market.
Solar photovoltaic systems

Forecasting the future adoption of solar PV systems is of significant interest to governments
(due to the cost of incentives and feed-in tariffs), and to utilities due to impacts across the
electricity grid. Utilities require forecasts with a much higher geographical granularity, as the
different rates of uptake at different locations will have implications on substations and
feeders.
Over the past few years, there have been several reports on national or state level PV
projections in Australia. AEMO (2012) provides projections out to 2022 based on basic
financial features of PVs and solar generation. These national level trends cannot be
disaggregated into small units representing substations (a few thousand dwellings) or feeders
(a few hundred dwellings or less), due to large differences in uptake rates and installation
sizes between locations. They usually do not accommodate physical features/constraints of
PV installations such as roof space and PV/inverter size. In a more recent report, AEMO (2015)
produced PV and battery storage forecast uptakes to 2017/2018 and 2034/2035 at state level.
The report provides no information about the mathematical forecasting model used, other
than noting an economic model with the primary variable being payback period for each PV
and battery option. A major limitation of using payback period without other variables is that
it does not account for the high purchase price barrier to households for many options. A
further limitation of the method is that forecasts at state level do not accommodate the
Page 252
typology of building stock and roof space availability, which would produce a ceiling of
adoption significantly lower than forecasts based on financial variables.
From a utility’s perspective, medium-to-long to long-term forecasting of PV adoption requires
knowledge of number and distribution of size of installation at each location, along with how
the size of new and existing installations change over time. Application of choice and
diffusion models at high geographical sensitivity in the Australian residential sector is made
possible using ABS census data. The 2011 ABS census data contains valuable information about
the number of households, housing types, number of vehicles and other demographics for
each SA1 geographical unit in Australia. Each SA1 unit contains about 180 households, which is
granular enough to derive valuable load and peak demand information for feeder and
substation analysis. Choice or diffusion models with high geographical sensitivity have
received very limited attention in the literature to date, let alone in the case of technologies
such as EVs and PVs. Bhat and Guo (2004) and Sener et al. (2009) develop spatially explicit
choice models that accommodate socio-demographic differences and zone dependencies.
In the case of solar PV, Higgins et al. (2014a and b) implemented a choice-diffusion model for
uptake of different size installations of solar PV across each ABS SA1 in Townsville and New
South Wales. Since consumers may update or replace the PV/inverter over time, an additional
variable was included to accommodate the reduced risk or cost of replacing the existing
technology with the same type. A consumer can upgrade their PV system by either adding
more panels or upgrading to a larger inverter at a lower cost than a complete replacement.
Diffusion models, particularly the Bass and logit models, have been applied extensively to
forecast uptake of solar PVs with and without consumer incentives (Yamaguchi et al. 2013;
Guidolin and Mortarino, 2012; Higgins et al. 2011). These models contain features or
additional parameters (e.g. Generalised Bass Model) based on marketing, prices and costs,
which allow sensitivities to interventions (e.g. incentives, tariffs) to be tested. In practice,
the adoption of solar PV systems and batteries is more complicated than represented by a
basic diffusion model, since households often upgrade their solar PV or add a battery later on.
There have been some diffusion model extensions that incorporate repeat purchases either
due to an upgrade or upon product failure (Revelt and Train, 1998; Olson and Choi, 1985).
Distributed electricity storage in the form of batteries connected to residential dwellings,
with or without solar PV, has the potential to shift electricity load away from the peak
periods. This has been shown for battery storage via vehicle-to-grid strategies, through trials
and simulations in Brazil (Drude et al. 2014) and Australia (Paevere et al. 2014), given price
tariffs that provide adequate incentives. Based on industry sources, Ernst et al. (2011)
consider two end-user price scenarios for lithium-ion batteries in large production. For 4, 12
and 20-kWh batteries, the end-user prices are €2,000, 6,000 and 10,000 according to the first
scenario, and €4,000, 12,000 and 20,000 for the second scenario. Rudolf and Papastergiou
(2013) suggest 1,500 €/kWh for lithium-ion batteries, which is a very high cost. However, this
technology has the highest energy density of all available storage technologies.
Forecasts of battery storage (connected to PV) uptake have had much less attention
compared to PV systems alone. The choice-diffusion model developed by Higgins et al.
(2014b) for uptake of solar PV and batteries under different price tariffs required several
technical innovations by accommodating the:
• compatibility of each solar PV or battery option to dwelling types
Page 253
• price of adding a battery to an existing solar PV or upgrading/downgrading a solar PV
(e.g. increasing the number of panels), which is less than the price for a first-time
adopter
• representative effects of modelled choice probabilities in the presence of large
differences in price and annual electricity savings between the PV/battery options.
Also, solar PVs now have an established market, whereas batteries have so far had negligible
market penetration and are much more expensive. Incorporating both solar PV and batteries
in a single choice-diffusion model was more difficult as a result. The study by Higgins et al.
(2014b) indicated that about 5.5% of households in Townsville would own a battery by 2025
under the current flat tariff structure. The results were sensitive to the price tariff, and more
so to assumptions of future price projections of batteries and PVs. The percentage of
households in Townsville with PV in 2022 was forecast to be about 37%, up from 18% in 2013.
The AEMO (2015) report indicated that there would be approximately 450,000 new PV
installations in Queensland by 2025, which would lead to 45% of building stock having PV
systems.
The results on PV-plus-battery uptake shown by Higgins et al. (2014b) were considered
conservative by some experts in Australian utilities. Under the current price tariffs, about 5%
of Townsville households would have a PV-plus-battery installation, which is less than the
