Energy Reports 5 (2019) 310–326

Contents lists available at ScienceDirect

Energy Reports
journal homepage: www.elsevier.com/locate/egyr

Research paper

Integrated Economic Adoption Model for residential grid-connected

photovoltaic systems: An Australian case study
∗
Omar Ellabban , Abdulrahman Alassi
Iberdrola Innovation Middle East, Qatar Science & Technology Park, Tech 1 Building, Office 204, PO Box 210177, Doha, Qatar

highlights

• EAM for distributed PV systems has been implemented based on real data.
• ToU-NB-BB scheme achieves the maximum saving compared with other metering schemes.
• 89% probability of getting PV-LCOE less than the retail rate has been achieved.

article info a b s t r a c t

Article history: Photovoltaic (PV) systems, like most renewable energy resources, are characterized by large initial
Received 23 October 2018 capital investments by the energy user. The high upfront cost (for which many residential customers
Received in revised form 23 January 2019 are unwilling to take risks) is and will remain one of the main obstacles that need to be tackled in
Accepted 11 February 2019
order to achieve a faster and a greater deployment of the distributed PV technology in the residential
Available online xxxx
sector as a local generation source. Understanding the economics of the residential PV installation is a
Keywords: key aspect to determine whether the financial investment in this sector is economically feasible and
Economic Adoption Model reasonable or not. Thus, it is essential to conduct a comprehensive economic analysis to encourage
Sensitivity analysis residential consumers to be PV adopters. The main purpose of this work is to present an Integrated
Profitability analysis Economic Adoption Model (IEAM) for distributed PV systems. This IEAM is composed of four modules:
Monte Carlo analysis
(1) economic assessment of four different proposed electricity retail tariff structures; (2) profitability
Feed in tariff
analysis of these proposed electricity retail tariff structures based on the most common economic
Net metering
indicators: Net Present Value (NPV), Internal Rate of Return (IRR), Discounted Payback Period (DPBP),
and Profitability Index (PI); (3) sensitivity analysis of different variables which strongly affect the PV
system profitability; and finally, (4) Levelized Cost of Electricity (LCOE) probability distribution based
on the Monte Carlo Analysis (MCA). The model is implemented using Microsoft Excel spreadsheet
based on a real dataset for one year of PV generation and residential load consumption for 54
customers in the Ausgrid’s electricity network, New South Wales (NSW), Australia (as case study)
under the current market arrangements and PV support policies. The model can be easily adapted to
any location by changing the generation and load profiles and the relevant economic assumptions and
electricity regulatory policies. The outcomes of this IEAM should guide the decision makers in different
distribution networks in designing their retail tariff structure and PV support policies and to conduct
economic feasibility analysis for residential and commercial PV projects with different sizes in different
locations within their networks. Also, the IEAM answers the investment profitability questions for the
household owners under real integrated system uncertainty.
© 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction energy to drive up the predicted global energy demand by about

25% by the year 2040. The electricity demand is expected to
Energy is no longer a luxury but a necessary commodity for increase by 60% between 2015 and 2040, representing 55% of the
the modern life. The global economic growth coupled with pop-
global energy demand growth. Therefore, the global electricity
ulation increase, mainly in developing countries; require more
generation capacity is expected to increase by 93% between 2010
∗ Corresponding author. and 2040 (Anon, 2018b). Shifting to renewable energy resources
E-mail addresses: oellabban@iberdrola.com (O. Ellabban),
can help the globe to meet the dual goals of reducing greenhouse
aalassi@iberdrola.com (A. Alassi). gas (GHG) emissions, thereby limiting future extreme weather

https://doi.org/10.1016/j.egyr.2019.02.004
2352-4847/© 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326 311

that, the costs and benefits of the PV systems are distributed

Abbreviations among different industry participants including customers (PV
BN-S Net Billing Simple owners), retailers and network service providers. Furthermore, PV
DPBP Discounted Payback Period based energy trading may as well add considerable complexity in
financial transfers between industry participants.
DPPs Dynamic Pricing Plans
In the literature, many authors have explored the techno-
FiT feed-in tariff
economic feasibility analysis for a PV system. Therefore, differ-
FR Flat Rate ent techno-economic feasibility studies for a variety of solar PV
FR-NB-BB Flat Rate-Net Billing with Buy Back systems had been conducted in different regions and countries
FR-NB-S Flat Rate-Net Billing Simple (Pillaia and Naserb, 2018; Audenaert et al., 2010; Camilo et al.,
GHG Greenhouse Gas 2017; Lomas et al., 2018; Lee et al., 2017; Haegermark et al., 2017;
IEAM Integrated Economic Adoption Model Li et al., 2017; Okoye and Oranekwu-Okoye, 2018; Ayadi et al.,
IRR Internal Rate of Return 2018; Tarigan et al., 2015; Kazem et al., 2017; Edalati et al., 2016;
LCOE Levelized Cost Of Electricity Kumar, 2015; Koumparou et al., 2017; Honrubia-Escribano et al.,
MCA Monte Carlo Analysis 2018; Akter et al., 2017; Chapman et al., 2016). These studies
tackled the economic assessment analysis of PV systems from
NB Net Billing
various viewpoints such as, profitability analysis, metering and
NB-BB Net Billing with Buy Back
feed-in tariffs (FiTs) schemes and PV support polices.
NM Net Metering
In Pillaia and Naserb (2018), the economic performance of
NPV Net Present Value a 1 MW grid-connected PV system is analyzed based on the
NSW New South Wales levelised cost of electricity (LCOE), net present value (NPV), pay-
PI Profitability Index back period (PBP) and energy payback time (EPBT) calculations
PR Performance Ratio with a conclusion that the LCOE for the PV system was 43% less
PV Photovoltaic than the present actual cost of a kWh in Bahrain. In Audenaert
RET Renewable Energy Target et al. (2010), using a cash flow projection method, the economic
SBS Solar Bonus Scheme evaluation of a grid connected PV systems (GCPVS) in Flanders
ToU Time of Use is analyzed based on NPV, internal rate of return (IRR), PBP,
discounted payback period (DPBP), profitability index (PI), yield
ToU-NB-BB Time of Use-Net Billing with Buy Back
unit cost, yield unit return and break-even turnkey cost with a
ToU-NB-S Time of Use-Net Billing Simple
conclusion that, it is financially responsible to invest in a GCPVS
for a Flemish company. In Camilo et al. (2017), an economic
assessment of residential PV systems with different configuration
in Portugal is made using the NPV, IRR, PI and DPBP and the
and climate impacts, and ensuring reliable, timely, and cost-
LCOE is calculated in each case. And if the results point to non-
efficient delivery of energy, therefore, investing in renewable
profitable solutions, the required reduction in the investment cost
energies can have significant advantages for world future energy
to make the project a viable one is computed. Lomas et al. (2018)
security, (Ellabban et al., 2014).
undertakes a detailed analysis of the energy and economic perfor-
The global renewable energies consumption is expected to
mance of a PV plant in Spain (based on NPV and IRR) in order to
increase annually by 2.3% between 2015 and 2040 representing
assess the influence of the promotion policies on the profitability
(under most optimistic predictions) 50%–60% of global electricity
of this investment. The conclusion of this work is, although the
consumption by 2050 (Kuo and Pan, 2018). Solar Photovoltaic
(PV) power is surging on the back of scaled-up production and project is economically viable; it is financially unfeasible due to
continuously falling costs. The amount of new solar power ca- a negative cumulative annual liquidity as a result of the recent
pacity has experienced remarkable growth across the globe, led governmental actions. In Lee et al. (2017), a life cycle cost analy-
by China, the United States and Japan. Technological learning sis of a residential PV system in the U.S is conducted to establish
as well as economics of scale have reduced costs throughout the minimum rate for purchasing electricity generated from a PV
the PV value chain. Prices continue to fall; meaning power from system to achieve the target payback period and to improve the
new solar plants is now cheaper than new coal plants in several solar incentives policies. Haegermark et al. (2017) presents an
locations. During 2016, at least 75 GW of solar PV capacity was economic feasibility study (based on NPV and PI) of PV systems
added worldwide, making the global solar PV capacity at least in Swedish multifamily buildings including, a tax rebate and an
303 GW, (Anon, 2017b) and (Anon, 2017c). The PV market was investment subsidy. The study showed that both the investment
multiplied by almost 40 in 10 years and the solar PV capacity subsidy and tax rebate have a large effect on profitability and
which has been installed in 2016 (up 48% over 2015) is more sizing of PV systems. Furthermore, the current Swedish incentives
than the cumulative world capacity five years earlier. Solar energy make the viability of PV rooftop systems difficult to predict.
costs have fallen 58% in five years (2010–2015) and are expected Li et al. (2017) presents a techno-economic assessment model
to continue to fall by additional 40%–70% by 2040, (IRENA, 2016) (based on LCOE, PBP and PI) of grid-connected PV systems at
and (International Energy Agency, 2016). Furthermore, in 2016, Northeast England for different PV technologies (monocrystalline
the global investment in solar energy was USD 112.4 billion, (s-Si), polycrystalline (p-Si) and Cadmium telluride (CdTe)). The
down about a third compared to 2015 investment in large part study concludes that, the p-Si system has the best performance.
due to falling costs of production. Investment in solar energy In Okoye and Oranekwu-Okoye (2018), a generalized economic
made up about 46% of all investment in renewable energies for model (based on NPV) is developed to assess the cost-benefit of
2016 (Frankfurt School-UNEPCentre/BNEF, 2017). an off-grid PV system in Sub-Sahara Africa. The study concludes
Investigating the economics of the distributed PV system is that, the PV technology is eligible for project financing as it
a highly complex exercise given the intermittent nature of PV can repay its loan within the specified time under the current
generation, location and generation and consumption behavior, energy policies in Nigeria. In Ayadi et al. (2018), an economic
in addition to a wide range of economic variables, a variety of assessment (based on IRR and PBP) is carried out for a proposed
support polices and different metering arrangements. Adding to 15 MW grid-connected PV system at a University of Jordan to
312 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326

