Sunteți pe pagina 1din 13

`

Mitigation of Voltage Fluctuations by Controlling Reactive Power of DER

Nokhum Markushevich
Smart Grid Operations Consulting

Introduction
The injections of kW by solar and wind generators vary with the changes in cloud cover and
wind. The variation can range from 10% to 100% of the rated kWs (see e.g., [1] and [2]). When
such a Distributed Energy Resource (DER) is located downstream from significant impedance,
the variability of the kWs may cause unacceptable voltage fluctuations. In many cases, the
voltage fluctuations can be compensated either by a corresponding increase of the generation of
kvars, or by a reduction in the absorption of kvars. To mitigate the voltage fluctuations by the
change of reactive power, there must be a sufficient reserve of the reactive power in the subject
circuits. When it comes to DERs with reactive power capability, it should be noted that the
available DER’s kvars might exceed the rated kvars under the rated kWs when the kWs are
below the full capacity. Whether this increase in available kvars is enough of a reserve for
compensating the voltage fluctuation needs to be checked.

Methods of Mitigation of voltage fluctuations due to Intermittent Operations


of DER considered in the paper.
Two methods of mitigation of voltage fluctuations due to DER intermittency were discussed in
[3]:

a. Constant absorbing power factor of DER [4]. As stated in [3], if the DER is set to follow
a given power factor (PF) to absorb kvars, the reserve for mitigation of the voltage
fluctuations is limited by the ability to reduce the absorbing kvars to zero, but it cannot
be extended into generating kvars. Operating DERs in kvar-absorbing modes increases
the need for compensating the reactive loading of the distribution and transmission
circuits. It also may increase the power losses in distribution and transmission, the
loading of the circuit elements, and the voltage drops in portions of the distribution
circuits. This method of mitigating the voltage fluctuations is an open-loop method and
can provide adequate compensation of the voltage fluctuations, as long there is a
sufficient range for reducing the absorbing kvars, i.e., as long as there is no need to
generate kvars. Different values of the absorbing power factor may be needed to adjust
the kvar control to different circuit designs.

b. Constant voltage at one or more given points of the circuit [3] and [5]. As stated in [3], in
this mode of operations, the reactive power of DER can change from generating kvars to
absorbing, and vice versa, within the DER capability curve, following the deviations of
voltage from a given setpoint. This means that the generating or absorbing kvars can

1
`

exceed the rated kvars, when the kWs are below the rated value. This is a close-loop
control of vars with the voltage as a target. A droop-based control was considered in [3].
It was assumed in [3] that under the “constant voltage” mode there is no constant
predefined voltage setpoint for the DER to follow. The voltage at the DER terminals (or
at another target node) is established for any time based on the current operating
conditions of the corresponding distribution circuits and on the operations of the
particular DER. The initial volt/var requirements for the DER under steady- state
conditions can be defined by the controlling entity based on the current operational
objectives. The objective of the “constant voltage” mode of DER operations assumed in
[3] is to keep the voltage close to the steady-state level by compensating the voltage
deviations from this level by adjusting the reactive power of the DER. For instance, the
“constant voltage” setpoints can be derived as a running average of previous voltage
measurements under steady-state conditions, where the average is either of a given
number of measurements, or over some predefined time interval. The setpoints of the
voltage control, including the droop, are the adjustable parameters for this method.

In this paper, two other methods are discussed, as follows:


1. Maximum limit of DER’s kvars (Qlim method). In this mode of operations, the DER
controls its kvars according to the following formula:

Qi,t = Ai,t x Qmaxi,t (kWi,t, Volti,t) (1)

where,

Qi,t – is the injection of the reactive power by the 𝑖-th DER under condition t
Ai,t – is the selected portion of the maximum available reactive power of the 𝑖-th DER
(the adjustment coefficient), for the t-th condition (steady-state or variability)
Qmaxi,t (kWi,t, Volti,t)– is the maximum available reactive power of the 𝑖-th DER under t-
th condition as a function of the actual kW of the 𝑖-th DER and the voltage at the 𝑖-th
DER under t-th condition.

This is an open-loop method of reactive power control. The coefficients can be used to
adjust the control to match the specifics of the circuit and to the condition of steady-state
or variability.

As follows from the DER’s capability curve, when the kWs of the DER drop (e.g., due to
the variability of the source), the maximum available kvars of the DER increase. Hence,
the injection of kvars increases, which results in an increase in voltage that can
compensate for the voltage reduction due to the drop in kWs.

2
`

The coefficients Ai,t can be different for the condition of steady-state or variability. In
order to automatically change the coefficients, an automatic detection of the transition
from steady-state to variability and vice versa should be available.

