Aalborg Universitet
Evaluating Maximum Wind Energy Exploitation in Active Distribution Networks
Siano, Pierluigi; Chen, Peiyuan; Chen, Zhe; Piccolo, A.
Published in:
IET Generation Transmission and Distribution
DOI (link to publication from Publisher):
10.1049/iet-gtd.2009.0548
Publication date:
2010
Document Version
Publisher's PDF, also known as Version of record
Link to publication from Aalborg University
Citation for published version (APA):
Siano, P., Chen, P., Chen, Z., & Piccolo, A. (2010). Evaluating Maximum Wind Energy Exploitation in Active
Distribution Networks. IET Generation Transmission and Distribution, 4(5), 598-608. https://doi.org/10.1049/iet-
gtd.2009.0548
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
- Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
- You may not further distribute the material or use it for any profit-making activity or commercial gain
- You may freely distribute the URL identifying the publication in the public portal -
Take down policy
If you believe that this document breaches copyright please contact us at vbn@aub.aau.dk providing details, and we will remove access to
the work immediately and investigate your claim.
Downloaded from vbn.aau.dk on: August 10, 2022
Published in IET Generation, Transmission & Distribution
Received on 29th September 2009
Revised on 18th January 2010
doi: 10.1049/iet-gtd.2009.0548
ISSN 1751-8687
Evaluating maximum wind energy exploitation
in active distribution networks
P. Siano
1
P. Chen
2
Z. Chen
2
A. Piccolo
1
1
Department of Information and Electrical Engineering, University of Salerno, Fisciano, Italy
2
Department of Energy Technology, Aalborg University, Aalborg 9220, Denmark
E-mail: psiano@unisa.it
Abstract: The increased spreading of distributed and renewable generation requires moving towards active
management of distribution networks. In order to evalua te maximum wind energy exploitation in active
distribution networks, a me thod based o n a multi-period optimal power flow analysis is proposed. Active
network management schemes such as coordinated vo ltage control, energy curtailment and power factor
control are integrated in the method in order to investigate their impacts on the maximisation of wind energy
exploitation. Some case studies, using real data from a Danish distribution system, confirmed the effectiveness
of the proposed method in evaluating the optimal applications of active management schemes to increase
wind energy harvesting without costly network reinforcement for the connection of wind generation.
1 Introduction
The international concern over climate change is driving
European countries to reduce carbon-dioxide emissions by
means of political and regulatory pressure and to increase
the total electrical supply energy from renewable sources.
Electricity market liberalisation and the priority given to
renewable sources under EU directive 03/54/EC along with
the worldwide promotion of renewable encourage the
development of the distributed generation (DG) and
renewable sources.
The connection of large amounts of DG to distribution
systems presents a number of technical challenges to
distribution network operators (DNOs) [1–6]. These
challenges are partly caused by the mismatch between the
location of energy resources and the capability of local
networks to accommodate new generation. Particularly, the
location of wind turbines (WTs) is determined by the local
wind resources and geographical conditions. However, the
current capacity of the network to which the WTs will be
connected may not be sufficient to deliver the generated
wind power. As a result, network reinforcement needs to
be planned by the DNOs. Since such network
reinforcement usually calls for high capital investment,
DNOs would like to explore less costly means that can
improve the capability of the network to accommodate new
generation. One way is to ma ke the best use of the existing
network by encouraging development at the most suitable
locations [3–6]. In order to do this, DNOs require a
reliable and repeatable method of quantifying the capacity
of new DG that may be connected to distribution networks
without the need for reinforcement.
The challenge of identifying the best network location and
capacity for DG has attracted significant research effort,
albeit referred to by several terms: optimal ‘capacity
evaluation’ [3–7], ‘DG placement’ [8] or ‘capacity
allocation’ [9–11]. These optimisation problems apply
different numerical algorithms with various objectives and
constraints. For example, genetic algorithms are used to
find the optimal location of DG [12–14]. Several other
algorithms are adopted to handle optimisation problems
with discrete variables [9, 15]. Other approaches require
network locations of interest to be pre-specified with
algorithms guiding capacity growth within network
constraints [7–9, 16].