cover the institute electricity needs, the IRR for this choice was prediction methodology has been developed and tested on the
32% and the payback period is 3 years. Tarigan et al. (2015) Australian market by the authors of this work In Alassi et al.
simulates a techno-economic feasibility study of installing a grid- (2018). The cost distributions and outlook of a distributed PV
connected PV system in a typical residential in Surabaya. This system are mainly attributed to country-specific developments
study concludes that, without any financial support from the such as local demand, local economy indicators (such as: nom-
government, a grid-connected PV system is not financially viable inal discount rate, interest rate and inflation rate), the level of
to meet the entire electrical need of a typical residential house in market maturity and local experience and incentives (subsidy
Surabaya. In Kazem et al. (2017), a techno-economic feasibility increment/decrement). Therefore, it is very difficult to give a
analysis of 1 MW PV grid-connected system in Oman is presented. value range of the used economic indicators.
The assessment results show that, the PV system is economically On the other hand, by reviewing the literature, one can find
feasible and shows great promise for the city of Adam. In Edalati many similar techno-economic studies, based on the similar eco-
et al. (2016), a technical and economic assessments (based on nomic indicators used in this study with similar methodology,
NPV, PBP, IRR and LCOE) of grid-connected PV power plants which has been conducted for a variety of energy resources
in Iran has been conducted based on recent a statistical data and energy storage devices, such as: tidal energy, wind energy,
and not on real nor experimental projects. The study concludes nuclear energy, fuel cells and batteries, (Segura et al., 2017; Cali
that, the LCOE for PV power plants is higher than the average et al., 2018; Kharitonov and Kosterin, 2017; Nelson et al., 2005;
selling price of the electricity; therefore, the solar PV systems may Löbberding and Madlener, 2019) and (Masebinu et al., 2017).
reach the grid-parity only for peak prices. In Kumar (2015), an Segura et al. (2017) presents an economic assessment for tidal
economic assessment models (based on NPV, PBP and IRR) for a energy technologies based on NPV, IRR, DPBP and LCOE. Cali et al.
residential PV system is conducted. The result of the assessment (2018), implements a techno-economic analysis for an offshore
screens the economic feasibility of PV systems. Koumparou et al. wind farm based on NPV, LCOE, DPBP, and IRR. Kharitonov and
(2017), present a methodology to identify the appropriate net- Kosterin (2017) investigates the investment performance criteria
metering scheme for the Mediterranean region (Cyprus, Greece, (NPV, LCOE, IRR, DPPB and basic engineering-economic parame-
Spain, Slovenia, Portugal and France) given the particularities and ters of nuclear reactors characterizing a Nuclear Power Plant prof-
local conditions and real system where flat rate electricity tariffs itability and competitiveness at the microeconomic level. An as-
is only considered. The main conclusion of this work is that, for sessment of the economic benefits of using grid-connected Fuel-
the time being, net-metering schemes should be considered as a Cell Systems is presented in Nelson et al. (2005) and Löbberding
transition policy from the FiT schemes towards policies closer to and Madlener (2019) based on NPV, payback period, benefit-to-
the market involving dynamic cost-reflective pricing. Honrubia- cost ratio and the cost of electricity. A techno-economic viabil-
Escribano et al. (2018) performs a comprehensive assessment ity and environmental benefits of using Battery Energy Storage
of the economic performance of a PV system (based on NPV, (BES) with three different technologies (lithium ion, lead acid and
IRR and PBP) with different topologies in seven EU countries sodium sulfur) has been implemented based on NPV, LCOE, IRR
with the largest PV development. The main conclusion of this and payback period in Masebinu et al. (2017).
work is that, the direct replication of PV systems that were Like other countries around the world, the international and
successful in a particular country should be avoided due to the domestic concerns regarding GHG emissions and energy secu-
large amount of parameters involved in the economic perfor- rity have encouraged the successive Australian governments to
mance assessment. Akter et al. (2017), proposes the inclusion increase the renewable energies share in the Australian power
of a generalized tariff structure in its economic evaluations of a generation mix replacing fossil fuel based generation. Since, the
residential building with solar photovoltaic and battery energy electricity generation accounting for 33% of Australia’s GHG emis-
storage systems, which allow residential consumers to choose the sions, Australia has committed to a Renewable Energy Target
right one based on their different perspectives of the investments. (RET) that delivers 20% of Australia’s energy needs by 2020.
Chapman et al. (2016) discusses the successes, failures and im- The RET aims to reduce GHG emissions from electricity gener-
pacts of Australian residential solar PV policy from 2001 to 2012 ation by providing certificates for renewable energy generation.
specially the different FiT levels, across two distinct types; gross Small systems (<100 kW) are provided with upfront payment
and net, with a conclusion that this potential inequity of these deductions, while large systems are provided with certificates
policies warrants further research. Net metering (NM) allows the based on MWh generated. The RET has played a key role in
PV system owners to offset their energy consumed within their facilitating Australia’s renewable energy deployment, it fueled the
premises with energy produced from their locally installed PV rapid deployment of household rooftop PV system installations
system and to sell surplus electricity to their electricity service across Australia, (Byrnes and Brown, 2015) and (Commonwealth
provider at a rate equal, less or greater than the retail rate Government of Australia, 2015). Additionally, feed-in tariffs (FiT)
according to the country and network operator policy. Different were introduced in 2008 by most Australian state governments
net metering (fully or partially) and Net Billing (NB) policies has to encourage residential sector and small businesses (<10 kW) to
been discussed in Christoforidis et al. (2016), Dufo-López and install solar PV technology, where, a household gets a premium
Bernal-Agustín (2015) and Comello and Reichelstein (2017). payment for the electricity supplied to the electricity grid, this
These studies addressed the profitability analysis of a PV sys- FiT can be a gross or a net FiT, (Zahedi, 2010) and (Poruschi
tem during its lifetime based on the common economic indicators et al., 2018). Moreover, Australia has the highest average solar
(NPV, PBP, DPBP, IRR and PI) based on estimated or average radiation per square meter of any continent in the world ranging
operating values, mostly without considering the real generation from 1500 to 1900 kWh/m2 /yr and more hours of sunlight per
and consumption data, real market economic parameters, differ- day, (Geoscience Australia and BREE, 2014), which leads to high
ent financing mechanisms, different tax schemes, the effect of PV electricity yield ranging from 1200–1600 kWh/kWp/yr (with
variation of different system parameters, different metering and an average 1460 kWh/kWp/yr) at Australia compared with other
retail tariffs schemes, in addition to the associated investment countries, (Talavera et al., 2010). As a result, Australia became one
risk and the expected monthly saving from such investment. of the top ten countries for total solar installed capacity in 2015
Furthermore, in order to understand the cost distribution and and this continued in 2016, (Anon, 2017c). By the end of 2017,
the cost outlook of the different components of a distributed Australia’s total solar PV capacity exceeded 6.4 GW with close
PV system in different countries, a generic market-based cost to 1.8 million installations, the amount of rooftop PV capacity
 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326 313

