Fundamentals of Reactive Power and

Voltage Regulation in Power Systems

Course No: E03-011

Credit: 3 PDH

Boris Shvartsberg, Ph.D., P.E., P.M.P.

Continuing Education and Development, Inc.

P: (877) 322-5800
[email protected]

www.cedengineering.com
 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

Introduction
One of the main goals that every electrical utility company has is transportation of
electrical energy from generating station to the customer, meeting the following main
criteria:

• High reliability of power supply

• Low energy cost
• High quality of energy (required voltage level, frequency etc.)

This course is concentrated on accomplishing the 2nd and 3rd goals through regulation of
reactive power and voltage. Reliability of power supply is a subject of a different course.

To better understand why the regulation of reactive power and voltage makes power
systems more efficient, let’s start with discussion about the structure of the power
systems and their main components.

Power System Structure

The typical power system structure is shown in Figure 1.

Fig. 1 - Power System Structure and Main Components

Where the numerical symbols represent the following components:

(1) Generator
(2) Generating station’s step-up transformer substation
(3) Extra high voltage step-down transformer substation
(4) High voltage step-down transformer substation
(5) Distribution substation
(6) Distribution Transformer
(7) Transmission and Distribution Lines
(8) Customer

© Boris Shvartsberg, 2022 1

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

The elements from 2 to 7 are the components of utility company Transmission and
Distribution (T&D) Systems, with typical power system voltages as follows:

• Generation: Up to 25 kV
• Transmission: 115 – 1500 kV
• Subtransmission: 26 – 69 kV
• Distribution: 4 – 13 kV
• Customer: Up to 600 V

Justification for Voltage Transformation

As we can see from the Figure 1, along the route from the source to the customer,
electricity is undergoing numerous transformations, with generating voltage getting
stepped-up to transmission level with a follow-up decrease down to distribution and
eventually customer levels. Why do we need to do it? As previously was mentioned,
utility company wants to keep energy costs down. One of the ways to do it is to reduce
power and energy losses, which may be accomplished by raising the voltage level,
because power losses ∆P have a functional relationship with voltage level described in
the following equation:

∆P = F (S2xL/V2) (1)
Where:

S is transported apparent power

L is distance to the customer
V is system voltage level

As we can see, there is a reverse proportion between power losses and voltage level in the
2nd degree. For example, if we increase voltage 10 times, power losses will be 100 times
smaller.

Another benefit from raising the voltage is a reduction of voltage drop, which is related to
voltage level ∆V as follows:

∆V = F (SxL/V) (2)
Having smaller voltage drop in the system helps a utility company meet its other
objective which is to provide customer with a high quality electrical energy meeting
specific voltage level requirements.

That’s why we increase voltage for transmission of electrical energy, but after it is
delivered to the area where customers are located, we gradually lower the voltage to the

© Boris Shvartsberg, 2022 2

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

safe utilization level (208/120 V, for example). Number of steps in raising and lowering
the voltage is being defined through optimization studies performed by utility company
Planners.

Fundamentals of Reactive Power Regulation

Besides changing the voltage level, there is another way to reduce power and energy
losses through a reactive power regulation. Let’s see how it may be done.

An apparent power S carried by a power line has two components active power P and a
reactive power Q, which are related as follows:

S = P + jQ = (P2 + Q2)1/2 x ej arctan Q/P , (3)

Where arctan Q/P = ϕ, which is an angle between P and S. Relationship between P, Q

and S is shown in Figure 2 in so called “Power Triangle”. The ratio of P to S is called
the power factor, which from Figure 2 is equal to cosϕ. For the inductive load, current
looking counterclockwise lags the voltage and the power factor is correspondingly called
“lagging”. For the capacitive load, current leads the voltage, and the power factor is
called “leading”. For a 100% active load (Q = 0), P = S → cosϕ = 1, and for a 100%
reactive load (P=O), cosϕ = 0. Usually, a power system has a wide mix of customers
with different shares of active and reactive loads, which makes the combined power
factor 0 ≤ cosϕ ≤ 1.