Nevertheless, as values associated with WTs are time- and
location-dependent, methods that simply consider one
specific power value at a specific moment are not able to
account for time dependence. Therefore, WTs optimal
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 5, pp. 1 – 11 1
doi: 10.1049/iet-gtd.2009.0548
&
The Institution of Engineering and Technology 2010
www.ietdl.org
IET Corrected Proof
allocation should consider their capability of delivering power
at the right time and WTs should be located at the right place
to be able to deliver energy while satisfying network
constraints. Simulating load and generation variations
during a year and computing the WTs delivered energy
allows including the time dimension, when compared with
methods that simply consider the power at one specific
point in time [16]. In order to account for load and
generation time interdependences, some approaches focused
on the concept of energy from DG [1, 17]. In [1], a
method that maximises the amount of energy that may be
reaped from a given area, on the basis of the available
energy resources, and minimises DG connection costs,
losses and technical constraints has been proposed and
implemented. In [17], a method that evaluates annual
energy in order to measure the risk of unserved energy for
each planning option is described. The added value of
computing the annual energy in determining the extent of
the distribution planning problem is demonstrated.
However, active management schemes are not considered
in either method to exploit higher energy from DG.
In this work, the maximum wind energy exploitation in a
distribution network is evaluated under different active
management schemes during a given time horizon. The
evaluation is based on a multi-period optimal power flow
(MP-OPF) algorithm, which takes into account
distribution network constraints. Such a MP-OPF, derived
from the OPF methods of [3–6, 18] considers the time-
varying characteristics of the load demand and wind power
generation. The algorithm also integrates active
management schemes such as coordinated voltage control,
energy curtailment and power factor control. The analyses
are demonstrated using a 69-bus 11 kV radial distribution
network. Section 2 describes the active management
schemes adopted in the MP-OPF which is described in
Section 3. Sections 4 and 5 present and comment some
case studies. A discussion on the presented results is given
in Section 6. Conclusions are drawn in Section 7.
2 Evaluating maximum wind
energy exploitation in active
distribution networks
Active management represents an alternative approach to
enable national targets for renewable energy and increase
the penetration of WTs into the existing distribution
networks [19, 20]. It has, indeed, the potential to maximise
DG penetration level while minimising DG-related
network reinforcements [21–30]. In [28], it is
demonstrated that networks endowed with active
management schemes can potentially accommodate up to
three times as much generation.
Active management can be realised, for instance, through
generation dispatch, transformer tap adjustment or reactive
power compensators. In [20], WT’s generation curtailments
during low demand, reactive power management using a
reactive compensator and area-based on-load tap-changer
(OLTC) coordinated voltage control have been used in the
active management.
In [21, 22], a multi-period steady-state analysis for
maximising the capacity of wind generation through an
OPF-based technique with active ma nagement features has
been proposed. However, since wind capacity rather than
wind energy is maximised, WTs allocation does not allow
maximum wind energy exploitation. Moreover, short-circuit
level is computed with a simplified approach.
The more advanced and emerging concept of active
management is based on real-time measurements of the
distribution network parameters and employs real-time
control of generators, tap-changing transformers, reactive
power compensators and communication among the
generators and voltage control devices [29].
The MP-OPF proposed here improves the methods
proposed in [18, 21, 22] by accounting for load and
generation time interdependences and by focusing on the
concept of energy from WTs. The proposed method
allows, in fact, finding the optimal WTs capacities
allocation in order to maximise wind energy exploitation
under different active management schemes, briefly
described in the following section.
2.1 Coordinated OLTC voltage control
Traditional control strategies of OLTCs are either based on
the voltage regulation at a single busbar or voltage drop
compensation on a particular line [20]. Such voltage
control strategies are based on local measurements and are
suitable for traditional distribution systems with
unidirectional power flow. However, these strategies may
cause problems in distribution networks with bi-directional
power flows. On the other hand, the area-based control
strategy of OLTCs is based on measurements from various
locations of the network. In this way, the voltage regulation
of OLTCs can be based on the voltage information of the
bus that has the most severe over-voltage problem [20].
Consequently, the maximum wind energy penetration level
may be increased by the implementation of the control
strategy.
2.2 Energy curtailment
In order to alleviate the over-voltage problem, it may be
necessary to curtail a certain amount of wind energy
injected into the network [20]. Although the output wind
energy is reduced, the WT developer may still gain more
profits due to the possibility of installing more WTs [29].
In the proposed method, wind energy may be curtailed
during certain periods in order to alleviate any voltage or
thermal constraint violation. For example, for a specific
2 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 5, pp. 1 – 11
&
The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2009.0548
www.ietdl.org
period, there are different possible combinations of load
demand and wind power. Wind energy is curtailed at the
combination of minimum demand and maximum wind
power. The same strategy is applied to each of the periods
analysed.