installed during 2017 is estimated to have reached over 1 GW, 2. Model description
(Anon, 2018a). The Australasian cumulative installed PV power,
from 2006–2016, is distributed among four sub-markets, where, This section contains a detailed description for the proposed
almost 90% of the cumulative installed capacity by 2016 comes IEAM to provide an integrated residential PV system economic
from the residential sector, (IEA-PVPS, 2017). The average number adoption model. The model should be able to: economically eval-
of Australian households with solar PV systems reached 18.7%, uate different retail tariffs under different metering schemes by
and exceed 30% in some states, (Stock et al., 2017). Solar PV elec- computing the PV system monthly saving for different tariff sce-
tricity systems produced about 2.6% of Australia’s total electricity narios; calculate the life time profitability of the residential PV
generation in 2016, while renewable electricity represents 16.3% system based on the traditional financial indicators (NPV, IRR,
PI and DPBP); compute the LCOE of the PV system; conduct
of the Australian total electricity generation (Jaeger-Waldau,
a sensitivity analysis for the variables affecting the PV system
2017). Moreover, the small scale PV installed capacity is expected
profitability the most, considering all the policy and regulatory
to surge to 44 GW by 2040 representing 31% of the Australian
issues in the distribution network and produce probabilistic dis-
power mix and the Australia’s electricity system will become one
tribution of its profitability based on MCA. The proposed IEAM
of the most decentralized in the world, with around 45% of power is structured into four different modules; Fig. 1 illustrates the
generating capacity located behind the meter (BNEF, 2017). structure and the different modules of the proposed IEAM. This
Therefore, Australia has been selected to validate the developed model is implemented using Microsoft Excel Spreadsheet based of
IEAM for distributed PV systems. real data for a complete year of PV generation and residential load
To the best authors’ knowledge, no integrated economic as- consumption for 54 customers in the Ausgrid’s electricity net-
sessment study of the distributed PV system is presented in work, NSW, Australia. However, the model can be easily adapted
the literature, which takes into account all the variables around for any location by updating the PV system parameters (tech-
the PV system (technical, economic, financial and environmental) nical, economic, financial and environmental), generation and
based on real market conditions and derived from real generation load profiles and retail tariffs and metering schemes (distribution
and consumption data. In addition, the sensitivity of different sys- network regulations). These four different modules are described
tem variables affecting its profitability and the variability distri- as follows:
bution of the system inputs and outputs has not been thoroughly
investigated yet. 2.1. Economical assessment of different electricity retail tariff struc-
The main contribution of this paper is to develop an Integrated tures
Economic Adaption Model (IEAM), which provides a holistic eco-
nomic assessment framework for distributed PV systems under Electric utility companies have developed exciting Dynamic
real integrated uncertainty and can be used by both the house- Pricing Plans (DPPs) to encourage consumers to change their
consumption patterns. In many regions, consumers can choose
hold and the utility decision makers. The developed model is
between different electricity pricing schemes such as Time of
easy to use, modify, adapt to different operating conditions, policy
Use (ToU) and Flat Rate (FR). Through DPPs, the consumers are
and regulatory variations. The proposed IEAM is developed and
encouraged to efficiently manage their energy consumption while
validated based on real PV generation and consumption data
attempting to flatten the overall load profile. Thus, DPPs are
from a real distribution network to serve as a full economic beneficial for both consumers and utilities through saving costs
assessment framework for distributed PV systems covering all the and reducing peak-to-average ratio, respectively. Therefore, it
aspects of a PV system (technical, environmental, economic and is essential to consider all possible tariff structures in order to
financial). The IEAM considers all the policy and regulatory issues conduct the cost–benefit analysis (Pillaia and Naserb, 2018).
in any distribution network through its four modules; economic Furthermore, in order to encourage households and small
assessment of different retail tariff structures as the first module, business owners to install PV systems in their premises, differ-
profitability analysis as the second module, sensitivity analysis as ent countries had implemented FiT for paying a premium price
the third module and probability distribution analysis based on for the generated electricity from PV systems in residential and
Monte Carlo Analysis (MCA) as the fourth module. The proposed commercial sectors under different Net Metering (NM) policies.
IEAM can be used by the household to economically assess the The output of a PV system can be recorded into two different
different metering schemes offered by the distribution network configurations, as illustrated in Fig. 2 (Zahedi, 2010), known as;
operator, to financially optimize the installed PV system size Net Billing (NB) and Net Metering (NM). In case of NB, the PV
according to the desired monthly saving, export rate and the pay- system owner, which has two meters installed at his premises,
back period and to estimate the effects of varying the different PV injects all his PV system generated electricity into the grid and
system variables based on its profitability under real integrated gets FiTNB (FiTNB can be equal to, below or above the retail tariff)
uncertainty. Also, the IEAM can be used by utility decision makers and imports his local electricity consumption from the grid with
to design their retail tariffs, to conduct profitability analysis for the retail tariff. In NB, the balance (imported electricity minus ex-
ported electricity) is done in monetary terms. However, in case of
different residential and commercial PV projects with different
NM, the PV system owner, which has one bidirectional meter in-
sizes and to estimate the effect of different incentives on the PV
stalled at his premises, injects the surplus PV generated electricity
system profitability under real integrated uncertainty.
(the PV system generated electricity minus the local consumed
This paper is organized as follows; Section 2 presents a de-
electricity) into the grid and gets FiTNM (FiTNM is normally equal
tailed description for the proposed IEAM and its four differ-
to the retail tariff) and the balance (imported electricity minus
ent modules. Section 3 presents the model validation based on exported electricity) is done in energy terms (kWh), this exported
the Australian case study of the Ausgrid’s electricity distribution energy typically varies from 10% to 50% of the total production of
network, New South Wales (NSW), Australia. Where the input the PV system and depends on the local consumption profile and
parameters for the four different modules of the IEAM are defined the installed PV system capacity, (Poruschi et al., 2018). The NB
based on: the Ausgrid’s customer load and generation profiles, will be used in this study; however, there are many NB and NM
retail and FiT tariff structures, PV support policies and current (fully or partially) policies used by different countries which have
market arrangements. Then, the results of this IEAM are pre- different economic impact on the PV system profitability, the
sented and discussed in Section 4. Finally, Section 5 presents the following two approaches will be used in this study, (Dufo-López
main conclusion of this study. and Bernal-Agustín, 2015):
314 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326

Fig. 1. Structure of the developed Integrated Economic Adoption Model for a distributed PV system.

• Net Billing Simple (BN-S): at the end of the billing period, with the initial installation, payback times tend to be relatively
if the imported electricity cost is higher than the exported long depending on the local tariff rates and incentives. Therefore,
electricity cost, the PV system owner must pay the grid the objective of this module is to provide the reader with a com-
operator for the difference between the two terms. If the prehensive understanding of the profitability analysis for the PV
cost of the energy imported is higher than cost of energy investment throughout its lifetime. The economic assessment is
exported, the PV owner receives no compensation. made using traditional economic indicators based on discounted
• Net Billing with Buy Back (NB-BB): at the end of each billing cash flows analysis, such as NPV, IRR, PI and DPBP. In addition,
period, the PV system owner pays the grid operator for all the LCOE is also computed. These economic indicators will be
imported electricity (at retail rate) and the grid operator calculated for the four proposed retail tariff structures from the
purchases any exported electricity at a certain FiT rate using previous module to benchmark their profitability. The different
a buy-back scheme. economic indicators are explained as follows.
Net Present Value (NPV) is a very important parameter in deter-
In both cases, the PV system owner may be required to pay the mining the economic feasibility of an investment as it represents
daily supply charges for grid connectivity according to Ausgrid’s the net profit generated. A project is economically viable if the
electricity network operator. NPV is positive, therefore, the project is economically justified if
In this IEAM, the two different retail tariff structures, which its NPV>0 and rejected if its NPV<0. When comparing multiple
are the FR and ToU tariffs, are combined with two different NB projects, the one with the highest NPV is the best option to make
schemes, which are NB-S and NB-BB, to propose four different the investment decision. The NPV calculation relies on the initial
retail tariff structures: (i) Flat Rate-Net Billing Simple (FR-NB-S), investment, the total accumulated cash flow and the discount
(ii) Flat Rate-Net Billing with Buy Back (FR-NB-BB), (iii) Time of rate. The NPV is the sum of present values of all cash inflows (rev-
Use-Net Billing Simple (ToU-NB-S), (iv) Time of Use-Net Billing enues) and outflows (costs) related to the investment (discounted
with Buy Back (ToU-NB-BB). Collectively, and based on the real cash flow over the lifetime of the project), thus accounting for
PV generation and load profiles, the cost balance between the the time value of money and the NPV is given by, Caamaiio and
imported and exported electricity is calculated on hourly basis to Lorenzo (1996), Nofuentes et al. (2002) and Talavera et al. (2007):
compute the monthly saving in the electricity bill (as percentage)
for the four proposed retail tariff structures for a household with
L
PV system compared with normal household. ∑
NPV = (It − Ot ) (1)
t =0
2.2. Profitability analysis of different electricity tariff structures
Where, I represents the discounted cash inflows, O represents
While costs of PV systems have declined rapidly, it is still diffi- the discounted cash outflows, and L is the lifetime of the project.
cult to determine the economic benefits of making an installation The project inflows (revenues) are the direct economic benefits
especially for the household owner who will personally finance of the project; in case of a subsidized PV system, it consists
his rooftop PV insulation. With the majority of costs being born of a onetime capital subsidy (S0 ), that may be provided by the
 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326 315

Fig. 2. Net Billing and Net Metering scheme for residential sector (Camilo et al., 2017).

