Improving Voltage Levels in Low-Voltage Networks With Distributed Generation - Case Study
Improving Voltage Levels in Low-Voltage Networks With Distributed Generation - Case Study
Improving Voltage Levels in Low-Voltage Networks With Distributed Generation - Case Study
Abstract—The use of distributed generation in low-voltage power of these systems usually does not exceed a few
networks may cause the voltage variation in them, within the kilowatts. In this paper only small wind power systems are
wide range. In unfavourable circumstances, the voltage may considered. Wind-based power systems are characterized by
reach unacceptable values. The paper presents the effect of high unpredictability of power production, which directly
distributed generation on voltage levels in a selected low- depends on weather/wind conditions. It means that maximal
voltage rural distribution network in Poland. An analysis of power generation as well as no power generation can occur
possible methods for improving voltage levels in this network is practically at any time of the day, and may not be coherent
conducted. The most suitable method is indicated. with maximal and minimal power demand of consumers.
Example waveforms of variability of wind power generation
Keywords—distributed generation, low-voltage distribution
network, microgeneration, voltage control
can be found in literature [7], [11]–[13].
MV
I. INTRODUCTION TR1
Nowadays many consumers of electric power are also LV W00
owners of power sources and, in fact, become prosumers. In L00-01
consequence, there are many microsources of electric power W01 W40
in low-voltage networks, based on gas and wind L01-02 O40
12.5 kVA
L37-40
microturbines, photovoltaics, fuel cells as well as W02 W37 W35
microturbines for biofuels [1]. The power surplus generated L02-03 O37a
12.5 kVA O37b
L30-37 L30-35
O35
16.5 kVA
The application of microgeneration also affects other power O16 L20-22 O20 L20-27 O34a L34-41
quality factors – it increases the amount of active power loss 12.5 kVA 12.5 kVA 16.5 kVA O34b
16.5 kVA
W22 W27 W41
[9] and distorts the voltage waveform, due to the presence of O22 L27-32 O27 L41-46 L41-45
in rural areas. These networks are relatively long, so it means W36 W38 10.5 kVA
W48 12.5 kVA
W51
that their impedance value is significant. O36a
10.5 kVA
O36b
12.5 kVA
L38-39 O48a
12.5 kVA
O48b
16.5 kVA
L51-52 O51
16.5 kVA
W39 W52
This paper presents the results of the analysis of the L39-42
L39-44 L39-43 O52a O52b
impact of distributed generation on the voltage levels in W42 W44 W43
16.5 kVA 10.5 kVA
W47 W49
proposals to improve the voltage levels are presented. L47-50
L47-54 L49-53
NETWORK
Fig. 1. Structure of the analyzed rural network. TR1 – power transformer,
The most popular microsources in Poland are small W00÷W54 – numbers of nodes, O54 (16.5 kVA) – number of a consumer
photovoltaic installations, and small wind turbines. The rated and its subscribed demand, L00-01÷L51-52 – numbers of power lines.
• 14 consumers with subscribed demand of 10.5 kVA, Vmin – minimal value of the voltage in the node.
• 23 consumers with subscribed demand of 12.5 kVA, Moreover, a voltage difference ΔVW between voltage in
a node with the highest value VW-max and voltage in a node
• 15 consumers with subscribed demand of 16.5 kVA. with the lowest value VW-min has been calculated:
The farthest consumer is located 713 meters from the
power transformer (node W54 in Fig. 1). The network load ΔVW = VW - max − VW - min (2)
has been determined on the base of known subscribed
demand of consumers and 2017’s load profiles included in
Annex No. 5 of the Power System Operator instruction [14]. where
The maximal and minimal load values for particular hours of VW-max – max{Vmax(W00); …; Vmax(W54)},
the day, resulting from the assumed load profiles, are
presented in Fig. 2. Based on them, it was assumed in the VW-min – min{Vmin(W00); …; Vmin(W54)}.
analysis that the maximal contribution of each consumer to
It should be stressed that, according to [14], permissible
the network load is equal to 20% of power indicated by the
voltage variation is within the range ±10% (with reference to
subscribed demand, and the minimal contribution is equal to
the nominal voltage of the network; 0.2 p.u. in total).
