You are on page 1of 10

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `


Coordination of Energy Storage Units and Distribution Network

Loading Management by Distributed Control strategy
C.Prakash1, DR. Rathanakumar2
1 2
PG Scholor, Professor, Department of Electrical and Electronics Engineering, SCSVMV University,
Enathur, Kanchipuram- 631561

This project proposes a distributed control approach to coordinate multiple energy storage units
(ESUs) to avoid violation of voltage and network load constraints ESU as a buffer can be a promising
solution which can store surplus power during the peak generation periods and use it in peak load
periods.In ESU converters both active and reactive power are used to deal with the power quality
issues in distribution network ESU’s reactive power is proposed to be used for voltage support, while
the active power is to be utilized in managing network loading.

I.Introduction high load and high generation periods,

As a sustainable solution for future causes the noted issues. As a result, the
energy crisis, it is anticipated that future introduction of energy storage unit (ESU)
distribution networks will see a wide- as a buffer can be a promising solution
spread use of renewable energy sources which can store surplus power during the
such as PV, wind turbine and fuel cell. peak generation periods and use it in peak
Distribution networks with renewable load periods. The main challenge in the
energy sources can encounter two main utilization of multiple ESUs is the
challenges. A typical load curve for NSW coordination control strategy. There are
shows that during the peak load period, three types of coordination strategies that
generation is normally low or zero, which can be taken. The strategy can be
may cause voltage drop along the provided through centralized manner in
network. On the other hand, in peak which a central controller coordinates
generation period, when generated power ESUs. The drawback of this approach is
exceeds the load, surplus power is that it would require extensive data base
injected to the grid. This will cause with high speed and fast calculating
reverse power and hence may result in computers, along with broadband
voltage rise along the network. networks. This can be too expensive for
Additionally, in both peak generation and the current state of art .This can also be
peak load periods, thermal constraints for less reliable due to communication failure
line and power transformer can be and computer freezing. The second
violated. approach is the localized control strategy,
The unbalance between the based on local measurements only. This
generated power and load, during both the control strategy is robust in the sense that

ISSN: 2395-1303 Page 112

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `
only local measurements are utilized.It avoid any upper permissible voltage limit
cannot effectively utilize all available violation.
resources in the network due to the lack
of broader information. A distributed
control strategy, the third approach, can
be as efficient as a centralized approach
while avoiding its drawbacks. The
robustness of this approach still depends
on the communication links.
The approach which can
coordinate multiple ESUs to manage and
control voltage and loading in distribution A.single-phase radial LV network
networks. As potential needs quick and
robust control, a combined localized and Inverter injects the surplus active power
distributed control approach is proposed in the DC link to the AC grid. In the
to regulate the ESUs reactive power to proposed RE system structure, there are
deal with voltage issues. It based on three main resources which can be used
consensus algorithm is proposed to by customer to deal with voltage rise.
manage network loading, which divides These possessions include PV inverter
the required active power equally among active (Pinv.i) and reactive (Qinv.i)
ESUs with respect to their maximum power and ESU active power (PESUi).
available active power. The proposed coordination approach uses
these three resources to manage voltage
II.Robust Resource Control rise. As shown, the first and second
LV networks typically are resources are coordinated by a localized
designed to allow a maximum of 5–10% control strategy to have a robust voltage
(depending on national standards) voltage reduction. To make the voltage reduction
drop from LV transformer secondary to more efficient, the charging of distributed
the last customer. Consider the single- ESUs along the network will be
phase radial LV network with PV coordinated using a distributed control
connection for each customer. strategy based on a consensus algorithm.
Transformer tap setting is designed in
such a way that the voltage level in
critical node is within the standard limits
in normal operation condition. However,
as more and more PVs get connected to
the LV networks, significant reverse
power flow ensues, especially as the PV
penetration level increases. To pass up
this problem, this paper proposes a new
robust and efficient approach for RE system for customer with ESU
customers to coordinate their resources to