approximate 9% (for Queensland) indicated by AEMO (2015). Uptake projections for both PV
and PV-plus-battery in Higgins et al. (2014b) were considerably less than those in the AEMO
study. This is likely due to the model in Higgins et al. (2014b) accommodating additional
variables (e.g. house size and type, purchase price, demographics) that would represent
barriers to households adopting PVs and batteries, and thus leading to more conservative
forecasts.
Future adoption of PV-plus-batteries is likely to be influenced by additional variables to those
considered in choice-diffusion models. These include differences in energy use behaviour
within a particular demographic or housing type, which may make batteries more attractive
in lower-consumption households. Non-financial variables that should considered for battery
forecasts include remote connections (where it is expensive to extend an electricity network)
and behaviour towards fast electricity price rises.
Electric vehicles
Unlike solar PV systems, consumer decisions to purchase technologies such as EVs are
influenced by a larger range of variables, including costs/benefits, performance,
appeal/status, risk, psychographics (e.g. attitude to reducing energy consumption),
demographics and physical limitations. Gardner et al. (2011) identified the importance of
such criteria for EVs using surveys and focus groups.
Over the past 10 years, a large number of studies have forecast future adoption and vehicle
stock of battery electric vehicle (BEV) and plug-in hybrid electric vehicles (PHEV) across the
world. This includes California (Becker et al. 2009), the broader United States market (Orhach
and Fruchter 2011; Won et al. 2009) and Iceland (Shafiei et al. (2012). A range of
methodologies are used to produce these state and national level forecasts, including Bass
diffusion and agent-based modelling. In Australia, national EV (BEV and PHEV) sales forecasts
Page 254
to 2040 were produced by AECOM (2009) based on choice modellings. Graham et al. (2009)
forecast different types of EVs (PHEV, BEV, mild hybrid) to 2050 using a partial equilibrium
modelling approach. Jarvinen et al. (2011) forecast uptake of passenger EV sales using a
simple method that allocates high and low scenarios to new car sales trends.
There were major differences in the results between these studies, with the AECOM report
indicating only 20% of vehicle sales in 2020 being internal combustion engine vehicles, and
about 25% of vehicle sales in 2015 being traditional hybrid vehicles. By 2040 about 85% of all
vehicle sales were PHEV or EVs. The Graham report provided scenarios based on oil price
projections, with all forecasts showing internal combustion engine vehicles comprising
90%+ km travelled in 2020, but dropping to 20% of all vehicle km by 2040. The Jarvinen report
indicated a massive difference in EV sales at 2020 between their low and high scenarios,
where 2% of total vehicle sales were EVs in the low scenario (also used in AEMO’s 2015
report), increasing to 22% under the high scenario. The large differences in these forecasts
would also have led to large differences in medium to long-term load forecasts on the
electricity grid.
Paevere et al. (2014) showed that to understand and plan for the impacts of EV charging loads
on local electrical distribution networks, and to potentially exploit the opportunities for peak
shaving using stored energy in EV batteries, it is necessary to consider at least four
interacting layers of spatial and temporal variability in the analyses and projections. These
four layers are distribution system capacity, EV uptake, EV usage and charging patterns, and
household energy demand (Paevere et al. 2014). The spatial variability in each of these layers
for the purposes of EV grid impacts assessment is described below.
Geographically sensitive diffusion models for EV uptake have been very limited. One example
is by Lin and Greene (2010), who partition the vehicle market into 1458 segments based on
region, driver type, technology attitude and home charging availability. Higgins et al. (2012)
accommodate geographical differences by introducing a location-by-demographic typology
into the model. This was done using more than 9000 ABS census geographical units of
residential housing stock (about 250 households per unit), which were further partitioned by
demographics and building features. This allows the population of consumers to be
partitioned into category combinations of location by income by house type by household
size, or by number of vehicles, etc. This is then linked to the financial and non-financial
variables used in the model. Unlike PV systems, which are mostly first-time purchases,
potential EV consumers face the decision of how long to wait for improved next-generation
technologies, thus creating a demand for dynamic diffusion models (Kim et al., 2005). An
important consideration for applying a diffusion model to EVs is the incorporation of multiple
financial and non-financial benefits, as well as the evolution of the demographics, building
stock or market size across the landscape. For the case of a single product, Higgins et al.
(2012) capture these considerations by integrating multicriteria analysis with an extension of
the Bass diffusion model. While the model showed geographical differences in forecast
adoption, it used the national forecast by Graham et al. (2009) to calibrate for an aggregated
forecast for Victoria to 2032.
To understand the impact of vehicle adoption on load forecasts, Paevere et al. (2014) models
the energy demand by time of day for each ABS SA1 geographical unit of housing using a
household energy model (Ren et al., 2012), as well as a range of information on travel
Page 255
patterns (VISTA 2009). The paper showed the sensitivity of peak loads to off-peak charging
and vehicle-to-grid scenarios.
Overall, the biggest limitation to understanding the impacts of future EV adoption on
electricity grid loads is the ability to develop a reliable aggregated forecast. The two national
level studies outlined above showed major differences to 2020. EV forecasting relies on
assumptions on vehicle/battery price projections, future vehicle options available, and their
features and energy costs. EV forecasts are sensitive to uncertainty in these assumptions,
regardless of technical methodology used. A large range of non-financial drivers that affect
the type of EV a consumer would buy (and when) are still not well understood. Studies such as
Gardner et al. (2011) and Axsen and Kurani (2010) have used surveys, focus groups and trials
to help better understand the consumer perceptions and social interactions to purchasing an
EV in the future. The research challenge is how to use this type of information in various
methodologies to improve forecasts of EV adoption at a national level.
Technology forecasting — non-residential sector
Commercial buildings contributed to about 4.7% of gross energy consumption (or