4.7%, respectively. Rated power of every wind-based
mircosource is equal to 5 kW – the voltage analysis has been
performed for this value of generated power. B. Voltage Levels without Voltage Control
Firstly, voltage levels in the low-voltage network have
As a result of the previous consideration [15], the been calculated for the following cases:
microsources are connected to the following nodes (Fig. 1):
W06, W14, W17, W23, W27, W30, W35, W40, W45, W50, • no microgeneration (0%) and minimal load of the
W54. The computer model of the rural network, and all network,
simulations have been performed with the use of DigSilent
PowerFactory software. • no microgeneration (0%) and maximal load of the
network,
III. RESULTS OF THE COMPUTER ANALYSIS • maximal microgeneration (100%) and minimal load of
the network,
A. The Scope of the Analysis
• maximal microgeneration (100%) and maximal load of
First step of the analysis is to evaluate voltage reference the network.
levels, i.e. when no microgeneration is applied. Next, the
following network states and measures for voltage levels Results of the calculation are presented in Fig. 3. There
improvement are considered: are significant differences in the voltage levels between
nodes. The lowest voltage variations are observed in the
• microsources are installed and no voltage control is supply node (node W00 – with transformer). It is obvious
applied – voltage levels depend on actual power demand that, when no microgeneration occurs, the highest value of
of consumers and operation of microsources without voltage is in this node, and the lowest is in the node supplied
their power factor control (assum. power fact. value: 1), via a line with the highest impedance (here: at the end of the
network – node W54).
After starting microgeneration with full power, the C. Voltage Adjustment by Microsources Reactive Power
opposite tendency in voltage levels along the length of the Control
power network (blue dashed trace in Fig. 3) is observed. In In order to change voltage in the network, a variation of
nodes far from the power transformer, the voltage value can the power factor of microsources, within the range 0.95ind.÷
be much higher (the highest in node W54) than in node W00, 0.95cap., has been considered. It allows the voltage to be
and is very close to the upper permissible limit (1.1 p.u.). maintained at the connection point of the microsource during
Comparison of voltage values for the characteristic nodes is its operation in the given reactive power range, whereas
presented in Tab. I and Tab. II. when the limit of reactive power generation is reached, the
Analysis of the calculation results presented in Tab. I microsource operates with constant reactive power. It has
enables to conclude that voltage values are within the been assumed that the desired voltage value of the sources is
permissible range 0.9÷1.1 p.u. for every node. However, equal to the nominal voltage of the network.
after application of microgeneration, maximal voltage at Fig. 4 presents voltage levels for all nodes of the
node W54 is equal to 1.096 p.u. (1.046 p.u. without network, whereas Tab. III and Tab. IV present calculation
microgeneration) and is very close to the upper limit. results for selected, representative nodes. It is obvious that
Moreover, voltage difference ΔVW is equal to 0.165 p.u. for the case with no microgeneration (0% in Tab. III),
(Tab. II) and is higher than for the state with no voltage values are the same as in Tab I for the respective
microgeneration (0.140 p.u). In some circumstances case. When voltage adjustment by control of the reactive
microgeneration may be the cause of excessive voltage rise power is applied, for microgeneration 100% (Tab. III)
and variation, therefore voltage control is recommended to voltage levels are also almost the same as without the
be applied. adjustment (Tab. I). Comparison of voltage variation
between selected nodes (Tab. IV) gives similar conclusion.
1.12
1.1 1.12
1.08 1.1
1.06 1.08
1.04 1.06
Voltage [p.u.]
1.02 1.04
Voltage [p.u.]
1 1.02
0.98 1
0.96 0.98
0.94 0.96
0.94
0.92
W00
W02
W04
W06
W08
W10
W12
W14
W16
W18
W20
W22
W24
W26
W28
W30
W32
W34
W36
W38
W40
W42
W44
W46
W48
W50
W52
W54
0.92
W00
W02
W04
W06
W08
W10
W12
W14
W16
W18
W20
W22
W24
W26
W28
W30
W32
W34
W36
W38
W40
W42
W44
W46
W48
W50
W52
W54
Node numbers
Node numbers
Fig. 3. Voltage levels without voltage control: ● no microgeneration (0%)
and max load of the network, ♦ no microgeneration (0%) and min load of Fig. 4. Voltage levels with voltage adjustment by microsources reactive
the network, ▲ max microgeneration (100%) and max load of the network, power control: ● no microgeneration (0%) and max load of the network,
■ max microgeneration (100%) and min load of the network. ♦ o microgeneration (0%) and min load of the network, ▲ max
microgeneration (100%) and max load of the network, ■ max
microgeneration (100%) and min load of the network.
