Optimal Management of Micro Grid in Restructured Environment

2012 Second Iranian Conference on Renewable Energy and Distributed Generation
Optimal Management of Micro Grid in Restructured

Environment
Ehsan Nasrolahpour*, Meysam Doostizadeh*, Hassan Ghasemi*, Senior member, IEEE
*School of Electrical and Computer Engineering
*University of Tehran
*Tehran, Iran
Emails: e.nasrolahpour@ece.ut.ac.ir, m.doostizadeh@ece.ut.ac.ir, h.ghasemi@ut.ac.ir
AbstractElectric utility deregulation and power system

restructuring encourage micro grids penetration into distribution
networks. Micro grids are low voltage distribution networks that
are designed to supply electric and heat loads for small
communities. A micro grid which is aggregation of various
distributed generators (DGs), storage devices and controllable
loads, can be operated and dispatched flexibly. Micro grids can
work in an interconnected or autonomous mode. In this paper,
component models of micro grids are determined by using real
data and cost of micro grid is minimized in order to determine
the optimal generation of each DG and the amount of
controllable loads during a day. The problem formulation
includes purchased energy from the main grid, DGs operational
cost, start-up and shut-down costs, and finally the cost of
interrupted loads. Also, the ramp rate and other constrains of
each DG are considered. The micro gird considered in this paper
consists of a wind turbine, a micro turbine, a photovoltaic array
and a fuel cell.
Micro grids can provide energy at lower price because of

their optimal use of waste heat. In addition, they do not pay any
cost for congestion, customer service, distribution and
transmission losses and other costs unlike traditional power
systems. The generation of micro grid should be changed with
regards to the variation of grids energy price. For better
performance, we need to manage and control the micro grid.
A successful management of micro grid must provide a
strong financial incentive for both micro grid owners and the
main gird. In order to manage micro grids, the following
information should be known as input data:
The energy price of main grid at any time
Maximum amount of interruptible load
All characteristics related to DGs such as wind speed,
temperature, irritation and etc.
Keywords- micro grid; component model; DG; energy

management.
Many studies have been done in the area of micro grids,

which some of them deal with optimization of micro grid
operation. Decentralized control of micro grid is discussed in
[4]. Ref. [5] suggested centralized control strategy. [6]
optimizes the operation of a micro grid at only 1 hour and in an
autonomous mode. In [7] optimizing the operation of an
autonomous micro grid is performed in 24 hours of a day but it
is done separately for each hour and there is no link between
them. [8] uses the model of [5] and optimizes the operation of
micro grid in 24 hours by using the genetic algorithm but the
component models of micro grid are not considered. In [9] the
cost function is minimized by MADS (Mesh Adaptive Direct
Search), but MADS alone is not sufficient to reach an optimal
solution, so MADS needs to be combined with SQP
(Sequential Quadratic Programming) to reach the optimal
solution and it should be considered too that MADS and SQP
methods are inherently unable to result in a global solution
[10]. The present paper proposes the component model of
micro grid, then develops the model of [5] and optimizes the
operation of the system considering constrains of network and
DGs. The objective function includes the purchased energy
from the main grid, DGs operational cost, start-up and shutdown costs, and finally the cost of interrupted loads. For
problem solution, the GAMS software is employed.
I.
INTRODUCTION
The need for more flexible and accessible electrical
systems, enhancement of local reliability, environmental and
economical issues derive development of micro grids, which
surely will play a significant role in the future of power system.
Micro grids are defined as intelligent distribution networks
include various distributed generators, storage devices and
controllable loads [1]. Micro grid can be operated in
autonomous mode in condition that it is disconnected from the
main grid or in non-autonomous mode if it is connected to the
main grid. A non-autonomous micro grid, due to its connection
to power utilities, can reach greater controllability and stability.
From the grids point of view, a micro grid is a controlled
entity that can be operated as a single aggregated load and as a
small source of power or ancillary services supporting the
network. On the other hand, micro grids are similar to
traditional local distribution networks that provide electricity,
from the customers point of view [2]. Since the micro grids
are designed to generate power at the voltage level of
distribution networks, they have limited energy handling
capability. Therefore, their maximum capacity is normally
restricted to 10 MVA [3].
978-1-4673-0665-2/12/$31.00 2012 IEEE
116
C.
Fuel Cell Cost

Fuel cell converts the chemical energy into electrical
energy. The following is the function used to calculate the cost
of fuel cell.
This paper is organized as follows: The components model

is discussed in section II. In section III, the objective function
and the constraints of problem are described. The case study is
introduced in section IV. The simulation results are presented
in section V. Finally, the section VI concludes the paper.
(5)
Pj
II. MICRO GRID COMPONENT