277 petajoules) in Australia in 2009 (ABARES 2011), with stand-alone offices representing a
25% share of this gross consumption (DCCEE 2012). Office buildings have the highest energy
savings opportunity compared with other commercial stock, with an estimated total energy
reduction opportunity of 5142 GWh by 2020 (Climate Works 2010). The largest opportunities
are in rationalisation, insulation, heating, ventilation and air conditioning, lighting and
electronics. However, retrofitting existing buildings is more difficult than creating new green
buildings, and requires the cooperation of a wide range of stakeholders such as landlords,
tenants, contractors, local and state government. Also, while the technology may be
available, issues such as cost and demand from landlords, tenants and policy makers will
determine the priority and timeframe in which existing buildings are retrofitted (Miller and
Buys, 2008).
A few studies have simulated future energy consumption in office buildings under an assumed
behaviour (e.g. replace at end of life) or scenarios of when retrofits are likely to take place
(Coffey et al. 2010). Foliente and Seo (2012) forecast energy consumption from commercial
office building in Australia and model different reduction opportunities, including solar PV.
However, uptake of retrofit options varies depending on social, financial and geographical
conditions.
Estimating the likely uptake of retrofit packages across building stock over time is currently
very difficult. Tools are needed to accommodate the:
 complex diffusion process of adoption over different time steps by spatial scales,
demographics and building types
 variables that influence landlord and tenant actions in different locations, such as
demographics, socio-economic indicators, income and debt levels
 evolution of building stock and demographics in different regions or precincts
 wide range of office building materials, sizes and usages, including mixed usage
 different incentive requirements for responsibilities for landlords versus tenants
Page 256
o some retrofits are the responsibility of the landlord depending on the
conditions of the lease agreement
o in some situations, retrofits are dismantled by the tenant at the end of the
lease
 soft and hard barriers to adoption, which vary with retrofit package; the main barrier
categories (Marquez et al. 2012) are
o capital constraints and investment priorities
o capital and implementation costs
o market structures and supply constraints
o regulatory structure and supply constraints
o information gaps
o workforce and skill barriers
 electricity metering in some buildings, which can make it difficult to distribute cost
savings to tenants to encourage adoption of energy-efficient retrofit.
The Australian residential sector has highly spatially granular ABS data on building stock and
demographics, but a similar dataset does not exist for commercial buildings. The most
reliable disaggregated datasets are by local government area, with a breakdown of building
stock (including floor area) into floor height and age categories.
The range of building electricity usage patterns is much higher than in residential buildings,
and ground-up methods to electricity modelling are more difficult. As a result, there has been
very limited research forecasting uptake of appliances and retrofit in commercial buildings
that affect electricity usage. Higgins et al (2013) developed and implemented a diffusion
model to forecast uptake of different energy-efficient retrofits in office buildings, with
application to New South Wales office buildings. One of the biggest limitations was the lack of
information on existing retrofits across office building stocks, particularly when disaggregated
below state level.
References
[1] ABARES. (2011). Energy in Australia 2011. Australian Bureau of Agricultural and Resource
Economics.
[2] Achanta R. (2012). Long term electric load forecasting using neural networks and support
vector machines. IJCST,3, 266-269.
[3] AECOM (2010). Forecast uptake and economic evaluation of electric vehicles in Victoria.
Report 60149263 submitted to the Victorian Department of Transport.
[4] AEMO (2012). Solar PV Forecast for AEMO 2012-2022. Report by Sunwiz Consulting for
Australian Energy Market Operator.
[5] AEMO (2015). Emerging Technologies Information Paper. National Electricity Forecasting
Report. Available at www.aemo.com.au.
[6] ARENA (2013). Drivers of Domestic PV Uptake. Report by ACIL Allen Consulting for the
Australian Renewable Energy Agency.
[7] Axsen J, Kurani K. (2010). Interpersonal influence in the early plug-in hybrid market:
Observing social interactions with an exploratory multi-method approach.
Transportation Research D, 16, 150-159.
Page 257
[8] Aydinalp M, Ugursal V I, Fung AS. (2003). Modelling of residential energy consumption at
the national level. International Journal of Energy Research, 27, 441-453.
[9] Bartels R, Fiebig DG, van Soest A. (2006). Consumers and Experts: An Econometric Analysis
of the Demand for Water Heaters. Empirical Economics, 31(2), 369-391.
[10] Bartels, R, Fiebig DG, McCabe A. (2004). The value of using stated preference methods: a
case study in modelling water heater choices. Mathematics and Computers in
Simulation 64, 487-495.
[11] Bass FM. (1969). A new product growth model for consumer durables. Management
Science, 15(5), 215–217.
[12] Bass FM, Krishnan TV, Jain DC. (1994). Why the Bass model fits without decision
variables. Mark Science, 13, 203–273.
[13] Bhat CR, Guo JY. (2004). A Mixed Spatially Correlated Logit Model: Formulation and
Application to Residential Choice Modeling. Transportation Research Part B, 38, 147-
168.
[14] Boulaire F, Higgins A, Foliente G, McNamara C. (2014). Statistical modelling of district
level residential energy use in NSW, Australia. Sustainability Science, 9, 77–88.
[15] Climate Works. (2010). Australian Carbon Trust Report: Commercial Buildings Emissions
Reduction Opportunities. Climate Works Australia.