ISSN: 2395-1303 Page 113

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `
B.Localised Control Strategy
In contained control strategy
will be used to coordinate the contribution
of PV inverters for voltage support using
their active and reactive power. Different
localized techniques are proposed for
voltage support based on reactive power
control. It is to be noted if a PV inverter
VA rating is increased by just 11.8%, it
will have the ability to supply 50%
reactive power while supplying full rated
power. Consequently, customers can use
this capability for voltage reduction. The
category which use inverter reactive
power for voltage decrease can be
summarized as follows:
Fixed reactive power control: In this
strategy, the inverter bus voltage is
monitored. If the bus voltage goes above
a predetermined limit, PV inverter starts Communication structure
to absorb a constant reactive power.
Reactive power control as a function of D.Control Objective:
inverter bus potential: This strategy is
based on fixing the inverter bus voltage to Radial LV network, the last
a desired value by absorbing or injecting node is the critical node. If the customers
reactive power in this node do not experience any voltage
C.Distributed control srategy limit violation, all other customers on the
Consensus algorithm is same phase will not face voltage problem.
projected for distributed control method The coordinated control strategy is
and to coordinate ESUs within the LV illustrated in fig, based on the voltage of
network for efficient voltage reduction this node. The chronological course of
while avoiding curtailment. Since single- events is described as follows
phase PVs in a particular phase can cause Between t=t2 and t=t2, the critical node
voltage rise in that phase, the algorithm is voltage is within desirable voltage range.
applied in per phase basis. Consider a Consequently, the network is in normal
phase that has n distributed ESUs. operation condition. In this instant, all PV
inverters are injecting maximum available
active power and their reactive power is
based on characteristics. Moreover, ESUs
are controlled by customers EMS.
At t=t2, the upper critical voltage limit is
violated in the critical node. Therefore,
PV inverters, whose bus voltages pass

ISSN: 2395-1303 Page 114

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `
this, start to curtail their active power
base. In summing, the leader initiates the
communication among ESUs to find their
contribution for voltage reduction.

Based on the communication speed, at

t3+∆t, the ESUs reduce the voltage of
critical node below the upper critical
Two general objectives need to
limit, which results in preventing RE
be achieved to coordinate ESUs when
required. The first is to design a control
At t=t4, voltage of the critical node falls
forui of each ESU to reduce the network
back in the desirable range which means
apparent power to less than its critical
the voltage of all nodes is in desirable
limits. In other words, in peak load
range. As a result, VC flag becomes 0 and
period, ESUs need to be coordinated to
ESU control strategy is changed back to
bring the value ofS(t) to less
function based on customer’s EMS
thanSload_critical, as shown in equation
Similarly, in peak generation period
equation needs to be met.
S (t ) < S load _ critical (1.1)

S( t ) < Sgen _ critical (1.2)

The second objective is to
design a control for ui to share the
F.Distributed Control Scheme required active power according to the
The distributed control strategy for each same ratio among ESUs, as shown in
ESU can be written in a general form as: equation (1.5).
ui = fi(ci0(t ).ε 0(t ), ci(t ),...,cin(t ).εn(t )) PESU1 PESU 2 PESU n
wherecij(t) denotes the communication = = (1.3)
link betweenith andjth ESUs,cij =1; ifjth
ESU send information toith one, In real case, the interaction
otherwisecij =0. In addition, ci0=1; if among ESUs occurs at discrete time step
theith ESU can get information from the td. Therefore, the information state of
leader, otherwise ci0=0.cii=1 for all ESUs each ESU is updated using equation (3.6).
This time-varying coefficients ε i ( t ) = Σ d ij .ε j ( t - t d ) (1.4)
can be represented as a matrix presenting
the whole communication topology as in C ij (t - t d )
ΣC d ij = (1.5)
equation . n
j =0 C ij (t - t d )

PESU ( t ) = εi( t ) × PESUi (1.5)

III.Proposed approach of distributed

control strategy

ISSN: 2395-1303 Page 115

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `
The dashed arrows the µi [t ] = sij .µi[t ] + ∑ sij [t ].µ j [t - td ] `
information flow, where the neighbouring j = Ni
ESUs are communicated to coordinate
their operation. The proposed internal V.Proposed localized voltage control
control structure for each ESU is shown for each ESU
in Fig. The reference value for ESU’s
active and reactive power (PESUi(t) and With the proposed communication
QESUi(t)) depend in information state of structure, the customers are only aware of
each neighbor. As noted before, the the value of Dji corresponding to their
proposed control includes voltage and neighbours. Therefore, the sensitivity
network loading management. matrix is modified as:
 D i ∈ {N i ∪ j}
D ji =  ji
 0i ∉ {N i ∪ j}
Finally the transition weights are
calculated by the following equation:
.Dij .c ji [t ]
S ij [t ] =
∑ Dki .cki [t ]
j =0
As a result, the weights are predetermined
IV.Distributed control strategy flow for eachESU. If either of the critical
voltage limits is violated for any ESU, its
chart localizedcontrol term initiates the
distributed control strategy based on
u i ( t ) = k qi ( v max_ critical - Vi ( t ) )
u i ( t ) = k qi ( v min_ critical - Vi ( t ) )
Following the control
initiation, the information states of ESUs
are updated in each discrete time interval
based on equation. Finally, the required
reactive power contribution of each ESU
at each time interval is updated by:
Q ESU i [ t ] = μ i [ t ] × Q ESU i