Different types of DG technologies have been invented
which are utilized in distribution networks. The output power
of DG in each technology depends on various parameters such
as environmental conditions. In this part, the output power of
each type has been modeled as a function of their input
parameters [9].
where Cnl is the natural gas price; Pj shows the net electrical
power produced at interval j; nj represents the cell efficiency at
interval j.
A.
D.
C fc (t ) = C nl .
Micro Turbines Cost

Micro turbines are widely used in DG systems and as
energy producer in CHP (Combined Heat and Power) systems.
They are small and simple-cycle gas turbine. Micro turbine can
be modeled as fuel cells. The cost of micro turbine can be
expressed as:
Wind Turbine
Wind turbine converts wind energy into electrical one. The
output power of a wind turbine depends on several factors such
as availability of the wind, the wind turbine power curve, wind
speed, size and shape of turbine. The power developed is given
by (1) [11]:
0
Pwt (t ) = a.v(t )3 b.PR

P
R
v < vci , v > vco

vci < v < vr
vr < v < vco
(6 )
Pj
(1)
C mt (t ) = C nl .
OPTIMIZATION PROBLEM
III.
where Pwt and PR denote actual and rated power, respectively; v
and vt are actual and rated wind speed, respectively; vci and vco
present cut-in and cut-off wind speed, respectively; a and b are
obtained from following equations:
b=
a=
vr3
A.
Min.
vci3
OC =
p .P
t
t =1:24
pR
vr3 vci3
Gstc
(1 + k (Tc Tr ))
(C
dg,t .Pdg,t .wdg,t
(7 )
t =1:24 dgS
IL ( PIL,t )
t =1:24
where pt denotes the energy price of main grid; Pgrid,t is the

amount of power exchanged between main grid and micro
grid; Cdg,t represents generation cost of each DG in each hour;
Pdg,t is generation of each DG in each hour; wdg,t ,udg,t ,vdg,t are
binary values that denote commitment status, start up, shut
down cost of each DG in each hour; STCdg and SHCdg indicate
start up and shut down cost; pIL is cost of interruptible load,
PIL,t correspond to amount of interruptible load; S is set of
DGs.
Photovoltaic
Solar PV generates the electricity from free solar energy.
The output power of this module can be calculated as follows:
Ging
grid,t
STCdg .udg,t + SHCdg .vdg,t ) +
(3)
B.
Ppv = Pstc
Obective Function
In order to minimize operation cost and therefore,

maximize social welfare the objective function is defined as
below:
(2)
vci3
(4)
where Ging and Gstc indicate irradiance at actual time and

standard situation, respectively. Gstc is supposed as 1000
(w/m2); Tc and Tr corresponding to cell temperature and
reference temperature, respectively; k is thermal coefficient of
photovoltaic; Ppv and Pstc denote the output power and the
module maximum power at standard test condition (STC),
respectively.
B.
Constraints
The problem constraints such generation limits and
capacity reserve limits are formulate as follows:
1- The supply-demand balancing constraint:
117
Pgrid ,t +
dg ,t
= PD (t ) Pcurt,t
t = 1 : 24
(8)
dgS
2- DG constraints:
Pdg ,t , min Pdg ,t .wdg , t Pdg , t , max t = 1 : 24
(9)
Pdg , t +1 Pdg ,t Rumdg
t = 1 : 24
(10)
Pdg , t Pdg , t +1 Rdmdg
t = 1 : 24
(11)
[Ton,dg,t MUTdg ].[wdg,t1 wdg,t ] 0
t = 1: 24 dg S
(12)
[Toff,dg,t MDTdg ].[wdg,t1 wdg,t ] 0 t =1: 24 dgS
(13)
[ wdg , t wdg , t 1 ] udg , t
(14)
t = 1 : 24 dg S
[ wdg , t wdg , t 1 ] vdg ,t
t = 1 : 24 dg S
Figure 1. Study case LV network [3]
(15)
Figure 2.
[wdg,t wdg,t 1 ] udg,t vdg,t t = 1 : 24 dg S
(16)
Two scenarios are described for energy price of main grid

as shown in Fig. 2. Also, for each scenario two strategies is
considered.
where Pdg,t,min and Pdg,t,max represent the minimum and

maximum generation of each DG in each hour, repectively;
Rumdg and Rdmdg are the ramp up and ramp down limit of each
DG; Ton,dg,t and Toff,dg,t denote the on and off times of each DG,
respectively; MUTdg and MDTdg are minimum up and down
limits of each DG, respectively.
Strategy 1: micro grid can buy energy from main grid but
cannot sell energy to the grid.
Strategy 2: micro grid can exchange power with the main
grid freely and can sell energy to main grid.
3- The constrains of interruptible load:

0 Pcurt m PD
V. RESULTS
The result of simulation has been shown in below figures. It
should be noted that since energy price are different in two
scenarios, demand of the network differ in each scenario. As it
can be seen for the presented results, at the hours that the
energy price of main grid is high, the majority of load demand
is provided by DG units of micro grid.
(17)
where m is the maximum of interruptible load.
Also, at hours that the DGs operational costs are higher than
the energy price, DG units reduce their power generation and
the energy is mostly injected from the grid into micro grid to
diminish total cost of system.
4- Reserve capacity:
dg ,t ,max wdg ,t
+ Pgrid,t + Pcurt,t = Res
(18)
dgS
Furthermore, at pick load hours that the load demand is

more than the generable energy by DGs, and the energy price
of grid is too high, some part of controllable loads are
interrupted so that the system total cost be reduced.
where Res denotes the reserve capacity of micro grid.

IV.
Wholesale market price
CASE STUDY
Since the generation cost of micro turbine and fuel cell is

low in comparison to other DGs, and also their generated
power is independent from environmental conditions, thus,
most of the generated energy by DGs is dedicated to them.
A.
System Description
The micro grid shown in Fig. 1 is used as the test case,
which consists of photovoltaic array, a wind turbine, a fuel cell
and one micro turbine. The characteristics of above
components are given in Tables A. I-III.
Table I compares the total cost of system in two strategies

and two scenarios. As seen, in each scenario, the cost of
strategy 2 is lower than that of strategy 1. This can be
118
concluded from the fact that in scenario 2, DG units after

meeting the load demand, they can sell their extra energy to the
main grid so that all their capacity is utilized.
reason for this matter is that in the second scenario, the

cheapness of wholesale market energy is continued for more
hours than that for first scenario (see Fig. 2).
VI. CONCLUSION
Based on component model of distributed energy
resources, environmental conditions, energy price of main
grid, maximum amount of interruptible load and other
constraints related to the system, this paper proposes
optimization and dispatch model of micro grids. The proposed
model is validated by simulation of a micro grid by GAMS
software. The simulation results are discussed in different
interactive operational control strategies and market prices.
The results have verified the effectiveness of the control and
management method in optimized operation of micro grids.
Figure 3. First senario and first strategy
APPENDIX
Figure 4. First senario and second strategy
Figure A.1 Wind speed during a day
Figure 5. Second senario and first strategy
Figure A.2 Sun irritation during a day
Figure 6. second senario and second strategy

Table I. total system cost in different strategies and scenarios
Strategy 1
Strategy 2
First scenario
248.081 ()
226.472 ()
Second scenario
169.063 ()
131.591 ()
Figure A.3 Temperature during a day
Also, it is observed that in each strategy, total system cost

in the second scenario is less than that of first scenario. The
119
Table A.I Wind turbine specifications
[2]
VR (m/s)
13.8
PR (kW)
300
Vci (m/s)
3.1
Vco (m/s)
25
T. Logenthiran, D. Srinivasan, A. M. Khambadkone, Multi-agent

system for energy resource scheduling of integrated microgrids in a
distributed system, Electric Power Systems Research, vol. 81, pp. 138148, Jan. 2011.
[3] R. Lasseter, A. Akhil, C. Marnay, H. Stephens, J. Dagle, R. Guttromson,
A.S. Meliopoulous, R. Yinger, J. Eto, The CERTS Microgrid Concept,
White Paper for CERTS, 2002.
[4] A. L. Dimeas, N. D. Hatziargyrious, Operation of a multiagent system
for microgrid control, IEEE Trans. Power Syst., vol.20, no.3, pp. 1447
1455, Aug. 2005.
[5] A. G. Tsikalakis, N. D. Hatziargyriou, Centralized control for
optimizing microgridsoperation, IEEE Trans. Energy Conv. vol.23,
no.1, pp. 241-248, Mar.2008.
[6] C.A. Hernandez-Aramburo, T.G. Green, N. Mugniot, Fuel consumption
minimizationof a microgrid, IEEE Trans. Ind. Appl., vol 41, no.3, pp.
673681, May 2005.
[7] F.A. Mohamed, H.N. Koivo, System modeling and online optimal
management of microgrid using multiobjective optimization,
International Conference on Clean Electrical Power, Capri, Italy, May
2007.
[8] E. Mashhour, S.M. Moghaddas-Tafreshi, Integration of disributed
energy resources into low voltage grid: A market-based multiperiod
optimization model, Electric Power System Research. vol.80, pp. 473480, Apr. 2010.
[9] F. A. Mohamed, Heikki N. Koivo System modelling and online optimal
management of Micro grid using Mesh Adaptive Direct search
Electrical Power and Energy Systems. vol. 32, pp.398-407, Jun. 2010.
[10] R. Bhuvaneswari, C. S. Edrington, D. A. Cartes, S.Subramanian,
Online Economic Environmental Optimization of a Micro grid Using
an Improved Fast Evolutionary, North American Power Simposium
(NAPS), 2009.
[11] R. Belfkira, G. Barakat, C. Nichita Sizing optimization of a stand-alone
hybrid power supply unit: wind/PVsystem with battery storage,
International Review of Electrical Engineering (IREE) vol. 3, no. 5,
Mar. 2008.
Table A.II Photovoltaic panel specification