government to promote renewable energy installation, and two size, used technology, market maturity level and other conditions,
operational components; savings from deferring purchases of typically ranging from 1%–5%.
grid electricity (Qd ) at the retail price (P r ) and sales of surplus Annual financing cost (Ft ) refers to the reimbursement of the
electricity to the grid (Qe ) at the FiT (PFiT ), as represented by: PV system installation loan, if there is one. The house owner may
L
decide to get a loan to finance the PV system installation partially
∑ Qd Pr + Qe PFiT or in total, which is incorporated by Equity and debt ratios in this
I = S0 + (2)
(1 + da )t study, therefore, the annual loan payment should be included in
t=1
the system costs (outflows). For each year, the loan payment will
Where, da is the actual discount rate, which is derived from consist of some principal P(t) and interest on this principal for
the nominal discount rate (dn ) and the annual inflation rate (g) t years at specified interest rate il . The amount of the nth year
by: principle P(t) and interest I(t) can be given by, Singh and Singh
(2010):
da = (dn − g)/(1 + g) (3)
Pl (t)
Most of PV system revenue depends on its energy delivery, P (t) = (6)
(1 + il )t
EPV . The useful energy generated by a PV system is a function of
many parameters, such as: the nominal power of the PV system, I (t) = Pl (t) − P (t) (7)
the average global solar irradiance at the installation location,
where, Pl (nt ) is the value of tth loan payment, which is given by:
the PV array orientation, the environmental conditions (possible
shadowing, temperature, soiling, snow, etc.) and the different
system components efficiencies (i.e. PV array and inverter). The Pl (t) = Pl (1)(1 + e)t−1 , for t = 1 to N (8)
effects of all these parameters are combined into one variable
where, e is escalation in the annual loan installment (%) and N is
called the Performance Ratio (PR), (Chabot, 1998). The PR is a
the loan repayment period (years). It has been assumed that the
measure for the degree of PV system utilization, it describes the
annual loan payment is due at the end of each year; therefore, the
relationship between the actual and theoretical energy output of
household can do the loan payment from the earning of selling
a PV system. Typical ranges of the PR have evolved from 50%–75%
surplus PV electricity to the grid. Therefore, the financing cost can
in the late 1980s, via 70%–80% in the 1990s, to >80% nowadays,
be giving by:
with some systems reaching 90%, (van Sark et al., 2012; de Lima {
et al., 2017) and (Cherfa et al., 2015). Therefore, the PV system Pl (t) , t = 1 to N
output can be given by: Ft = (9)
0, t = N + 1 to L
Gdt(β0 )
Epv = Pnom . .PR (4) Typically, in most countries with active distributed PV grid
Gstc
integration, the PV systems installed at domestic sites are con-
where Pnom is the nominal installed power of the PV system, Gstc is sidered non-taxable (for personal use/hobby) as they are not in
the solar irradiance under standard test conditions (1 kW/m2 ) and the nature of a business or profit making scheme.
Gdt(β0 ) is the solar irradiance at optimum position and is typically Finally, as the PV system owner is supplied in part by the PV
available from the Solar Radiation Atlas. system and in part by the distribution network, the NPV must
The project outflows (costs) can be divided into one-time take into account the respective payments (supply charge cost or
and reoccurring costs. The one-time costs include the PV system grid connection cost) for the distribution network operator and
installation I0 = CkWp .Pnom , (where, CkWp is the system initial this cost should be part of the retail tariff. This supply charge
cost per kWp), while reoccurring costs include operations and cost may vary depending on the tariff scheme and the user
maintenance (OM t ) and financing (Ft ). A one-time VAT tax on the consumption level, (Vilaça Gomes et al., 2018).
system (T0 ) may be included depending on the installation area, The Profitability Index (PI) is an investment assessment tech-
as well as the required payment of VAT on income from oversold nique which is defined as the ratio between its net present value
electricity (Tt ). The discounted lifetime cost (outflows) for a PV and its initial investment cost (benefit cost/ratio) and is given by:
system is given by:
L NPV
∑ OMt + Ft + Tt PI = (10)
O = I0 + T0 + (5) I0
(1 + da )t
t=1
The NPV is an absolute measurement in cash while the PI is
The annual OM t cost is normally given as a percentage of the a relative measurement, which can be used to compare several
PV system initial cost, this percentage depends on PV system investments with different cost and to rank the relative returns.
316 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326

The investment breakeven point is when the PI is equal to one,

therefore, a PI greater than one indicates a profitable investment,
and when a PI is equal to two the investment is doubled, etc.
(Drury et al., 2011).
Discounted Payback Period (DPBP) is the required number of
years to recover the initial investment made on a project by
considering the discount rate. Therefore, the present values of the
net cash flows (cash inflows minus cash outflows) are summed
year by year until the total equals zero to get the DPBP, which
can be represented by:
⎧ T
⎨∑
(It − Ot ) = 0
⎪
DPBP = T (11)
⎪ t=0
dn > 0
⎩

Obviously, the profitability means that the DPBP should be

shorter than the serviceable life of the system; otherwise, the ini-
tial investment cannot be recovered within the project’s lifetime.
Internal Rate of Return (IRR) of an investment project is the
value of the discount rate that leads to zero NPV. In general, no
analytical solution can be found for the equation NPV=0, so the Fig. 3. Monte Carlo Analysis outline.
value of IRR can be determined by iteration numerical calculation
and it can be given by:
L and LCOE by varying a selection of major input parameters by
± 10%. Furthermore, this sensitivity analysis demonstrates the
∑
0= (It − Ot ), where dn = IRR (12)
robustness of the economic analysis and gives solidity to the
t=0
obtained results. This sensitivity analysis is considered as an
The higher the value of IRR, the more positive the value of NPV important tool for gauging the importance of different factors
appears to be and the more stable is the project (less risks), (No- in driving the overall financial result of the PV system. The four
fuentes et al., 2003). most influential input variables (combining technical, economic
The Levelized Cost of Electricity (LCOE) is a measure for defining and financial inputs) are the solar irradiance, initial investment
the cost of electrical energy generated from a PV system; it is the cost, the operation and maintenance cost and the electricity retail
net present value of the unit-cost of electricity. The LCOE calcu- tariff (the discount rate has been considered in (Audenaert et al.,
lation is defined as the discounted production cost of installing 2010)). The impact of varying each variable will be graphically
and operating a project divided by the energy generated from this visualized on both the NPV and LCOE, (Ren et al., 2009) and
project over its lifetime, as given by: (Cucchiella et al., 2016). Collectively, this module highlights the
Lifecycle cost most significant variables impacting the NPV and the LCOE of the
LCOE = (13) PV system while varying a selection of major PV system input
Lifetime energy production
parameters by predefined percentage.
L
∑ 2.4. Levelized Cost of Electricity (LCOE) probability distribution
Ot
t=0
LCOE = (14) Sensitivity analysis can help to determine a reasonable range
L
∑ of outcomes, or identifying most critical factors, but cannot pro-
Et /(1 + dn ) t
vide insight into the probability of such outcomes taking place.
t=0 However, Monte Carlo Analysis (MCA) is a more valuable plat-
Where, Et is the energy yield in year t. As the solar PV technology form for assessing long term uncertainties in deregulated markets
becomes more mature, the economic feasibility of PV projects is since it can quantify risk and present it in a way that is easy
evaluated using the LCOE generation to compare it with other for non-professional investors to understand, (Sommerfeldt and
electricity generation technologies, (Branker et al., 2011) and Madani, 2017). In MCA, a calculation is performed many times,
(Hernández-Moro and Martínez-Duart, 2013). and each set of parameters is chosen randomly from predefined
Finally, this module presented the methodology for perform- distributions for each input, creating thousands of possible sce-
ing the profitability assessment for the distributed PV adoption. narios. The choice of probability distributions of the inputs and
Based on the PV system technical, environmental, economic and their limits are very important for accurate MCA results, since, in-
financial parameters, the NPV, IRR, PI, DPBP and LCOE are calcu- appropriate choices may lead to mistaken decisions. Inputs can be
lated and compared for the four proposed retail tariff structures. completely independent or linked, allowing causal relationships
that better represent real world conditions (e.g. a reduction in
2.3. Sensitivity analysis for a distributed photovoltaic grid connected electricity usage due to high prices). MCA results are presented
system as a probability distribution to show the occurrence possibility of
a certain situation rather than a point value. The outline of MCA
There is a wide range of inputs that go into the PV system is illustrated in Fig. 3. Three steps are required in the simulation
profitability analysis and their values may change according to process: (1) sampling on random input variables, (2) evaluating
government policies, technology used, location, country economic model output, and 3) statistical analysis on model output. In this
conditions, etc. The objective of this module is to demonstrate work, the MCA is also implemented using Microsoft Excel.
the influence of certain PV system variables on its profitability; This is particularly valuable for PV investment, which are
therefore, a sensitivity analysis is performed mainly on the NPV subject to many unforeseen negative events to occur during its
 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326 317

Fig. 4. Aggregated generation and load of 54 customers at Ausgrid distribution network, July 2012 to June 2013.