Fig. 5. Simplified diagram of the network indicating the Static E. Voltage Control by the use of Transformer Tap-Changer
Compensators (SC) location.
Voltage control in low-voltage networks with the use of
on-load tap-changer of transformer is considered. Usually, in
this type of voltage control, a reference node (a node to
TABLE V. TECHNICAL DATA OF THE INSTALLED COMPENSATORS achieve desired value of voltage) is the node with the
No. of Rated Reactive power Droop Desired transformer (here W00). However, in this investigation other
node power range voltage nodes have also been adopted as reference nodes: W52, W54
kVA kvar % p.u. (Fig. 7). For node W00, the desired value of voltage is
W52 30.7 -30.7 ÷ +30.7 1 0.980
assumed to be 1.065 p.u., and for nodes W52, W54: 0.98 p.u.
W54 11.9 -11.9 ÷ +11.9 1 0.980 Fig. 8 presents voltage levels in the network when node
W00 is a reference node. The obtained results are almost
identical to those for the case of network operation without
voltage control (Fig. 3). Voltage variation in this node is very
1.12 small, resulting from the proximity of the medium-voltage
1.1 network with low impedance. Unfortunately, voltage in the
1.08
nodes far from the transformer varies within a relatively wide
1.06
range. Much better results have been obtained when a node
1.04
far from the transformer (W52 – Fig. 9 and W54 – Fig. 10) is
Voltage [p.u.]
1.02
adopted as a reference node. Admittedly, in node W00 the
voltage variation is wide, but for nodes starting from W06, it
1
is relatively narrow.
0.98
0.96
R
max 1.056 0.931 0.940
1.02
0
1 min 1.071R 1.044 1.046
R
0.98 max 1.060 0.984 1.000
0.96 100
min 1.073R 1.087 1.096
0.94
R
0.92
max 1.085 0.964 0.972
0
W00
W02
W04
W06
W08
W10
W12
W14
W16
W18
W20
W22
W24
W26
W28
W30
W32
W34
W36
W38
W40
W42
W44
W46
W48
W50
W52
W54
R
min 1.018 0.990 0.992
Node numbers
max 1.060 0.984R 1.000
100
Fig. 8. Voltage levels with voltage control by the use of tap-changer of min 0.973 0.988 R
0.998
transformer. Node to achieve desired (reference) voltage – W00: ● no
microgeneration (0%) and max load of the network, ♦ no microgeneration max 1.085 0.964 0.972R
(0%) and min load of the network, ▲ max microgeneration (100%) and 0
min 1.018 0.990 0.992R
max load of the network, ■ max microgeneration (100%) and min load of
the network. max 1.033 0.955 0.971R
100
min 0.973 0.988 0.998R
R
– reference node (node to achieve desired voltage)
1.12
1.1
1.02
node p.u. p.u. p.u. p.u.
1 W00 0.017 0.156 0.156 0.165
0.98 W52 0.112 0.026 0.028 0.121
0.96 W54 0.112 0.035 0.027 0.130
0.94
0.92
Therefore, the method with use of tap-changer of
W00
W02
W04
W06
W08
W10
W12
W14
W16
W18
W20
W22
W24
W26
W28
W30
W32
W34
W36
W38
W40
W42
W44
W46
W48
W50
W52
W54
1.02
Voltage levels can be very high even in nodes far from the
1
transformer supplying the network. In unfavourable
0.98
circumstances, microgeneration may cause voltage to exceed
0.96
the permissible level. Thus, application of voltage control
0.94 should be considered. In the analyzed low-voltage network,
0.92 the most suitable is the method that uses tap-changer of the
W00
W02
W04
W06
W08
W10
W12
W14
W16
W18
W20
W22
W24
W26
W28
W30
W32
W34
W36
W38
W40
W42
W44
W46
W48
W50
W52
W54