2. Addition of kvars in proportion to the reduction of kWs (kW Delta method). In this mode
of operations, the DER controls its kvars according to the following formulas:

Qj,t = Qconstj,t + Bj,t x (Pinitj – Pactj,t) (2)

Qj,t ≤ Qmaxj,t (3)


Where
Qj,t - is the injection of the reactive power by the j-th DER under the condition t
Qconstj,t – is the constant component of the reactive power injection by the j-th DER
selected to provide the desired compensation of the reactive load in the subject circuits
for the t-th condition
Bj,t - is the selected portion of the reduction of kWs from an initial value of kW Pinitj to
the actual value Pactj,t for the j-th DER and for the t-th condition..

This is also an open-loop method of reactive power control. The parameters Qconstj,t and
Bj,t can be used to adjust the control to the specific circuit. These parameters also can be
different for the steady-state and for the variability conditions.

Illustrative examples
The illustration of these methods is based on a simple circuit diagram presented on Figure 1.

Node 1 Node 2

G G

Figure 1. Example diagram with DER in the secondary circuits

As seen in the figure, there are two DERs connected to the secondary voltage busses. The
following three ratios of the reactance to the resistance of the primary circuits have been
considered: 1 to 1; 2 to 1; and 3 to 1. Two rated DER power factors were considered: 0.95 and
0.9. The voltage drops in the primary circuits, when the DERs are in steady-state operations, are
in the range from 4% through 7%, depending on the X/R ratio and DER power factor.

3
`

The reactance to resistance ratio of the distribution transformers was assumed to be 6 to 1, and
the voltage drop in the secondaries is about 2% of the nominal voltage. The CVR-watts factor is
1, and the CVR-vars factor is 4. The power factor of the load is 0.9.

In these examples, it was assumed that the kWs of the DERs randomly fluctuate from the full
capacity to 10%. It was also assumed that the ratio of the DER kW capacity to the load is 0.4.

In all scenarios, the feeding bus voltage was kept to provide the same minimum voltage at the
customer terminals (0.95 pu).

The following criteria were used for the comparison of the methodologies:

 Standard deviation of voltage (secondary and primary) at node 1 (StDev1)


 Standard deviation of voltage (secondary and primary) at node 2 (StDev2)
 Losses in the entire circuit (Losses)
 Maximum current in distribution primary circuit (Loading)
 Average kW loading of the distribution feeder (Avrg_Load+Loss)
 Average kvar required for the distribution circuit (Avrg_kvar_req)
 Maximum kvar required for the distribution circuits (Max_kvar_req)

The secondary voltages at the customer terminals were the target voltages. The fluctuations of
these voltages were considered acceptable, if their standard deviations did not exceed 0.25% of
the nominal voltage.

The parameters of the sample circuit are different from the ones used in [3]. Because of that, the
two methods of reactive power control described in [3] were tried again in the example of this
paper.

The results of the analyses are presented in

Table 1, Table 2 and in the figures below.