[16] Coffey B, Borgeson S, Selkowitz S, Apte J, Mathew P, Haves P. (2010). Towards a very
low energy building stock: Modelling the U.S. Commercial building sector to support
policy and innovation planning. Lawrence Berkeley National Laboratory Paper LBNL-
2543E.
[17] Couture T, Gagnon Y. (2010). An analysis of feed-in tariff remuneration models:
Implications for renewable energy investment. Energy Policy, 38, 955-965.
[18] DCCEE, 2012. Baseline Energy Consumption and Greenhouse Gas Emissions in Commercial
Buildings in Australia—Part 1, Department of Climate Change and Energy Efficiency.
Commonwealth of Australia.
[19] Deloitte. (2014). Residential electricity tariff review. Report commissioned by the Energy
Supply Association of Australia. Deloitte Access Economics.
[20] Delsante A. (2005). Is the new generation of building rating software up to the task? – A
review of AccuRate. Paper presented at the ABCB Conference ‘Building Australia’s
Future’, Surfers Paradise, 11-15 September 2005.
[21] DEWHA (2008). Energy use in the Australian residential sector 1986-2020. Report
published by the Australian Government Department of the Environment, Water,
Heritage and the Arts.
[22] Drude L, Pereira LC, Ruther R. (2014). Photovoltaics (PV) and electric vehicle-to-grid
(V2G) strategies for peak demand reduction in urban regions in Brazil in a smart grid
environment. Renewable Energy, 68, 443-451.
[23] Farahbakhsh H, Ugursal VI, Fung AS. (1998). A residential end-use energy consumption
model for Canada. International Journal of Energy Research, 22, 1133-1143.
[24] Faruqui A, Sergici S. (2010). Household response to dynamic pricing of electricity: a
survey of 15 experiments. Journal of Regulatory Economics, 38, 193-225.
[25] Foliente G, Seo S. (2012). Modelling building stock energy use and carbon emission
scenarios. Smart and Sustainable Built Environment, 1(2), 118-138.
[26] Gardner J, Quezada G, Paevere P. (2011). Social study on attitudes, drivers and barriers
to the uptake of electric vehicles. CSIRO, Australia.
Page 258
[27] Graham P, Brinsmead T, Dunstall S, Ward J, Reedman L, Elgindy T, Gilmore J, Cutler N,
James G, Rai A, Hayward J. (2013). Modelling the Future Grid Forum scenarios. CSIRO.
[28] Greene AL, Patterson PD, Sing M, Li J. (2005). Feebates, rebates and gas-guzzler taxes: a
study of incentives for increased fuel economy. Energy Policy, 33, 757-775.
[29] Guidolin M, Mortarino C. (2010). Cross-country diffusion of photovoltaic systems:
modelling choices and forecasts for national adoption patterns. Technological
Forecasting & Social Change, 77, 279–296.
[30] Hayward J, Graham P. (2012) Australian Electricity Generation Technology Cost
Projections: Application of a Global and Local Learning Model. CSIRO, available at
http://www.csiro.au/~/media/CSIROau/Files/PDF/Australian electricity generation
technology cost projections.pdf.
[31] Hens H, Verbeeck G, Verdonck B. (2001). Impact of energy efficiency measures on the
CO2 emissions in the residential sector, a large scale analysis. Energy and Buildings,
33, 275-281.
[32] Herter K. (2007). Residential implementation of critical-peak pricing of electricity.
Energy Policy, 35, 2121-2130.
[33] Herter K., Wayland S. (2010). Residential response to critical-peak pricing of electricity:
California evidence. Energy, 35, 1561-1567.
[34] Higgins A, Syme M, McGregor J, Marquez L, Seo S. (2014). Forecasting Uptake of Retrofit
Packages in Office Building Stock Under Government Incentives. Energy Policy, 65,
501–511.
[35] Higgins A, Paevere P, Gardner J, Quezada G. (2012). Diffusion of competing vehicle
options across a landscape of heterogeneous consumers. Technological Forecasting &
Social Change, 79, 1399-1412.
[36] Higgins A, Foliente G, McNamara C. (2011). Modelling intervention options to reduce GHG
emissions in housing stock—a diffusion approach. Technological Forecasting & Social
Change, 78, 621–634.
[37] Higgins A, McNamara C, Foliente G. (2014a). Modelling the future uptake of solar photo
voltaics and water heaters under different government incentives. Technological
Forecasting & Social Change, 83, 142-155.
[38] Horne M, Jaccard M, Tiedemann K. (2005). Improving behavioural realism in hybrid-
economy models using discrete choice studies in personalised transportation decisions.
Energy Economics, 27, 59–77.
[39] Horsky D, Simon L. (1983). Advertising and the diffusion of new products. Marketing
Science, 2(1), 1-17.
[40] Howard B, Parshall L, Thompson J, Hammer S, Dickinson J, Modi V. (2012). Spatial
distribution of urban building energy consumption by end use. Energy and Buildings,
45, 141-151. doi: 10.1016/j.enbuild.2011.10.061.
[41] Hsiao C, Mountain DC, Illman KH. (1995). A Bayesian integration of end-use metering and
conditional demand analysis. Journal of Business & Economic Statistics, 13(3), 315-
326.
[42] Islam T, Meade N. (2013). The impact of attribute preferences on adoption timing: The
case of photo-voltaic (PV) solar cells for household electricity generation. Energy
Policy, 55, 521-530.
[43] Jaccard M, Dennis M. (2006). Estimating home energy decision parameters for a hybrid
energy-economy policy model. Environmental Modelling & Assessment, 11, 91-100.
Page 259
[44] Jarvinen J, Orten F, Nelson N. (2011). Electric Vehicles in the NEM: energy market and
policy implications. AGL Applied Economic and Policy Research. Working Paper 27.
[45] Jones PJ, Lannon S, Williams J. (2001). Modelling building energy use at urban scale.
Paper presented at the Seventh International IBPSA Conference, Rio de Janeiro, 13–15
August 2001.
[46] Jun DB, Kim J. (2011). A choice-based multi-product diffusion model incorporating
replacement demand. Technological Forecasting and Social Change, 78, 674-689.
[47] Kavgic M, Mavrogianni A, Mumovic D, Summerfield A, Stevanovic Z, Djurovic-Petrovic M.
(2010). A review of bottom-up building stock models for energy consumption in the
residential sector. Building and Environment, 45(7), 1683-1697. doi: 10.1016/
j.buildenv.2010.01.021.
[48] Kim W, Lee J, Kim T. (2005) Demand forecasting for multigenerational products