long lifetime and payback period. MCA produces a probability 3.1. Load and generation profiles
distribution of different outcomes for the main economic perfor-
mance indicators (such as the LCOE) providing a more accurate Ausgrid provides a free dataset called ‘‘Solar home electricity
projection as a result (Jeon and Shin, 2014) and (da Silva Pereira data’’, which can be used by organizations and individuals for a
et al., 2014). The purpose of this module is to demonstrate a variety of purposes, including research, policy-making and pro-
complete MCA for a residential PV system owner given the cur- viding information about solar PV system performance. This data
rent market conditions and support policies under real integrated has been sourced from 300 randomly selected solar customers in
system uncertainty. MCA answers the question: ‘‘what is the Ausgrid’s electricity network area that were billed on a domestic
probability of getting a LCOE (for the local installed PV system) tariff based on net-billing scheme for the period from 1 July 2010
that is less than the current electricity retail rate?’’. Therefore, to 30 June 2013. The chosen customers had a full set of actual
five input variables have been selected which exhibit random data (collected every half-hour), which went through some data
variability and represented with probability distributions, these quality checking and resulting in excluding customers on the high
and low ends of household consumption and solar generation
variables are; PV system installation cost, operation and main-
performance, (Anon, 2013). The Ausgrid dataset includes: the in-
tenance annual cost, average annual solar irradiance, electricity
stalled PV capacity (kWp), the solar PV generation (kWh) and the
retail price and FiT. Other studies has considered other variables
residential energy consumption (kWh) recorded every half hour
such as: tax refund rate, market risk premium, risk free rate and
for each of the 300 customers. Out of these 300 customers, a clean
foreign exchange rate, which are not affecting the distributed PV
dataset for 54 customers without any load and/or generation
system in the Australian market, (Jeon and Shin, 2014). Applying
abnormalities as described In Ratnam et al. (2015). Fig. 4 shows
all the possible probability distribution for these five variables the aggregated PV generation and load of these 54 customers over
and running the MCA for ten thousand iterations, the probabil- a one year period based on half-hour intervals daily real measure-
ity distribution of the LCOE, which is cheaper than the current ments from July 2012 to June 2013. The peak load occurred on 12
electricity retail price, can be obtained. Eventually, this module January 2013, which was recorded as the hottest day in summer
produces the probability distribution of the LCOE (or any other 2013 in NSW as recorded by the bureau of meteorology, also,
economic indicator) due to randomly varying a selection of PV the observed zero load point corresponds to changes in daylight
system input parameters by using MCA. saving time on 3 October. Fig. 5 illustrates the average monthly PV
generation and load, shown by season over the day based on half-
3. Model verification—Australian case study hour generation and consumption measurements. This dataset
has been used in this work to provide the average PV system
installed capacity, the average hourly, weekly, monthly and yearly
Following the description of the proposed IEAM and the de-
load and PV generation profiles for a typical Ausgrid residential
termination of the main inputs and outputs of its four different
customer.
modules in the previous section, this section presents the applica-
tion of the proposed IEAM for the Ausgrid’s electricity distribution 3.2. Tariff structures
network, NSW, Australia, under the current market arrangements
and PV support policies. Ausgrid’s network supplies electricity to The Australian electricity retailers provide different DPP op-
more than 1.7 million customers and covers an area of 22,275 tions to the consumers. These DPPs include three different tariffs
km2 which includes some of the most densely populated and such as ToU, FR and FiT. Figure 6 shows the three different tariff
fastest growing areas in NSW. Ausgrid supports the installation structures; the FR and FiT which are constant though the day
and connection of small scale PV systems. As of 2017, around and the latter is negative as the consumers are paid, while the
108,000 solar customers are connected to its network. The follow- third structure, which is the ToU scheme is divided into three
ing subsections include the technical, environmental, economic different rates: peak, shoulder and off-peak. Table 1 illustrates the
and financial parameters of a residential PV system connected NSW residential energy price, in Ausgrid Distribution Zone (Anon,
to the Ausgrid’s network under the current market arrangements 2017a). Through these DPPs, the customers are encouraged to
and PV support policies, (Ausgrid, 2017). achieve financial benefits by utilizing the offered incentives.
318 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326

Fig. 5. Ausgrid average monthly PV generation and load per customer during the day, July 2012 to June 2013.

Fig. 6. Different retail electricity tariffs, NSW 2017.

The NSW Solar Bonus Scheme (SBS) began in 2009 initially 3.3. PV system technical parameters
offering a FiT to eligible customers for all PV generation (based
on net-billing scheme) at a rate of a 60 Australian cent/kWh with The PV technical system parameters include, the peak installed
NB, reduced to 20 c/kWh after October 2010. The SBS 60 and capacity (kWp), the expected life time of the PV system and the
SBS 20 schemes ended on December 2016. From April 2011, all annual system degradation rate. In the current study, the peak
generation received an amount set by electricity retailer in the installed capacity is the average installed PV system capacity for
the 54 customers at Ausgrid distribution network. Most solar
range of 6–12c/kWh and it is not compulsory for retailers to offer
panels manufacturers offer a standard 25-year warranty, guar-
this FiT, (Anon, 2017d) and (Jacobs, 2017). The current study
anteeing that their energy output will remain at 80% of capacity
is using a FiT of 9 Australian cent/kWh and this FiT is applied
for 25 years. Moreover, recent solar inverters carry 10–15 year
only in the retail tariff structures with buy back, as indicated in warranty, which means that the solar inverter should be replaced
Table 1. There is no grid costs associated with the existence of once during the lifetime of the PV system and the replacement
the PV system, however, there is daily supply charge for every cost is normally included in the annual system operation and
customer based on his tariff scheme, and in this work the daily maintenance cost. The PV system degradation rate defines the
supply charge is included in the tariff calculation. percentage of system output reduction from year to year and
 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326 319

Table 1
NSW residential energy price, AusGrid distribution zone, September 2017, (Anon, 2017a).
Tariffs Time periods of the day Electricity prices Daily supply charge
(Australian Cents/kWh) (Australian Cents/Day)
Flat rate 24 h 28.52 83.39
Peak 2pm–8pm on working weekdays 53.01
Time of
Shoulder 7am–2pm, 8pm–10pm working weekdays, 23.79 96.85
Use (ToU)
7am–10pm on weekends & public holidays
Off peak All other times 14.42
Solar feed-in tariffs 24 h 9 –

Table 2
Technical, economic and financial parameters for the average residential PV system at Ausgrid.
Parameter Value Dimension
PV system technical information
Peak installed capacity 2.45 kWp
PV system life time 25 year
PV system degradation rate 0.5 %/year
PV system economical information
Cost/Wp 1.72 AUD
O&M cost/Wp /year 1% of the capital cost per Wp AUD
PV system financial information
Annual inflation rate 2.5 %
Nominal discount rate 10.29 %
Electricity and O&M cost escalation rate 2.5 %
Loan interest 8.5 %
Loan period 15 Year
Equity and debt ratio 50% to 50% %
Amortization period 5 Year
Tax 0 %

Fig. 7. Solar Global Horizontal Irradiation (GHI) in three different cites at NSW, Jul 12–Jun 13.

it is approximately range from 0.1%–1% depending on the solar of incentives which reduce the price to consumers by a further
panel technology. The Australian Energy Council has used 0.5% as 60–70c/Wp , depending on installation area, therefore, the average
a constant degradation rate, (IEA-PVPS, 2017). watt cost is considered as 1.72 AUD/Wp , (Anon, 2017d).

3.4. PV system economic parameters

3.5. PV system financial parameters
The PV system economic parameters include the capital ex-
penditures (CAPEX), which is the installing cost of the PV system The PV system financial data includes the annual inflation rate,
per watt (I0 ) and the operating expenses (OPEX ) which is the the nominal discount rate, electricity and O&M cost escalation
operation and maintenance cost (OMt ) of the PV system per watt rate, the loan interest rate, the loan period, the equity to loan
per year. The best-practices for O&M cost are generally around ratio, the depreciation period and the income taxation. These
0.5% of system initial cost per year for large systems and 1% parameters are market dependent and vary from one country
for small systems, In this work, the O&M costs are assumed to or investment sector to another. In the case of Ausgrid, these
be 1% of total system costs, (NREL, 2016) and (IRENA, 2017). financial and other technical and economic parameters are given
According to Australian PV institute, the average price per watt is in Table 2, (Mills et al., 2011; Oliva et al., 2013) and (Yu and
2.42 Australian Dollar (AUD), March 2017, this price is exclusive Halog, 2015).
320 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326

Fig. 8. Average monthly load, local generation, exported energy and exported rate per each of the dataset Ausgrid’s customers.

Fig. 9. Average monthly electricity cost without local PV generation (Ausgrid).