4
`

Table 1. Results of mitigation of voltage variability due to DER intermittency


Mitigating Secondary Voltages
X/R Results PF=1 PF=0.9 pf=0.95
no comp. ConsPF-abs Q-lim*A DeltaP*B+C ConstV ConsPF-abs Q-lim*A DeltaP*B+C ConstV
StDev1 0.41% 0.04% 0.24% 0.07% 0.15% 0.11% 0.22% 0.10% 0.05%
StDev2 0.56% 0.04% 0.23% 0.05% 0.01% 0.21% 0.23% 0.20% 0.01%
Losses 8.33% 8.90% 7.61% 7.66% 7.67% 8.70% 7.75% 7.83% 7.77%
MinVolt 0.95 0.95 0.95 0.95 0.95 0.95 0.95 0.95 0.95
1
Max loading 1.268 1.279 1.106 1.116 1.133 1.28 1.128 1.152 1.139
Avrg_Load+Loss 1.093 1.100 1.084 1.084 1.083 1.10 1.086 1.087 1.086
Avrg_kvar_req 0.620 0.746 0.322 0.351 0.359 0.71 0.400 0.440 0.409
Max_kvar_req 0.635 0.815 0.467 0.451 0.412 0.75 0.556 0.525 0.476
StDev1 0.41% 0.18% 0.18% 0.21% 0.05% 0.01% 0.21% 0.05% 0.05%
StDev2 0.56% 0.18% 0.21% 0.21% 0.06% 0.06% 0.24% 0.08% 0.02%
Losses 8.31% 8.86% 7.77% 7.64% 7.7% 8.69% 7.86% 7.84% 7.79%
MinVolt 0.95 0.95 0.95 0.95 0.95 0.95 0.95 0.95 0.950
2
Max loading 1.283 1.286 1.155 1.113 1.142 1.291 1.165 1.160 1.158
Avrg_Load+Loss 1.097 1.105 1.089 1.086 1.085 1.102 1.090 1.090 1.089
Avrg_kvar_req 0.653 0.784 0.446 0.376 0.392 0.742 0.485 0.469 0.449
Max_kvar_req 0.681 0.848 0.534 0.469 0.430 0.779 0.590 0.562 0.496
StDev1 0.41% 0.18% 0.19% 0.20% 0.02% 0.09% 0.17% 0.10% 0.00%
StDev2 0.56% 0.21% 0.24% 0.24% 0.09% 0.09% 0.22% 0.14% 0.04%
Losses 8.3% 8.71% 7.83% 7.66% 7.68% 8.66% 7.96% 7.82% 7.77%
MinVolt 95.0% 0.95 0.95 0.950 0.95 0.95 0.95 0.950 0.95
3
Max loading 1.299 1.297 1.179 1.130 1.154 1.30 1.21 1.164 1.165
Avrg_Load+Loss 1.101 1.107 1.092 1.088 1.088 1.11 1.09 1.092 1.091
Avrg_kvar_req 0.684 0.793 0.506 0.416 0.426 0.78 0.56 0.496 0.467
Max_kvar_req 0.726 0.836 0.569 0.486 0.448 0.81 0.62 0.580 0.516

Table 2. Impact on primary voltages, while mitigating fluctuations of secondary voltages

Impact on Primary Voltages


X/R Results PF=1 PF=0.9 pf=0.95
no comp. ConsPF-abs DeltaP*B+C ConstV ConsPF-abs DeltaP*B+C ConstV
StDev1 0.33% 0.17% 0.15% 0.20% 0.22% 0.20% 0.18%
1
StDev2 0.49% 0.25% 0.23% 0.27% 0.33% 0.31% 0.25%
StDev1 0.33% 0.01% 0.06% 0.10% 0.11% 0.11% 0.11%
2
StDev2 0.49% 0.02% 0.08% 0.12% 0.17% 0.16% 0.15%
StDev1 0.33% 0.05% 0.12% 0.03% 0.00% 0.14% 0.07%
3
StDev2 0.49% 0.08% 0.18% 0.02% 0.01% 0.22% 0.08%

Figure 2 through Figure 4 present the standard deviations of voltages at the target buses and at
the primary voltage buses. As seen in the figures, all methods can provide the required voltage
quality at the target buses in this example circuits. As seen in Figure 2, in some cases, mitigating
the voltage fluctuations in the secondary circuits does not prevent undesirable voltage
fluctuations at some primary buses.

5
`

0.60%

0.55%

0.50%

0.45%

% of Nominal Voltage
0.40%

0.35%

0.30%

0.25%

0.20%

0.15%
StDev1
0.10%
StDev2
0.05%

0.00%
ConsPF-abs

ConsPF-abs

ConsPF-abs

Q-lim*A

ConsPF-abs

Q-lim*A
no comp.
constPF-delta

ConstV

constPF-delta

ConstV

DeltaP*B+C

ConstV

DeltaP*B+C

ConstV
(no compensation)

PF=1 PF=0.9 pf=0.95 PF=1 PF=0.9 pf=0.95


Impact on Primary Voltages Mitigating Secondary Voltages

Figure 2. Standard Deviations of Voltage Fluctuations for X/R=1

0.60%
0.55%
0.50%
0.45%
% of Nominal Voltage

0.40%
0.35%
0.30%
0.25%
0.20%
0.15%
StDev1
0.10%
StDev2
0.05%
0.00%
ConsPF-abs

ConsPF-abs

ConsPF-abs
Q-lim*A

ConsPF-abs
Q-lim*A
no comp.
constPF-delta
ConstV

constPF-delta
ConstV

DeltaP*B+C
ConstV

DeltaP*B+C
ConstV
(no compensation)

PF=1 PF=0.9 pf=0.95 PF=1 PF=0.9 pf=0.95


Impact on Primary Voltages Mitigating Secondary Voltages

Figure 3. Standard Deviations of Voltage Fluctuations for X/R=2.