combining discrete choice and dynamics of diffusion under technological trajectories.
Technological Forecasting & Social Change, 72, 825-849.
[49] Lin Z, Greene D. (2010). Who will more likely buy PHEV: A detailed market segmentation
analysis. The 25th World Battery, Hybrid and Fuel Cell Electric Vehicle Symposium &
Exhibition, Shenzhen, China, 5–9 November 2010.
[50] Lins MPE, da Silva ACM, Rosa LP. (2002). Regional Variations in Energy Consumption of
Appliances: Conditional Demand Analysis Applied to Brazilian Households. Annals of
Operations Research, 117(1), 235-246. doi: 10.1023/a:1021533809914.
[51] Mansfield E. (1961). Technical change and the rate of imitation. Econometrica, 29, 741–
776.
[52] Marquez L, McGregor J, Syme M. (2012). Barriers to the Adoption of Energy Efficiency
Measures for Existing Commercial Buildings. CSIRO Energy Transformed Flagship.
[53] Mills, D. (2010). Greenhouse gas emissions from energy use in Queensland homes (S.
Innovation, Trans.): Department of Environmental and Resource Management,
Queensland Government.
[54] Murphy R, Rivers N, Jaccard M. (2007). Hybrid modelling of industrial energy consumption
and greenhouse gas emissions with an application to Canada. Energy Economics, 29,
826-846.
[55] Olson J, Choi S. (1985). A product-diffusion model incorporating repeat purchases.
Technological Forecasting and Social Change, 27, 385–397.
[56] Orbach Y, Fruchter GE. (2011). Forecasting sales and product evolution: The case of the
hybrid/electric car. Technological Forecasting and Social Change, 78, 1210-1226.
[57] Paevere P, Higgins A, Ren Z, Horn M, Grozev G, McNamara C. (2014). Spatial modelling of
electric vehicle charging demand and impacts on peak household electrical demand.
Sustainability Science, 9 (1), 61-76.
[58] Ren Z, Foliente G, Chan W, Chen D, Syme M. (2011). AusZEH Design: Software for low-
emission and zero-emission house design in Australia. In: Proceedings of Building
Simulation 2011: 12th Conference of International Building Performance Simulation
Association, Sydney, 14-16 November 2011.
[59] Ren Z, Foliente G, Chan W, Chen D, Ambrose M, Paevere P. (2013). A model for
predicting household end-use energy consumption and greenhouse gas emissions in
Australia. International Journal of Sustainable Building Technology and Urban
Development, 4, 210-228.
Page 260
[60] Ren Z, Paevere P, McNamara C. (2012). A local-community-level, physically-based model
of end-use energy consumption by Australian housing stock. Energy Policy, 49, 586-
596.
[61] Revelt D, Train K. (1998). Mixed logit with repeated choices: household’s choices of
appliance efficiency level. Reviews of Economics and Statistics, 80(4), 647-657.
[62] Rogers E. (1962). Diffusion of Innovations. Glencoe: Free Press. ISBN 0612628434.
[63] Rudolf V, Papastergiou KD. (2013). Financial analysis of utility scale photovoltaic plants
with battery energy storage. Energy Policy, 63, 139-146.
[64] Sener I, Pendyala R, Bhat C. (2009). Accommodating spatial correlation across choice
alternatives in discrete choice models: Application to modelling residential location
choice behaviour. Transportation Research Board Annual Meeting, Paper #09-2281.
[65] Shafiei E. et al (2012). An agent-based modeling approach to predict the evolution of
market share of electric vehicles: A case study from Iceland. Technological
Forecasting & Social Change, 79, 1638-1653.
[66] Shimoda Y, Fujii T, Morikawa T, Mizuno M. (2004). Residential end-use energy simulation
at a city scale. Building and Environment, 39, 959-967.
[67] Sijm JPM. (2002). The performance of feed-in-tariffs to promote renewable electricity in
European countries. ECN-C-02-083.
[68] Soderholn P, Klassen G. (2007). Wind power in Europe: A simultaneous innovation-
diffusion model. Environmental & Resource Economics 36, 163-190.
[69] Swan LG, Ugursal VI. (2009). Modeling of end-use energy consumption in the residential
sector: A review of modeling techniques. Renewable & Sustainable Energy Reviews,
13(8), 1819-1835. doi: 10.1016/j.rser.2008.09.033
[70] Swan L, Ugursal VI, Beasuoleil-Morrison I. (2009). Implementation of a Canadian
Residential Energy End-Use Model for Assessing New Technology Impacts. Paper
presented at the Eleventh International IBPSA conference, Glasgow, Scotland,
27-30 July 2009.
[71] VISTA (2009). Victorian Integrated Survey of Travel and Activity, Victorian Department of
Transport. Available at www.transport.vic.gov.au.
[72] Wang CH, Grozev G, Seo S. (2012). Decomposition and statistical analysis for regional
electricity demand forecasting. Energy, 41(1), 313-325.
[73] Won J, Yoon Y, Lee K. (2009). Prediction of electricity demand due to PHEVs distribution
by Korea by using diffusion model. IEEE Transmission & Distribution Conference &
Exposition: Asia and Pacific, 26-30 October, 2009.
[74] Yamaguchi Y, Akai K, Shen J, Fujimura N, Shimoda Y, Saijo D. (2013) Prediction of
photovoltaic and solar water heater diffusion and evaluation of promotion policies on
the basis of consumers choices. Applied Energy, 102, 1148-1159.
Page 261
Appendix C – Maximising network benefits:
overview of relevant literature
This appendix presents a brief description of the aim and method for the references reviewed
in the report. Each reference is tagged with the relative embedded generation (EG)
technology. Even though a reference may use only one particular EG technology for its study,
the findings of the study are generally applicable across different EG technologies. All EG
technologies can provide power and voltage levelling, frequency support, are dispatchable
and can defer infrastructure upgrades. In the case of uncontrollable energy sources, such as
PV and wind, some form of storage may also be required. The effectiveness of each EG
technology in providing these benefits and the technical complexity of operation will differ
between different EG technologies, as will the cost.
Photovoltaics and storage
Transmission and distribution deferral using photovoltaics and energy storage