3.6. PV system environmental parameters represented in Fig. 6 and Table 1 for the AusGrid distribution net-
work service provider in NSW, Australia, an economic assessment
The PV system environmental parameters include solar irradi- of the four proposed retail tariff structures (FR-NB-S, FR-NB-BB,
ance level, location and tilt angle and temperature. The combina- ToU-NB-S and ToU-NB-BB) has been conducted with the results
tion of Australia’s dry climate and latitude gives it high benefits illustrated in Figs. 8 to 11.
and potential for solar PV energy production. The Solar Global Fig. 8 illustrates the total monthly average customer load,
Horizontal Irradiation (GHI) is the most important parameter local generation, exported energy and the export rate, where
in calculating PV system generated electricity, Fig. 7 shows the the exported energy is the local generated energy which is not
average monthly solar irradiance all over a year in different cities used locally and exported into the distribution network and
at NSW, (Morrison and Litvak, 0000). the export rate is the ratio between the exported energy and
the local consumed energy from the local generated energy.
4. Results The average annual local load and generation are 6944 and
3363 kWh/customer, respectively. The average PV yield is 1373
This section details the results of applying the proposed IEAM kWh/kWp which is in agreement with average PV yield in Aus-
under current prices, regulation and irradiation conditions tralia (1400 kWh/kWp), (Anon, 2017d). The export rate changes
throughout Austigrid distribution network, NSW, Australia. The monthly ranging from 20%–50%, however, the average annual
results of the different four modules of the IEAM will be presented export rate is 39%. Thus, almost two thirds of the local generated
and discussed as follows. energy (based on 2.45 kWp average PV system size for the given
data) is consumed locally. Fig. 9 shows the average monthly
4.1. First module electricity cost per customer without local PV generation for ToU
and FR schemes, respectively.
Using the annual generation and consumption real data rep- Fig. 10 shows the average monthly saving for the four pro-
resented in Figs. 4 and 5 and the retail prices and tariff schemes posed scenarios, where the saving is high during the summer
 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326 321

Fig. 10. Percentage average monthly saving for different scenarios (Ausgrid).

Fig. 11. Percentage average annual saving for different scenarios (Ausgrid).

Table 3 are varied from 7.6 to 10.75 years depending of the proposed
Summary of economic indicators value for profitability analysis of the
retail tariff scenarios, with the ToU-NB-BB scheme giving the
presidential PV system.
Economic indicator Proposed retail tariff scenarios
highest NPV, project and Equity IRR, profitability index and short-
est payback period. It is noted that, all schemes are profitable
FR-NB-S FR-NB-BB ToU-NB-S ToU-NB-BB
within the Ausgrid network and the details of the different eco-
NPV (AUD) 2842.39 4419.88 3421.33 4998.83
Project IRR (%) 14 17 15 18 nomic indicator for the different proposed retail pricing schemes
Equity IRR (%) 17 23 19 25 are illustrated in Fig. 12 and summarized in Table 3.
PI (Unitless Ratio) 1.68 2.05 1.82 2.19
DPBP (years) 10.75 8.28 9.67 7.61

4.3. Third module

months (December, January and February) and low in winter

months (June, July and August), where the weather is cold and the Fig. 13 and Fig. 14 show the impact of the variation of key
energy consumption is high, which is line with the solar radiation parameters by ± 10% on the LOCE and NPV ranking the inputs
illustrated in Fig. 7. Fig. 11 illustrates the average annual saving from the ones with the highest impact to the least, respectively.
for the four different presented electricity retail tariff scenarios, Fig. 13 shows the impact of the four inputs on the NPV output, as
where the ToU-NB-BB achieves the maximum saving and FR-NB- shown the solar irradiance and tariff rate have the highest impact
S achieves the minimum saving; getting financing compensated on the NPV change followed by the initial cost. The O&M cost
for the exported electricity to the grid will increase the saving by has very low impact that is almost insignificant. As noticed from
21.3% and 19.8% for the FR and ToU schemes respectively. Fig. 14, the most significant variables impacting the LCOE are the
solar irradiance and the system initial cost, therefore, an extended
4.2. Second module
evaluation is carried out to investigate the mutual initial cost and
The profitability analysis resulted in LCOE equal to 0.1959 solar irradiance effect beyond the 10% variations as illustrated in
AUD/kWh which is less than the flat rate at AusGrid (28.52 Aus- Fig. 15, where, average values’ boundaries are marked by orange,
tralian Cents/kWh), the payback period results of the PV system encircling LCOE value around the current: 0.1959 AUD/kWh.
322 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326

Fig. 12. Economic indicators for the profitability analysis of a residential PV system: (a) net present value; (b) project and equity internal rate of return; (c) profitability
index and (d) discounted payback period.

Fig. 13. Impact on the Net Present Value associated with variation of key parameters ± 10%, ranking the inputs from the highest significant impact to the least.

Fig. 14. Impact on LCOE with variation of key parameters ± 10%, ranking the inputs from the highest significant impact to the least.
 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326 323

Fig. 15. Levelized cost of electricity sensitivity with variation of average annual solar irradiance and initial investment cost per watt.

4.4. Fourth module install PV systems and reduce their electricity purchases from the
distribution network. However, the payback from PV system is re-
The LCOE calculation is highly sensitive to the following five alized over a long term and PV system owners could be uncertain
variables: installation cost, O&M cost, average annual solar ir- about many factors that impact the system payback over such
radiance, electricity retail price and FiT. Therefore, these vari- a long period. Therefore, the main objective of this study is to
ables have been modeled to have random variability with differ- develop an intensive IEAM for distributed photovoltaic systems.
ent probability distributions as follows. Anon (2018a) and IEA- The model is implemented using MS Excel Spreadsheet based on
PVPS (2017). Also, Table 4 summarizes the assumed probability a real case study with one year real load and generation data from
distributions for these five variables. Ausgrid distribution network (NSW, Australia). The developed
IEAM is composed of four modules: economic assessment of four
• A wide variety of factors can influence the installation cost, different proposed electricity retail tariff schemes; profitability
such as; equipment quality, installer experience, and roofing
analysis of these proposed tariffs schemes; sensitivity analysis
conditions. Therefore, in this study, the installation cost
of different variables which strongly affect the distributed PV
per kW is assumed to be randomly linearly varied, from
system profitability; and finally, LCOE probability distribution
500–2500 AUD.
based on MCA due to random variations of a certain set of the
• Solar irradiance is subjected to many changes daily, monthly,
distributed PV system input variables.
seasonally and annually due to different climatic conditions
The following findings have been obtained (based on Ausgrid
and sun related conditions. Therefore, the average annual
cleaned dataset with the current retail and FiT tariffs, 2.45 kWp
solar irradiance is represented by normal distribution, with
average PV system size and 39% average annual export rate):
2000 (kWh/m2 /year) as mean and 150 as standard deviation
(STD) to cover all the possible variations in solar irradiance ◦ ToU-NB-BB achieves the maximum saving (40%) and FR-
across NSW. NB-S achieves the minimum saving (26%) out of the four
• The annual O&M cost per kW/year is assumed to be ran- proposed retail tariff scenarios; getting financing compen-
domly linearly varied between 1%–5% of the total initial cost
sated for the exported electricity to the grid will increase
of a PV system.
the saving by 21.3% and 19.8% for the FR and ToU schemes,
• The residential retail tariff is assumed to randomly vary
respectively.
linearly from 0.1–0.35 AUD/kWh.
◦ All retail tariff scenarios are profitable, and the ToU-NB-BB
• The FiT is assumed to linearly vary from 0–0.15 AUD/kWh.
achieves the highest NPV, project and equity IRR and PI,
Applying all the possible probability distribution for the five and shortest payback period. The profitability analysis re-
variables and running the MCA for ten thousand iterations fol- sulted in LCOE equal to 0.1959 AUD/kWh which is 31% less
lowing the steps illustrated in Fig. 3 using Microsoft Excel, the than the flat rate at Ausgrid (28.52 Australian Cents/kWh),
probability distribution of the LCOE can be obtained as illustrated the payback period of the PV system is varied from 7.6 to
in Fig. 16, where for 89.19% of the probability distribution, a 10.75 year depending of the proposed retail tariff scenarios.
LCOE less than the current retail rate can be achieved under ◦ The NPV of a PV system is very sensitive to the varia-
the wide range of variation in the five selected input variables tion on the solar irradiance and the initial cost; while,
which proves the system profitability under real integrated sys- the LCOE is very sensitive to solar irradiance and retail
tem uncertainty. Also, MCA has been conducted to other output tariff. Therefore, the extended sensitivity analysis using
variables NPV, PI and DPBP The results mean values are presented main influencing parameters is important to predict the
here, respectively, to provide the reader with additional analytic market behavior corresponding to potential climatic and/or
insights: 4,724.7253 AUD, 1.7683 and 9.6237 year. economical variations.
◦ Taking all the possible probability distribution into ac-
5. Discussion and conclusions count: PV system installation cost, operation and mainte-
nance annual cost, average annual solar irradiance, elec-
Due to falling PV systems cost and rising electricity tariff rates, tricity retail price and FiT simultaneously, the probability
it is becoming increasingly attractive for residential consumers to distribution of getting a LCOE from the local installed PV
324 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326