Irritation at (stc)
1000
Termal coefficient
-0.004
Enviroment temperature ( c)
25
Table A.III Installed DG sources

Unit ID
Min power (kw)
Max power (kw)
MT
30
FC
30
WT
15
PV1
13
REFERENCES
[1]
N. W. A. Lidula, A. D. Rajapakse, Microgrids research: A review of

experimental microgrids and test systems, Renewable and Sustainable
Energy Reviews, vol. 15, pp. 186-202, Jan. 2010.
120

Optimal Management of Micro Grid in Restructured Environment

Uploaded by

Copyright:

Available Formats

Optimal Management of Micro Grid in Restructured Environment

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

Optimal Management of Micro Grid in Restructured Environment

Uploaded by

Copyright:

Available Formats

2012 Second Iranian Conference on Renewable Energy and Distributed Generation

Optimal Management of Micro Grid in Restructured

AbstractElectric utility deregulation and power system

Micro grids can provide energy at lower price because of

Keywords- micro grid; component model; DG; energy

Many studies have been done in the area of micro grids,

978-1-4673-0665-2/12/$31.00 2012 IEEE

Fuel Cell Cost

This paper is organized as follows: The components model

II. MICRO GRID COMPONENT

Micro Turbines Cost

Pwt (t ) = a.v(t )3 b.PR

v < vci , v > vco

dg,t .Pdg,t .wdg,t

where pt denotes the energy price of main grid; Pgrid,t is the

STCdg .udg,t + SHCdg .vdg,t ) +

In order to minimize operation cost and therefore,

where Ging and Gstc indicate irradiance at actual time and

1- The supply-demand balancing constraint:

Pdg ,t , min Pdg ,t .wdg , t Pdg , t , max t = 1 : 24

Pdg , t +1 Pdg ,t Rumdg

Pdg , t Pdg , t +1 Rdmdg

[Ton,dg,t MUTdg ].[wdg,t1 wdg,t ] 0

[Toff,dg,t MDTdg ].[wdg,t1 wdg,t ] 0 t =1: 24 dgS

[ wdg , t wdg , t 1 ] udg , t

[ wdg , t wdg , t 1 ] vdg ,t

Figure 1. Study case LV network [3]

[wdg,t wdg,t 1 ] udg,t vdg,t t = 1 : 24 dg S

Two scenarios are described for energy price of main grid

where Pdg,t,min and Pdg,t,max represent the minimum and

3- The constrains of interruptible load:

where m is the maximum of interruptible load.

+ Pgrid,t + Pcurt,t = Res

Furthermore, at pick load hours that the load demand is

where Res denotes the reserve capacity of micro grid.

Wholesale market price

Since the generation cost of micro turbine and fuel cell is

Table I compares the total cost of system in two strategies

concluded from the fact that in scenario 2, DG units after

reason for this matter is that in the second scenario, the

Figure 3. First senario and first strategy

Figure 4. First senario and second strategy

Figure A.1 Wind speed during a day

Figure 5. Second senario and first strategy

Figure A.2 Sun irritation during a day

Figure 6. second senario and second strategy

Also, it is observed that in each strategy, total system cost

Table A.I Wind turbine specifications

T. Logenthiran, D. Srinivasan, A. M. Khambadkone, Multi-agent

Table A.II Photovoltaic panel specification

Table A.III Installed DG sources

Min power (kw)

Max power (kw)

N. W. A. Lidula, A. D. Rajapakse, Microgrids research: A review of

You might also like