Reference [4] examines real data for a number of distribution networks in Salt Lake City. The
data consisted of 15-minute load and solar resource data. A load growth rate of 4% is assumed
and substations examined are rated around 10 MVA. Photovoltaic (PV) generation profiles
were generated using solar resource data. The paper investigates how PV reduces peak load,
both commercial and residential. The reduction in peak load reduces the number of overload
hours experienced by the transformer. In this system, a transformer is defined as requiring
replacement once the number of overload hours for the year is greater than 1%. The increase
in the number of years before replacement is required due to PV reducing peak load,
compared with no PV, is the number of deferred years. The paper does not define the PV
penetration level. The study finds that for a projected 2016 load, the number of hours that a
substation transformer is overloaded decreases from 510 to 179 for 10% penetration and to 35
for 20% penetration. This defers the transformer upgrade by two and four years, respectively.
Mitigation of rooftop solar photovoltaic impacts and evening peak support by

managing available capacity of distributed energy storage systems
Reference [5] proposes a new control method for managing the charging/discharging of a
battery in a battery/PV system. The system is shown to reduce fluctuations in voltage due to
PV and reduce voltage rise during midday. Redirecting PV real power to charge the battery
also mitigates voltage rise, while energy stored in the battery during the day can then be
used to reduce peak load in the evening. A simplified description of the charging/discharging
approach is that charging is gradually increased from low levels in the morning to peak
charging around midday. Charging is then gradually decreased from peak at midday into the
afternoon; this is essentially matching the charging target profile to the expected PV
generation profile. Battery discharging begins at low levels at the start of the evening peak
and gradually increases into the peak of the evening load period. The results show a reduction
Appendix C – Maximising network benefits: overview of relevant literature | Page 262