Table 4
MCA inputs variation range and their recommended probability distribution functions.
Parameter Distributions method Variability range Units
Solar irradiance Random (Gaussian) Mean: 2000, STD: 150 kWh/m2 /Year
Installation costs Linear 0.5–2.5 AUD/Wp
Annual operation and maintenance cost Linear 1%–5% % of initial cost/Wp/year
Electricity retail Tariff Linear 0.1–0.35 AUD/kWh
Feed in Tariff Linear 0.05–0.15 AUD/kWh

Fig. 16. Probability distribution of getting levelized cost of electricity less than the current retail rate.

system less than the current retail rate is 89.19%, which and evaluates the investment risk (through the sensitivity and
validates the system profitability under real integrated un- the probability distribution modules under real integrated uncer-
certainty. tainty). This in turn encourages further distributed PV penetration
level in the distribution network, which result in different bene-
The presented analysis in this paper mainly target standalone fits for the network operator such as reduce the severity of peak
PV systems, which are widely used on a global scale. Though, an electricity demand, reduce network distribution losses, reduce
emerging trend for distributed scale consumers in some countries CO2 emissions, provide firm capacity to the distribution network
(e.g. in USA and Australia) is to increase their PV system size (the with high degree of reliability at a high level of penetration
2017 average residential PV system size in Australia increased to and reduce the length of peak electricity demand which in turn
6 kW), to and couple it with distributed energy storage units to will reduce the stress in the distribution network and wholesale
increase their PV self-consumption and minimize export to the
prices.
grid. This trend is a result of lowering the FiT for the exported
energy by some distribution network operators, which makes it
Acknowledgment
more economically attractive for customers to maximize their
self-consumption rate rather than exporting their surplus PV
This publication is supported by Iberdrola S.A. as part of its
energy at lower rates. For such systems, an extended version
innovation department research activities. Its contents are solely
of the presented IEAM could be developed to take into account
different energy storage integration scenarios. the responsibility of the authors and do not necessarily represent
Collectively, the developed IEAM is of primordial importance the official views of Iberdrola Group.
for understanding and implementing the economic feasibility
study for any residential PV system under real integrated un- References
certainty. In addition, the IEAM can be applied to a variety of
Akter, M.N., Mahmud, M.A., Oo, Amanullah M.T., 2017. Comprehensive eco-
electricity market designs, retail rate structures, and PV compen-
nomic evaluations of a residential building with solar photovoltaic and
sation mechanisms by changing the generation and load profiles battery energy storage systems: An Australian case study. Energy Build. 138,
and relevant economic and financial parameters. The outcomes of 332–346.
this IEAM (based on retrospective scenario) will guide different Alassi, Abdulrahman, Ellabban, Omar, Bañales, Santiago, 2018. Generic dis-
utility, in emerging countries where PV technology is becoming a tributed photovoltaic cost outlook methodology: Australian market appli-
cation example. In: 7th International Conference on Renewable Energy
reality in their energy generation mix, by offering some useful
Research and Applications (ICRERA), pp. 144–149.
lessons learned to design their time-varying rates and policies Anon, 2013. www.ausgrid.com.au/Common/About-us/Corporate-information/
for billing residential electricity customers (net-billing policy, Data-to-share/Solar-home-electricity-data.aspx.
net-metering policy, FR tariffs and ToU tariffs) and to conduct Anon, 2017a. www.originenergy.com.au/content/dam/origin/residential/
profitability analysis for residential and commercial PV projects docs/energy-price-fact-sheets/nsw/12017/NSW_Electricity_Residential_
Endeavour%20Energy_Origin%20Supply.PDF.
with different sizes in different locations within their networks
Anon, 2017b. RE Thinking energy 2017, IRENA 2017. www.irena.org.
to estimate the adoption rate of residential and commercial solar Anon, 2017c. Renewables 2017 Global Status Report, REN21 2017, www.irena.
PV systems. Adding to that, the IEAM gives a clear answer for org.
the PV investment profitability question for the house owner Anon, 2017d. PV in Australia Report 2016, APVI.
 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326 325