6
`

0.60%
0.55%
0.50%

% of Nominal Voltage
0.45%
0.40%
0.35%
0.30%
0.25%
0.20%
0.15%
0.10% StDev1
0.05%
StDev2
0.00%
ConsPF-abs

ConstV

ConstV

Q-lim*A

ConstV

Q-lim*A

ConstV
ConsPF-abs

ConsPF-abs

ConsPF-abs
constPF-delta

constPF-delta

no comp.

DeltaP*B+C
(no compensation)

DeltaP*B+C
PF=1 PF=0.9 pf=0.95 PF=1 PF=0.9 pf=0.95
Impact on Primary Voltages Mitigating Secondary Voltages

Figure 4. Standard Deviations of Voltage Fluctuations for X/R=3.

Figure 5 presents the real power losses in the distribution circuit for different methods of
mitigating voltage fluctuations. As seen in the figure, the losses are approximately at the same
level for the constant voltage methodology and for the two additional methodologies presented in
this paper (Qlim and kW Delta methods).

9.50%

9.00%

8.50%
Losses, %

8.00%

Losses_x/r=1
7.50%
Losses_x/r=2
Losses_x/r=3
7.00%

6.50%
ConsPF-abs
no comp.

Q-lim*A

ConsPF-abs

Q-lim*A
DeltaP*B+C

DeltaP*B+C
ConstV

ConstV

PF=1 PF=0.9 pf=0.95


Mitigating Secondary Voltages

Figure 5. Energy Losses

7
`

Figure 6 presents the loading of the distribution circuits (Amps in pu) for the mitigation methods
considered here. As seen in the figure, in this particular solution, the loading in the case of the
Qlim method slightly exceeds the loading in the Delta and Constant Voltage methods. The
difference, probably, can be reduced by adjusting the A coefficient in the Qlim case.

1.350

1.300

1.250
Loading, pu

1.200

1.150
Loading_x/r=1
1.100 Loading_x/r=2
Loading_x/r=3
1.050

1.000
ConsPF-abs
no comp.

Q-lim*A

ConsPF-abs

Q-lim*A
DeltaP*B+C

DeltaP*B+C
ConstV

ConstV
PF=1 PF=0.9 pf=0.95
Mitigating Secondary Voltages

Figure 6. Maximum circuit Loading

Figure 7 presents the load plus loss values for the circuit under consideration. These values
ultimately reflect the customers’ electricity bills. As seen in the figure, these values are close for
the Constant Voltage, Qlim, and kW Delta cases.

8
`

1.120

1.110

1.100

Load+Loss, pu 1.090

1.080

1.070 Load+Loss_x/r=1
Load+Loss_x/r=2
1.060
Load+loss_x/r=3
1.050

1.040
ConsPF-abs

ConsPF-abs
no comp.

Q-lim*A

Q-lim*A
DeltaP*B+C

DeltaP*B+C
ConstV

ConstV
PF=1 PF=0.9 pf=0.95
Mitigating Secondary Voltages

Figure 7. Real Load+Losses in Distribution

Figure 8 presents the maximum required reactive power for the circuits. As seen in the figure,
the required kvars are slightly greater in the Qlim and kW Delta cases in comparison with the
Constant Voltage case. The differences, probably, can be reduced by adjusting the A and B
coefficients.

0.900

0.800
Maximum kvar required, pu

0.700

0.600

Max_kvar_x/r=1
0.500
Max_kvar_x/r=2
Max_kvar_x/r=3
0.400

0.300
ConsPF-abs

ConsPF-abs
Q-lim*A

Q-lim*A
no comp.

DeltaP*B+C

DeltaP*B+C
ConstV

ConstV

PF=1 PF=0.9 pf=0.95


Mitigating Secondary Voltages

Figure 8. Maximum kvar required for the distribution circuit

9
`

As it was mentioned above, the adjustment coefficients can be different for the steady-state and
for the variability conditions. Consider an illustrative example for the Qlim method.

In this example, the setup of the volt/var control by the two DERs and by the controller of the
load tap changer (LTC) has the following two objectives:

 Keep the voltage at the customer terminals close to the minimum standard limit for
energy conservation, without violating the limit
 Keep the fluctuation (flicker) of the voltage at the customer terminals below ±0.5%

The following parameters were used in this example for the adjustment of the control:

 The bandcenter setpoint of the LTC controller.


 The A coefficients for DERs at node 1 and at node 2.

For the steady-state condition, the LTC bandcenter was set to 1.055pu, the coefficient A for the
DER at node 1 was set to 0, and it was set to 1 for the DER at node 2. The random components
of the bus voltage were assumed to be in the range of ±0.00125.

Under these settings, the minimum voltages at the customer terminals were 0.96pu and 0.95pu at
the node 1 and node 2 respectively.