By applying this control

structure, both objectives for voltage
support can be achieved and ESU
inverters will contribute with their
reactive power.
ε i ( t ) = Σ d ij (t - t d ).ε j ( t - t d )

ISSN: 2395-1303 Page 116

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `
Q ESU i [ t ] = μ i [ t ] × Q ESU i (1.7) PV generation and the three energy
storage units are considered. Let the
Pesu i [ t ] = ε i [ t ] × PESU i (1.8) energy storage unit (ESU-1) is placed in
between the PV-1 and PV-2, the ESU-2 is
Q ESU i [ t ] = Q ESU i ( t ) (1.9)
placed in between PV-4 and PV-3 and the
PESU i [ t ] = PESU i ( t ) (1.10) last ESU-3 is placed in between PV-4 and
Where equations 1.7 and 1.8 model the The general simulation diagram is shown
proposed distributed control strategy for in figure. In occurrence of voltage issues
ESUs, equations 1.9 and 1.10 model the and the overload problems, the energy
ESUs internal control. It is worth noting storage units are coordinated with the
that the internal dynamics of network are principle of distribution control strategy.
not considered here. So, it can be said that With the coordination of energy storage
the value of these variables diminish unit we can overcome the voltage issues
much faster than the output power. As a and network problems in the
result, the dynamic of ESU output power distributionnetwork. The block diagram
is determined by the controller designed of subsystem of the energy storage unit is
in this paper while the inner dynamic of shown in next figure. This subsystem
ESUs is ignored. Consequently, the ESUs consists of inverter and the output of the
reactive and active power are modelled whole subsystem is connected with the
VI.Proposed Similation Diagram For grid for overcome the voltage issues.
Coordination Of ESU Similarly, the identical subsystem is
The proposed distribution communicated with the each and every
control strategy architecture is developed subsystem in the station.
with the coordinated communication link.
In order to develop this architecture five

General Simulation Diagram

ISSN: 2395-1303 Page 117

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `

Simulation of Energy storage unit Subsystem

All the energy storage units are

interconnected with the communication The details about the parameter of the
leader corresponding to the distributed system such as PV generation ratings and
control strategy. The resulting the available active power for each
waveforms are presented, analyzed with generation units are shown in tables. The
the active and reactive power with permissible and critical voltage limits
respect to time. and the network loading limits are shown
in tables below.
a.System Parameters: b. Available active power for ESU
Parameter Voltage (p.u)
voltage limit
V max _permissible Active 550 600 500 450 650
max _critical 1.0
V max _desirable

ISSN: 2395-1303 Page 118

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `
c. PV’s ratings
Active 150 200 250
d. Network loading limits
Parameter Power (kva)
Energy storage unit bus voltage
S max _permissibl e 2200 waveform

S max _critical 1800 The waveform for the

energy storage unit voltage power factor
S max _desirable 1500 with respect to time is shown in Figure
The voltage is measured in terms of p.u.
and it keeps on increasing monotonously.
It can be seen between t= 0s
VII. ANALYSIS OF SIMULATION and t= 100s, all voltages and network
RESULTS loading are in the desirable range.
The proposed power system Therefore, no ESUs coordination is
model has been simulated in MATLAB / needed. However, at t=100s, as the PV
Simulink. The resulting waveforms for generation increases, the upper critical
active power and reactive power are limit for voltage of ESUs 2 and 3 and
shown network loading is violated. Therefore,
a. Simulation Result For ESU’s Bus the proposed control approach for both
Voltage And Network Loading Profile voltage and thermal constraints
The simulation results are management is initiated. To network
discussed below for both ESU’s bus loading management, it can be, at the
voltage and reactive power on basis of equilibrium point, its value is less than
distribution control strategy. Consider upper critical limit.
first the waveform for the active power.
In addition, it is stipulated that the power
factor of ESUs should be more than 0.9
and the upper limit of the reactive power
output is dependent of the active power
output as
Q max
ESU = 0 .4843 × PESUi i = 1,2,3.... (1.11)

It is assumed that all loads are Power factor waveform of ESU

in 15% of their maximum. In addition,
the PVs generation change from 75% to b.Simulation Results For ESU’s Active
95% at t=100 s. And Reactive Power
Next, the resulting waveforms
for each Energy storage units active