Anon, 2018a. Australian Energy Council, Solar Report, www.energycouncil.com. International Energy Agency, 2016. World energy outlook 2016. www.iea.org.
au. IRENA, 2016. The power to change: solar and wind cost reduction potential to
Anon, 2018b. 2017 outlook for energy: A view to 2040. exxonmobil.com. 2025.
(Accessed 21 January 2018). IRENA, 2017. IRENA Cost and Competitiveness Indicators: Rooftop Solar PV.
Audenaert, Amaryllis, De Boeck, Liesje, De Cleyn, Sven, Lizin, Sebastien, International Renewable Energy Agency, Abu Dhabi.
Adam, Jean-François, 2010. An economic evaluation of photovoltaic grid Jacobs, 2017. Projections of Uptake of Small-Scale Systems: Australian Energy
connected systems (PVGCS) in Flanders for companies: A generic model. Market Operator.
Renew. Energy 35 (12), 2674–2682. Jaeger-Waldau, Arnulf, 2017. PV Status Report 2017, Publications Office of the
Ausgrid, 2017. Distribution and Transmission Annual Planning Report, www. European Union, Luxembourg.
ausgrid.com.au. Jeon, Chanwoong, Shin, Juneseuk, 2014. Long-term renewable energy technology
Ayadi, Osama, Al-Assad, Rami, Al Asfar, Jamil, 2018. Techno-economic assess- valuation using system dynamics and Monte Carlo simulation: Photovoltaic
ment of a grid connected photovoltaic system for the University of Jordan. technology case. Energy 66, 447–457.
Sustainable Cities Soc. 39, 93–98. Kazem, Hussein A., Albadi, M.H., Al-Waeli, Ali H.A., Al-Busaidi, Ahmed H.,
BNEF, 2017. New Energy Outlook 2017: Global Overview. Chaichan, Miqdam T., 2017. Techno-economic feasibility analysis of 1MW
Branker, K., Pathak, M.J.M., Pearce, J.M., 2011. A review of solar photovoltaic photovoltaic grid connected system in oman. Case Stud. Therm. Eng. 10,
levelized cost of electricity. Renew. Sustain. Energy Rev. 15 (9), 4470–4482. 131–141.
Byrnes, Liam, Brown, Colin, 2015. Australia’s Renewable Energy Policy: The Case Kharitonov, V.V., Kosterin, N.N., 2017. Criteria of return on investment in nuclear
for Intervention, MPRA Paper 64977. University Library of Munich, Germany. energy. Nucl. Energy Technol. 3 (3), 176–182.
Caamaiio, E., Lorenzo, E., 1996. Modelling and financial analysis tools for PV Koumparou, Ioannis, Christoforidis, Georgios C., Efthymiou, Venizelos, Papa-
grid-connected systems. Progr. Photovolt. 4, 295–305. giannis, Grigoris K., Georghiou, George E., 2017. Configuring residential PV
Cali, Umit, Erdogan, Nuh, Kucuksari, Sadik, Argin, Mehmet, 2018. Techno- net-metering policies – A focus on the mediterranean region. Renew. Energy
economic analysis of high potential offshore wind farm locations in Turkey. 113, 795–812.
Energy Strategy Rev. 22, 325–336. Kumar, Deepak, 2015. Economic assessment of photovoltaic energy production
Camilo, Fernando M., Castro, Rui, Almeida, M.E., Fernão Pires, V., 2017. Economic prospects in India. Proc. Earth Planet. Sci. 11, 425–436.
assessment of residential PV systems with self-consumption and storage in Kuo, Way, Pan, Chin, 2018. A reliability look at energy development. Joule 2,
Portugal. Sol. Energy 150, 353–362. 1–9.
Chabot, Bernard, 1998. From costs to prices: Economic analysis of photovoltaic Lee, Minhyun, Hong, Taehoon, Yoo, Hyunji, Koo, Choongwan, Kim, Jimin,
energy and services. Progr. Photovolt. 6, 55–68. Jeong, Kwangbok, Jeong, Jaewook, Ji, Changyoon, 2017. Establishment of a
Chapman, Andrew J., McLellan, Benjamin, Tezuka, Tetsuo, 2016. Residential solar
base price for the solar renewable energy credit (SREC) from the perspective
PV policy: An analysis of impacts, successes and failures in the Australian
of residents and state governments in the United States. Renew. Sustain.
case. Renew. Energy 86, 1265–1279.
Energy Rev. 75, 1066–1080.
Cherfa, F., Hadj Arab, A., Oussaid, R., Abdeladim, K., Bouchakour, S., 2015.
Li, Tianqi, Roskilly, Anthony Paul, Wang, Yaodong, 2017. Life cycle sustainability
Performance analysis of the mini-grid connected photovoltaic system at
assessment of grid-connected photovoltaic power generation: A case study
algiers. Energy Procedia 83, 226–236.
of northeast England. Appl. Energy.
Christoforidis, G.C., Panapakidis, I.P., Papadopoulos, T.A., Papagiannis, G.K.,
Löbberding, Laurens, Madlener, Reinhard, 2019. Techno-economic analysis of
Koumparou, I., Hadjipanayi, M., Georghiou, G.E., 2016. A model for the
micro fuel cell cogeneration and storage in Germany. Appl. Energy 235,
assessment of different net-metering policies. Energies 9, 262.
1603–1613.
Comello, Stephen, Reichelstein, Stefan, 2017. Cost competitiveness of residential
Lomas, J.C., Muñoz Cerón, E., Nofuentes, G., de la Casa, J., 2018. Sale of profitable
solar PV: The impact of net metering restrictions. Renew. Sustain. Energy
but unaffordable PV plants in Spain: Analysis of a real case. Energy Policy
Rev. 75, 46–57.
117, 279–294.
Commonwealth Government of Australia, 2015. Energy White Paper, Department
Masebinu, S.O., Akinlabi, E.T., Muzenda, E., Aboyade, A.O., 2017. Techno-
of Industry, Innovation and Science, 2015.
economics and environmental analysis of energy storage for a student
Cucchiella, Federica, D’Adamo, Idiano, Gastaldi, Massimo, 2016. A profitability
residence under a South African time-of-use tariff rate. Energy 135, 413–429.
assessment of small-scale photovoltaic systems in an electricity market
Mills, Graham, Passey, Robert, Watt, Muriel, Franklin, Simon, Bruce, Anna, John-
without subsidies. Energy Convers. Manage. 129, 62–74.
son, Warwick, 2011. Modelling of PV & Electricity Prices in the Australian
da Silva Pereira, Edinaldo José, Pinho, João Tavares, Galhardo, Marcos André Bar-
Residential Sector. The Australian PV Association.
ros, Macêdo, Wilson Negrão, 2014. Methodology of risk analysis by Monte
Carlo method applied to power generation with renewable energy. Renew. Morrison, Graham L., Litvak, Alex, 0000. Condensed solar radiation data base for
Energy 69, 347–355. Australia UNSW, Solar Thermal Energy Laboratory, University of New South
de Lima, Lutero Carmo, de Araújo Ferreira, Leonardo, de Lima Morais, Francisco Wales, Report No 1/1999.
Hedler Barreto, 2017. Performance analysis of a grid connected photovoltaic Nelson, D.B., Nehrir, M.H., Gerez, V., 2005. Economic evaluation of grid-connected
system in northeastern Brazil. Energy Sustain. Dev. 37, 79–85. fuel-cell systems. IEEE Trans. Energy Convers. 20 (2), 452–458.
Drury, Easan, Denholm, Paul, Margolis, Robert, 2011. The Impact of Different Nofuentes, G., Aguilera, J., Munoz, F.J., 2002. Tools for the profitability analysis
Economic Performance Metrics on the Perceived Value of Solar Photovoltaics, of grid-connected photovoltaics. Prog. Photovolt: Res. Appl. 10, 555–570.
NREL/TP-6A20-52197. Nofuentes, G., Aguilera, J., Rus, C., Santiago, R.L., 2003. A short assessment on the
Dufo-López, Rodolfo, Bernal-Agustín, José L., 2015. A comparative assessment profitability of PV grid-connected systems using classical investment project
of net metering and net billing policies. Study cases for Spain. Energy 84, analysis. In: Proceedings of 3rd World Conference on Photovoltaic Energy
684–694. Conversion, Osaka, Japan, pp. 2632–2635.
Edalati, Saeed, Ameri, Mehran, Iranmanesh, Masoud, Tarmahi, Hakimeh, Gholam- NREL, 2016. Best Practices in Photovoltaic System Operations and Maintenance.
pour, Maysam, 2016. Technical and economic assessments of grid-connected Okoye, Chiemeka Onyeka, Oranekwu-Okoye, Blessing Chioma, 2018. Economic
photovoltaic power plants: Iran case study. Energy 114, 923–934. feasibility of solar PV system for rural electrification in Sub-Sahara Africa.
Ellabban, Omar, Abu-Rub, Haitham, Blaabjerg, Frede, 2014. Renewable energy Renew. Sustain. Energy Rev. 82 (3), 2537–2547.
resources: Current status, future prospects and their enabling technology. Oliva, Sebastián, MacGill, Iain, Passey, Rob, 2013. Estimating the Financial Costs
Renew. Sustain. Energy Rev. 39, 748–764. and Benefits of Distributed Grid-Connected Photovoltaics for Different Elec-
Frankfurt School-UNEPCentre/BNEF, 2017. Global trends in renewable energy tricity Industry Participants. Centre for Energy and Environmental Markets,
investment. http://www.fs-unep-centre.org. The University of New South Wales.
Geoscience Australia and BREE, 2014. Australian Energy Resource Assessment, Pillaia, Gobind, Naserb, Husain Ali Yaqoob, 2018. Techno-economic potential
second ed. Geoscience Australia, Canberra, (Chapter 10). of largescale photovoltaics in bahrain. Sustain. Energy Technol. Assess. 27,
Haegermark, Maria, Kovacs, Peter, Dalenbäck, Jan-Olof, 2017. Economic feasibility 40–45.
of solar photovoltaic rooftop systems in a complex setting: A Swedish case Poruschi, Lavinia, Ambrey, Christopher L., Smart, James C.R., 2018. Revisiting
study. Energy 127, 18–29. feed-in tariffs in Australia: A review. Renew. Sustain. Energy Rev. 82 (1),
Hernández-Moro, J., Martínez-Duart, J.M., 2013. Analytical model for solar PV 260–270.
and CSP electricity costs: Present LCOE values and their future evolution. Ratnam, Elizabeth L., Weller, Steven R., Kellett, Christopher M., Murray, Alan T.,
Renew. Sustain. Energy Rev. 20, 119–132. 2015. Residential load and rooftop PV generation: an Australian distribution
Honrubia-Escribano, Andrés, Javier Ramirez, F., Gómez-Lázaro, Emilio, Garcia- network dataset. Int. J. Sustain. Energy.
Villaverde, Pedro M., Ruiz-Ortega, Maria J., Parra-Requena, Gloria, 2018. Ren, Hongbo, Gao, Weijun, Ruan, Yingjun, 2009. Economic optimization and
Influence of solar technology in the economic performance of PV power sensitivity analysis of photovoltaic system in residential buildings. Renew.
plants in Europe. A comprehensive analysis. Renew. Sustain. Energy Rev. 82 Energy 34 (3), 883–889.
(Part 1), 488–501. Segura, E., Morales, R., Somolinos, J.A., López, A., 2017. Techno-economic chal-
IEA-PVPS, 2017. National Survey Report of PV Power Applications in AUSTRALIA lenges of tidal energy conversion systems: Current status and trends. Renew.
2016. Sustain. Energy Rev. 77, 536–550.
326 O. Ellabban and A. Alassi / Energy Reports 5 (2019) 310–326

Singh, Parm Pal, Singh, Sukhmeet, 2010. Realistic generation cost of solar Tarigan, Elieser, Djuwari, Kartikasari, Fitri Dwi, 2015. Techno-economic sim-
photovoltaic electricity. Renew. Energy 35 (3), 563–569. ulation of a grid-connected PV system design as specifically applied to
Sommerfeldt, Nelson, Madani, Hatef, 2017. Revisiting the techno-economic residential in Surabaya, Indonesia. Energy Procedia 65, 90–99.
analysis process for building-mounted, grid-connected solar photovoltaic van Sark, Wilfried, Reich, Nils, Müller, Björn, Armbruster, Alfons, Kiefer, Klaus,
systems: Part two – Application. Renew. Sustain. Energy Rev. 74, 1394–1404. Reise, Christian, 2012. Review of PV Performance Ratio Development. World
Stock, Petra, Stock, Andrew, Bourne, Greg, 2017. State of Solar 2016: Globally Renewable Energy Forum.
and in Australia. climatecouncil.org.au. Vilaça Gomes, P., Knak Neto, N., Carvalho, L., Sumaili, J., Saraiva, J.T., Dias, B.H.,
Talavera, D.L., Nofuentes, G., Aguilera, J., 2010. The internal rate of return of Miranda, V., Souza, S.M., 2018. Technical-economic analysis for the integra-
photovoltaic grid-connected systems: A comprehensive sensitivity analysis. tion of PV systems in Brazil considering policy and regulatory issues. Energy
Renew. Energy 35, 101–111. Policy 115, 199–206.
Talavera, D.L., Nofuentes, G., Aguilera, J., Fuentes, M., 2007. Tables for the Yu, Man, Halog, Anthony, 2015. Solar photovoltaic development in Australia—A
estimation of the internal rate of return of photovoltaic grid-connected life cycle sustainability assessment study. Sustainability 7, 1213–1247.
systems. Renew. Sustain. Energy Rev. 11 (3), 447–466. Zahedi, A., 2010. A review on feed-in tariff in Australia, what it is now and what
it should be. Renew. Sustain. Energy Rev. 14 (9), 3252–3255.