in reverse power flow during the PV peak generation period, as well as evening peak-load
support from battery discharging.
Sizing strategy of distributed battery storage system with high penetration of

photovoltaic for voltage regulation and peak-load shaving
Reference [6] presents a battery sizing strategy to obtain the most economically feasible
battery energy storage system (BESS) size. The sizing is determined by optimising the cost
benefit of the BESS. The cost of the BESS is weighed against the following operational cost
reductions: peak demand, workload of on-load tap changers (OLTCs) and step voltage
regulators (SVRs), and peaking power generation. Battery life is also taken into account. The
results show a reduction in OLTC and SVR operation as well as peaking power operation.
Despite these reductions, the authors find that no battery/PV investment brings about an
economically feasible investment.
Coordinated control of grid-connected photovoltaic reactive power and battery

energy storage systems to improve the voltage profile of a residential distribution
feeder
Paper [8] examines the potential for PV combined with storage to improve the voltage profile
of residential feeders. The improvement is to include both the mitigation of voltage rise due
to PV and voltage drop during peak demand periods. All EG devices are in communication and
control is coordinated. Control functions include reactive power support, and real power
support through charging/discharging of battery storage. The results show the change in
voltage profile for urban and rural cases when only reactive power support is provided. With
only reactive power support, breaches of the lower voltage limit still occur in the rural case.
When real power from the battery is also injected into the feeder, voltage levels remain
above the lower voltage limit.
Frequency support functions in large photovoltaic power plants with active power
reserves
The European Network of Transmission System Operators for Electricity is proposing that
large-scale, centralised PV (LSCPV) plants provide frequency support functions [9] (Figure 57).
It suggests that PV plants inherit some of the features provided by synchronous-based power
plants, including inertial response, synchronising power, primary frequency control,
secondary frequency control, and the provision of active power reserve and reactive power
reserve. To be able to provide frequency support, PV plants would need to operate in
frequency-sensitive mode (see Section 3.4.6). To accommodate frequency-sensitive mode,
the PV plant would either have to be integrated with storage or operate under their maximum
available capacity, or create the reserve using the curtailed power. The LSCPV plant model
frequency support efficacy, frequency-sensitive mode and inertial response is tested by
initiating a frequency event (major loss of generation) on the Institute of Electrical and
Electronics Engineers (IEEE) system. For the two test scenarios, with the LSCPV plant
providing frequency-sensitive mode and inertial response support, the system recovered more
quickly. A daily load profile was also applied to the system and the frequency deviation across
the day was recorded. The frequency distribution was shown to deviate less when LSCPV plant
frequency support was present.
Figure 57. European Network of Transmission System Operators for Electricity requirement for frequency-sensitive
mode for large-scale, centralised photovoltaic plant [9]
This work tests the concept of LSCPV providing frequency support functionality, frequency-
sensitive mode and inertial response. A model of an LSCPV plant, including controls to provide
the frequency support, is presented.
Network impacts of high penetration of photovoltaic solar power systems

Reference [10] presents a breakdown of the potential grid benefits offered by integrating PV,
BESS and a static synchronous compensator (STATCOM). The benefits include:
 MW and MVAR (volt-amps reactive) support during critical contingencies
 voltage regulation support
 low-voltage ride-through during faults, provided by PV inverters
 ancillary services
 postponement of transmission upgrades in constrained circuits.
Rule-based control of battery energy storage for dispatching intermittent renewable

sources
The work presented in paper [12] is similar to [11]; the main difference is that a BESS is used
instead of a hybrid energy storage system. A control strategy is presented that minimises the
power fluctuations of variable renewable EG (wind or solar) to make its power dispatchable.

The effectiveness of this control strategy and BESS is tested using an actual PV system and
wind farm data. For a 1.5 MW-PV system, the BESS reduces the percentage of undesirable
deviations (anything greater than 100 kW, positive or negative) from 16 to 4%.
Impact of high photovoltaic penetration on distribution transformer insulation life

Paper [13] estimates the extension in operating life for a typical distribution transformer
(DTx) (200-kVA, 22/0.415-kV) due to rooftop PV generation. The transformer supplies a
suburban area with a high penetration of rooftop PV systems. Using one year of residential
load data drawn from the Perth Solar City High Penetration PV Trial, the insulation ageing
impact is estimated. Each phase is analysed to investigate the impact of imbalanced load
across phases on transformer loss-of-life (LOL). One year of ambient temperature data is used
as an input into the LOL model of the transformer. The test scenario shows the saving in LOL
is 0.2, 14 and 160 days for phases A, B and C, respectively.
A vision and strategy for deployment of energy storage in electric utilities

In 2005, American Electric Power (AEP) pushed energy storage to the distribution level in the
form of sodium sulfur batteries located at distribution substations. Storage sizes are 1, 2 and
4 MW, with 7-hour discharge capability. AEP is now developing and planning to install smaller,
more broadly distributed, energy storage units for fringe-of-grid customers ([14]). Termed
community energy storage, the units will be 25 kW operated as fleets to provide aggregate
benefits at the MW scale using existing communication and control technologies. AEP found
that power system reliability was improved when storage is pushed further out on the system.
Diesel, gas, steam and hybrid generation systems
Quantification of the distribution transformer life extension value of distributed

generation
Paper [15] presents a method for quantifying the economic benefits of DTx life extension
through EG. The proposed approach is applied to actual distribution transformers installed in
five sample cities in Iran to provide realistic estimates of the benefits in monetary terms. The
types of EG considered in the study include PV, wind, combined heat and power (CHP) and
microturbines. The results give the life extension and estimated economic value (US$/year)
for each city for each EG type.
Impact of distributed generation on distribution investment deferral