In the method, energy curtailment is implemented in each
period by introducing a negative generation variable to
represent the curtailed energy from each WT. For a given
period, the maximum energy that can be curtailed from a
given WT is set to a fraction of the potential energy that
the WT could have produced without energy curtailment.
2.3 Coordinated generator reactive
power control
The recent grid codes of many countries, such as Denmark,
Germany, Italy, Ireland and the UK, require that WTs
should provide reactive power control capabilities and that
network operators may specify power factor or reactive
power generation requirement for grid-connected WTs [31].
In practice, a grid-connected WT needs to fulfil the
specific requirement depending on the regulation of the
country. For example, in the Danish grid code for grid-
connected WTs, reactive power generation is confined to a
control band with respect to active power generation (with
a power factor between 1.00 and 0.995 lagging). The
German grid code specifies different reactive power limits
according to voltage value at interconnection (with a power
factor ranging between 1.00 and 0.925 lagging). The Irish
grid code requires a power factor between 0.835 leading
and 0.835 lagging when the active power output level is
below 50% of the rated capacity. In Italy and the UK, the
power factor at a WT’s terminal should be between 0.95
leading and 0.95 lagging.
Although it is important to fulfil the grid code when
connecting a WT, this paper intends to illustrate the
concept of the proposed method, but not to design a WT
that fulfils a specific requirement. WTs, especially those
with power electronic controllers, are able to provide
necessary reactive power support to the grid. The reactive
power generation can be dispatched centrally by the DNOs
[32]. In other words, power factors of WTs can be
controlled so that wind energy penetration level in the
network is maximised. The proposed control scheme
requires WTs to generate reactive power during load peak
hours and low generation, and to absorb reactive power
during load off-peak hours and high generation.
3 Multi-period optimal
power flow
The optimisation method aims to find the optimal locations
and capacities of WTs so that the wind energy exploitation in
the network is maximised. Such an objective is subject to a
number of technical constraints imposed by regulations,
including bus voltage limits, line/transformer thermal limits
and system short-circuit levels. By fulfilling these
constraints, the network reinforcement due to the
connection of WTs may be avoided. In addition, such a
method can be used to investigate the impact of the
foregoing active management strategies on the maximum
wind energy penetration level in the network.
The proposed approach, based on the non-linear
programming formulation of the MP-OPF described in
[18, 21, 22], has been modified in order to maximise the
wind energy exploitation and to include active management
schemes, the time-varying characteristics of the load
demand and wind power generation and the system short-
circuit constraints.
The MP-OPF is formulated as
maximise
N
j
j=1
N
G
g=1
E
j
g
(P
g
, x
j
)
subject to
h(x
j
) = 0
g(x
j
) ≤ 0
(1)
where E
j
g
(P
g
, x
j
) is the wind energy generated during the
time period j by the g
th
WT with rated capacity P
g
, N
j
is
the total number of periods in a year corresponding to
different combinations of load demand and wind power
generation and N
G
is the number of WTs (indexed by g).
The vector x
j
consists of a set of controllable quantities and
dependent variables during each period j. The optimisation
variables include the capacity of each WT, and for each
period j, the secondary voltage of the OLTC, the power
factor angle, the curtailed energy of each WT and the
import/export power at the interconnection to the external
network.
The equality constraints h(x
j
) represent the static load flow
equations such as Kirchhoff current law ∀j [ J and ∀b [ B,
where J is the set of periods (indexed by j), B is the set of
buses (indexed by b) and Kirchhoff voltage law, ∀j [ J and
∀l [ L, where L is the set of lines (indexed by l ).
The inequality constraints g(x
j
) are listed in the following:
† Capacity constraints for the interconnection to external
network (slack bus) ∀j [ J, ∀x [ X
P
−
x
≤ P
x, j
≤ P
+
x
Q
−
x
≤ Q
x, j
≤ Q
+
x
(2)
where X is the set of external sources (indexed by x), P
x, j
and Q
x, j
are the active and reactive pow er outputs of x,respectivelyand
IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 5, pp. 1 – 11 3
doi: 10.1049/iet-gtd.2009.0548
&
The Institution of Engineering and Technology 2010
www.ietdl.org
IET Corrected Proof
P
−
x
, P
+
x
and Q
−
x
, Q
+
x
are the min/max activ e and reactive power
outputs of x,respectively.