ISSN: 2395-1303 Page 119

International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `
power and reactive power according to voltage, reactive power, real power is
the distribution control strategy. In this reached the stability condition.
case, the waveform reaches the stable Therefore, the main benefits of using this
condition and as shown below. Figure type of control has attained the steady
shows the active power of the generators state region earlier when compared with
using the controller. In this figure, the some other types of controlling strategy.
power first fluctuates and then aches a Conclution:
stable state Distribution control strategy is a new
approach to coordinate multiple Energy
storage units to manage voltage and
loading in distribution networks.
This method is designed to use the most
adjacent ESUs to the violated bus
voltage and for loading management, a
distributed control strategy based on
consensus algorithm is employed to
Energy storage unit active power coordinate ESU’s active power.
waveform The proposed distribution control
strategy has been designed to share the
required active power with the same
ratio among ESUs with respect to their
available active power. The proposed
structure can support any of the
precieved options for the future network
Energy storage unit reactive
Power waveform
1. Anna R. Di Fazio, Giuseppe Fusco,
Figure shows the waveforms of the
and Mario Russo "decentralized control
Energy storage unit reactive power for
of distributed generation for voltage
the three energy storage unit. The active
profile optimization in smart feeders,"
power is measured in terms of p.u. and it
IEEE Trans. Smart Grid, vol.4, no.3,
attains the stable condition due to the
pp.1586-1596, Sep. 2013.
presence of the distribution control
QESU 3 QESU 2 QESU1 2. Augusto C. Rueda-Medina and
< max
< max Antonio Padilha-Feltrin "distributed
3 2 1
generators as providers of reactive power
After the generation step change, the support a market approach," IEEE Trans.
ESU voltages follow the pattern based on Power Systems, vol.28, no.1, pp.490-
equation. Accordingly, reactive power 502, Feb. 2013.
sharing among ESUs needs to follow as 3. Edson Luiz da Silva, Jonathan J.
in to have effective voltage support. Hedgecock, Member, João Carlos O.
From the analysis of simulation results, it Mello, and João Carlos Ferreira da Luz
concluded that with the presence of "practical cost-based approach for the
distributed control strategy the terminal
ISSN: 2395-1303 Page 120
International Journal of Engineering and Techniques - Volume 3 Issue 2, March-April 2017 `
voltage ancillary service," IEEE Trans. ‘Economic and Efficient Voltage
Power Systems, vol.16, no.4, pp.806- Management Using Customer-Owned
812, Nov.2001. Energy Storage Systems in a Distribution
Network With High Penetration of
4. Guerrero. J. M, Kai. S, Li. Z and Yan.
Photovoltaic Systems,’ Power Systems,
X. (2011), ‘A Distributed Control
IEEE Transactions on, vol. 28, pp. 102-
Strategy Based on DC Bus Signaling for
Modular Photovoltaic Generation
Systems With Battery Energy Storage,’ 10. Teng. J. H, Luan. S. W, Lee. D. J and
Power Electronics, IEEE Transactions Huang. Y. Q. (2013), ‘Optimal
on, vol. 26, pp. 3032-3045. Charging/Discharging Scheduling of
Battery Storage Systems for Distribution
5. Hideharu Sugihara, Kohei Yokoyama,
Systems Interconnected with Sizeable
Osamu Saeki, Kiichiro Tsuji, and
PV Generation Systems,’ Power
Tsuyoshi Funaki "economic and efficient
Systems, IEEE Transactions on, vol.28,
voltage management using customer-
pp. 1425-1433.
owned energy storage systems in a
distribution network with high
penetration of photovoltaic systems,"
IEEE Trans. Power Systems, vol.28,
no.1, pp.102-111, Feb. 2013.
6. Katherine M. Rogers, Ray Klump,
Member, HimanshuKhurana, Angel A.
Aquino-Lugo, and Thomas J. Overbye
"an authenticated control framework for
distributed voltage support on the smart
grid," IEEE Trans. Smart Grid, vol.1,
no.1, pp.40-47, June 2010.
7. Mokhtari. G, Zare. F, Nourbakhsh. G,
and Ghosh. A. (2012), ‘A new DER
coordination in LV network based on the
concept of distributed control,’ in Power
Electronics for Distributed Generation
Systems (PEDG), pp. 1-8.
8. ReinaldoTonkoski, Dave Turcotte, and
Tarek H. M. EL Fouly "impact of high
pv penetration on voltage profiles in
residential neighborhoods," IEEE Trans.
Industry Sustainable Energy, vol.3, no.3,
pp.518-527, July 2012.
9. Sugihara. H, Yokoyama. K, Saeki. O,
Tsuji. K and Funaki. T. (2013),

ISSN: 2395-1303 Page 121