Paper [16] assesses the impact of EG on distribution network investment deferral in the long
term. Due to the randomness of the variables that affect such matters (e.g. load demand
patterns, EG hourly energy production, EG availability), a probabilistic approach using a
Monte Carlo simulation is adopted. Several scenarios characterised by different EG
penetration and concentration levels and EG technology mixes, are analysed. EG types

include PV, wind and CHP. For several test scenarios, the paper calculates the increase in
admissible load growth due to EG. For one particular scenario, the results show wind to be
the worst performer, with the other technologies performing equally well, providing an
increase in admissible load growth of nearly 250% at 100% penetration.
Distribution transformer loss-of-life reduction by increasing penetration of distributed

generation
The method described to calculate the reduction in LOL in a DTx due to EG in [15] is first
presented in this paper [17]. It is applied to an actual distribution system in Iran using
different combinations of EG including PV, wind, CHP and microturbines, half PV/Wind,
PV/CHP and wind/microturbine. The results show the reduction in LOL rate over the life of
the DTx; one figure plots the annual LOL rate for each technology and combination of
technology against penetration level. Relative to the reference LOL rate, the best result is
achieved through the microturbine at 15% penetration. The LOL rate is reduced from 1.2 to
0.15, an 87.5% reduction.
Evaluation of investment deferral resulting from micro-generation for extra-high-

voltage distribution networks
In paper [18], CHP is deployed at different locations for four different penetration levels to
assess the deferral in investment for a practical United Kingdom extra-high-voltage (EHV)
network (33—132 kV). The savings associated with investment deferral depend on the length
of time before which an asset needs upgrading. The results provide the reduction in
investment as a percentage of the required investment if no EGs were installed. Savings were
calculated for EG evenly distributed and allocated in proportion to load, and penetration
percentage is the installed EG capacity as a percentage of the total load. The savings
calculated were large; up to 82% reduction for a 32% penetration level.
Adequacy and security evaluation of distribution systems with distributed generation

Paper [19] calculates the increase in network reliability delivered by the installation of EG in
rural Brazil. The rural network consists of conventional generation with EG distributed among
subsystems. The paper investigates how effectively EG maintains subsystem voltage and
frequency after either i) a change in state of the EG within the subsystem, ii) a load state
transition, or iii) a protection device event. These events are termed islanding. Its
effectiveness is measured by comparing the System Average Interruption Duration Index
(SAIDI) and System Average Interruption Frequency Index (SAIFI), with EG and without. The
paper finds that EG does improve reliability within the rural network. The frequency and
duration of inadequate services for the whole system dropped from 3.6 occurances/year and
2 h/year to 0.18 occ./year and 0.0978 h/year, respectively. There was also an average of 1.4
islanding events a year, of which 92% were successful.

Embedded Generation Project - Final Report v1

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Embedded Generation Project - Final Report v1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Embedded Generation Project - Final Report v1

Uploaded by

Copyright:

Available Formats

Embedded Generation Project –

Energy Networks Association

2 Scope and approach.............................................. 17

3 Technical impact assessment ................................... 20

4 Regulatory impact assessment ................................ 160

5 Commercial impact assessment ............................... 168

6 Policy and regulatory options ................................. 213

7 Recommendations ............................................... 227

Appendix A – Survey responses ......................................... 237

Appendix B – Analysis of embedded generation/distributed energy resources

Appendix C – Maximising network benefits: overview of relevant literature

Key Technical Findings

Technical Constraints and Network Impacts

operation of protection and voltage regulation devices. In islanded networks, such as

Options for Mitigating Network Impacts and Risks

8. Improved network load prediction techniques that incorporate increased penetration

Options for Maximising Network Benefits

Australian Standards Gap Analysis

commissioning; increased cost and reduced consistency of type-testing; and potentially

Key Regulatory and Commercial Findings

The key findings are categorised below as either regulatory or commercial.

Key Regulatory Findings

Enabling Technologies and Pricing

 Working closely with a range of industry stakeholders (including EG proponents, the

o discounting connection charges, network charges or providing direct payment,

o implementing price signals or incentives for customers to undertake efficient

strategic deployment of medium-scale power electronics solutions, such as STATCOMS, at

 Supporting a national approach to cost-reflective pricing that removes inconsistent

 Identifying and supporting opportunities to restructure current incentives and subsidies so

 Investigating opportunities for the development of cost-effective policies and incentives

 Supporting the development of a new Australian Standard to manage the functionality of

 Supporting the development of standards relating to the safety of small-scale and

Enabling Business Models

There should be no inherent advantage or disadvantage for network service providers in

 Supporting changes to current state-based ring-fencing rules to create a national

 Supporting a review of relevant rules to ensure no barriers exist to the creation,

There should be no inherent advantage or disadvantage for network service providers to

Critical Further Work

Abbreviations used in this report

VRB vanadium redox battery

Source: Australian PV Institute

2 Scope and approach

2.3 Document maps

Document map key

The impact, constraint, mitigation option or opportunity applies at the customer

The impact, constraint, mitigation option or opportunity is relevant for the

The impact, constraint, mitigation option or opportunity is relevant for networks

The impact, constraint, mitigation option or opportunity is relevant for networks

The impact, constraint, mitigation option or opportunity is relevant for microgrids,

The impact, constraint, mitigation option or opportunity is of low significance given

The impact, constraint, mitigation option or opportunity is of moderate significance

The impact, constraint, mitigation option or opportunity is of high significance

The impact, constraint, mitigation option or opportunity is likely to be highly

The impact, constraint, mitigation option or opportunity is likely to be highly

The impact, constraint, mitigation option or opportunity is unlikely to be relevant

3 Technical impact assessment

3.2 Technical constraints and network impacts

Technical impact assessment | 24