If the same settings were left for the variability conditions, the voltage fluctuation at node 2
would be ±0.9%.

In order to reduce the voltage fluctuations to an acceptable level, the A coefficients for the
variability conditions were set to 0.3 and 0.6 for the DERs at node 1 and node 2 respectively. In
this case, if the LTC bandcenter were left unchanged, the minimum voltage would drop below
the standard limit to 0.94pu. To avoid violations of the voltage limit, the bandcenter was changed
to 1.065pu. For this case, the random components of the bus voltage were assumed to be in the
range of ±0.00185.

In this example, the detection of the transition from the steady-state to the variability conditions
and back was accomplished by monitoring the standard deviation of the DER’s kWs for a small
number of consecutive measurements. The criterion for the detection of the change was set to
10% of the DER’s rated kW capacity. If the standard deviation of the DER’s kW was below
10%, the condition was considered a steady-state condition. If it was 10% or above, it was
considered to be a variability condition.

It can be noted that if the change from steady-state to variability and vice versa conditions is
detected, other adjustments in the subject circuits can be made. For instance, to avoid excessive
operations of the LTC, the bandwidth and/or the time delay of the LTC controller could be made
less sensitive to the fluctuations of the real and reactive powers and voltages.

10
`

The results of this example are illustrated in Figure 9 through Figure 11.

0.25

0.2

0.15
DER kW, pu

0.1

0.05

0
1 21 41 61 81 101 121 141 161 181 201 221
Instances

Figure 9. The DER kW for the steady-state and the variability conditions

1.2

0.8
A1, A2, StDev

StDev
0.6
A1
A2
StDev-crit
0.4

0.2

0
1 21 41 61 81 101 121 141 161 181 201 221
Instances

Figure 10. Change of parameters with the change of DER operating conditions

11
`

1.08

1.07

1.06

1.05

1.04

1.03

1.02
Volt, pu

1.01
Volt 2
1
Volt 1
0.99 BusVolt
0.98

0.97

0.96

0.95

0.94

0.93
1 21 41 61 81 101 121 141 161 181 201 221
Instances

Figure 11. Voltages at the customer terminals and at the substation bus. Mitigated voltage fluctuations.

Conclusions
1. The following two additional methods of mitigating voltage fluctuations caused by DER
variability are discussed in the paper:
a. Qlim method, when the DER’s kvar injection is proportional to the maximum
reactive power limit as a function of the actual DER kW and the voltage at the
DER terminals
b. kW Delta method, when the DER’s kvar injection is proportional to the reduction
of DER’s kWs
2. These two methods of mitigation can provide the required voltage quality at the target
terminals similar to that of the Constant Voltage method, with approximately the same
level of losses, loading, and reactive power requirements.
3. All methods considered here need adjustable parameters to fit the specifics of the designs
of the subject circuits.
4. In some cases, the needed adjustment of the DER reactive power control may be different
for the steady-state operating conditions of the DERs and for the DER variability
conditions. In these cases, automated detection of the transition from the steady-state
condition to the variability and vice versa is needed.
5. To provide the capabilities of mitigating voltage fluctuations by reactive power control of
DERs, sufficient reserves of the reactive power for both the steady-state and variability
conditions should be provided. Hence, the choices of the reactive power control
capability of DERs should be the subject of distribution planning along with other
alternatives of reactive power sources, as well as the subject of interconnection studies.

12
`

Acknowledgment

The author would like to thank Mr. Martin Delson for his comments on this paper.

References.
1. Renewable Integration Benefits, by Energy Storage Association.
Available: http://energystorage.org/energy-storage/energy-storage-benefits/benefit-
categories/renewable-integration-benefits
2. Xuemei Cheng, Wesley Dose, Thomas Hempel, at al, Supercapacitors for peak shifting
over short durations in PV-systems. Available:
http://gronnfase.blogspot.com/2014/10/supercapacitors-for-peak-shifting-over.html
3. Nokhum Markushevich, Mitigating Voltage Fluctuation Caused by Variability of
Distributed Energy Resources. Available:
http://www.energycentral.com/gridtandd/gridoperations/articles/2602/
4. Reigh Walling and Gao Zhi, Eliminating voltage variation due to distribution-connected
renewable generation, DistribuTech 2010
5. Robert Passey, Ted Spooner, Iain MacGill, Muriel Watt, Katerina Syngellakis, The
potential impacts of grid-connected distributed generation and how to address them: A
review of technical and non-technical factors. Available:
https://www.michigan.gov/documents/energy/UCS_Final_response_to_question__37_419821_
7.pdf

13

S-ar putea să vă placă și