† Capacity constraints for the WTs: maximum capacity that
may be installed at each site ∀j [ J, ∀g [ G
P
−
g
≤ P
g, j
≤ P
+
g
Q
−
g
≤ Q
g, j
≤ Q
+
g
(3)
where G is the set of WTs (indexed by g), P
g, j
and Q
g, j
are the
active and reactive power outputs of g, respectively, and P
−
g
, P
+
g
and Q
−
g
, Q
+
g
are the min/max active and reactive power output
of g, respectively.
† Voltage-level constraints ∀j [ J, ∀b [ B
V
−
b
≤ V
b, j
≤ V
+
b
(4)
where V
b, j
is the voltage at b, V
+
b
and V
−
b
are the max/min
voltage at b, respectively.
† Flow constraints for lines and transformers ∀j [ J, ∀l [ L
( f
P
l, j
)
2
+ ( f
Q
l, j
)
2
≤ f
+
l
(5)
where f
P
l, j
and f
Q
l, j
represent the active and reactive power
injection onto l, respectively, and f
+
l
the maximum power
flow on l.
† Short-circuit-level constraint: the requirement of not
exceeding the design short-circuit capacity in typical radial
networks, fed by a MV/LV substation and with wind
generation, should be satisfied as it could constrain new
generation capacity. WTs connected to the distribution
network may contribute to the short-circuit level at the
distribution substation. The upstream grid provides the
dominant contribution to the short circuit capacity, which
rapidly diminishes downstream the network due to the
series impedance of the lines. The short-circuit requirement
normally needs to be checked at the MV (or LV) busbars
of the substation [33]. Therefore, given the typical radial
arrangement of distribution networks, the maximum short-
circuit level will be obtained when considering a three-
phase short-circuit at the low-voltage side of the substation.
The magnitude of the expected short-circuit current |I
cc
| at
the low-voltage side of the substation, calculated from the
phasor sum of the maximum short-circuit currents from
the upstream grid, through the step-down transformer, and
from the WTs connected to the distribution network, is,
therefore, limited by the design short-circuit capacity I
max
cc
.
|I
cc
|≤I
max
cc
(6)
The grid contribution is calculated according to IEC 60909
[34–37] and the contribution of WTs is computed
according to the method proposed in [33].
The additional constraints derived from the active
management schemes are coordinated OLTC voltage
constraint, curtailed energy and WTs power factor angles.
† Curtailed energy constraint ∀j [ J
CE
j
g
≤ CE
j
g max
(7)
where CE
j
g
represents the amount of curtailed energy from
generator g during period j and CE
j
g max
= C
j
f
× E
j−max
g
the maximum permitted curtailed energy from generator g
during j, where C
j
f
is the curtailment index, varying in
the range [0, 1] and E
j−max
g
is the maximum energy that
generator g could have produced during j without
curtailment.
† Coordinated OLTC voltage constraint ∀j [ J
V
−
OLTC
, V
j,OLTC
, V
+
OLTC
(8)
where V
j,OLTC
is the secondary voltage of the OLTC during
j, V
−
OLTC
and V
+
OLTC
are the (max/min) voltage of the OLTC,
respectively.
† Coordinated generator reactive power constraints, ∀j [ J,
∀g [ G
w
−
g
,
w
g, j
,
w
+
g
(9)
where
w
g, j
is the power factor angle of g during j,
w
−
g
and
w
+
g
are the (max/min) power factor angle of g, respectively.
The proposed method has been implemented in Matlab
w
and is based on MATPOWER suite [38] and demonstrated
through the study system described in the following section.
4 Study system
The following analyses are based on a 69-bus 11 kV radial
distribution system whose data are given in [39]. The four
feeders are supplied by two identical 6 MVA 33/11 kV
transformers. Fig. 1 shows the distribution system and the
potential WT locations, selected to demonstrate the
capabilities of the method.
4.1 Modelling of time-varying load and
wind power generation
For the modelling of time-varying load and wind power
generation, real data from the local distribution network in
Nordjylland in Denmark have been used and processed. In
order to account for the seasonal, weekly and daily variation
of load, the measured data are grouped by summer/winter,
weekday/weekend and 24 h. In order to account for the
seasonal and daily variation of wind power generation, the
measured data are grouped by summer/winter and 24 h. In
particular, the 365 days of the year have been divided into
153 winter days and 212 summer days and, for each week
4 IET Gener. Transm. Distrib., 2010, Vol. 4, Iss. 5, pp. 1 – 11
&
The Institution of Engineering and Technology 2010 doi: 10.1049/iet-gtd.2009.0548
www.ietdl.org