Fig. 2 - Relationship between S, P and Q

© Boris Shvartsberg, 2022 3

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

Active power losses ∆P and voltage drop ∆V may be found from the following
equations:

∆P = (P2 + Q2) x r / V2 (4)

∆V = 31/2 x (P2 + Q2)1/2 x r / V (5)

Where:

V is system voltage
R is circuit’s resistance

As we can see from Equations (4) and (5) reduction of reactive power transported from
generating station to the customers will lead to reduction of both active power losses and
voltage drops. To achieve this goal, local sources of reactive power may be used: either
shunt capacitors for inductive load, or shunt reactors for capacitive load. Let’s discuss
both options.

Shunt Capacitors
As it was mentioned before, shunt capacitors may be used to provide a local source of
capacitive reactive power Q C to reduce a value of inductive reactive power Q L carried
by the line (usually, an overhead line). The results achieved by the application of shunt
capacitors are shown in Figure 3.

Fig. 3 - Application of Shunt Capacitors for Power Factor Improvement

© Boris Shvartsberg, 2022 4

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

The results achieved are as follows:

• Without a shunt capacitor, apparent power carried by the line SL = PL + jQL,

and power factor cosϕ = PL /SL
• With a capacitor, line apparent power, SL1 = PL + j(QL – QC ) < SL, and
cosϕ1 = PL / SL1 > cosϕ
• Ultimately, power losses ∆P and voltage drop ∆V will be reduced after shunt
capacitor is installed, i.e. ∆P1 < ∆P, and ∆V1 < ∆V

Usually, shunt capacitors are coming as banks made up of a number of capacitor units
that should be connected in parallel and series sections to obtain desired ratings of the
bank, both voltage and capacity. One phase of the typical 26 kV capacitor bank is shown
in Figure 4. It consists of six “cans” that are going to be connected as per manufacturer
instructions.

Fig. 4 - One Phase of 26 kV Capacitor Bank

Let’s learn how to design these connections. Figure 5 shows possible connections for one
phase of a three-phase capacitor bank with capacitor units connected in parallel groups
and groups connected in series. Another option is to connect units in series to form a

© Boris Shvartsberg, 2022 5

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

string, and after this, connect strings in parallel. Combination of these two options is
possible as well. For any connection scheme utilizing capacitor units rated for a voltage
VU and a reactive power QU, the following equations may be used to calculate numbers
of units in each phase required to obtain for the 3-phase bank a total power rating of QB
at a system line voltage VL

VL QB
NSER = ---------------- NPAR = -------------------------, (6)
31/2 x VU 3 x NSER x QU

Where:

NSER = number of series sections

NPAR = number of capacitor units connected in parallel in each series section

Fig. 5 - Connections of Capacitor Units into a Single Phase Bank

It should be noted, that the following relationship exists between a reactive power QU of
every capacitor unit with a capacitance CU and a voltage VU connected to it:

QU = CU x VU2 (7)

If NSER, found from the Equation (6), is not the whole number, the closest higher number
of sections should be selected. A lower is not acceptable because the voltage per section
in this case will be more than the rated voltage for which a capacitor unit is designed. As
a result of such selection, the voltage VU’ per every series section will be lower than a
rated voltage VU. From Equation (7), it follows that the reactive power QU’ will be lower
than QU as well. The following equation may be used to find the derating of a capacitor
unit:

© Boris Shvartsberg, 2022 6

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

QU’ = QU x (VU’/VU)2 (8)

Then, QU’ should be used instead of QU in Equation (6) to find NPAR. The number of
capacitor units that may be connected in parallel and series has some limitations
described in the IEEE Guide for Protection of Shunt Capacitor Banks [1]1. The minimum
number of units in one series section is determined by overvoltage considerations, after
one unit in the section fails. Table 2 of this IEEE Guide [1]1 contains this minimum NPAR
for different values of NSER and connection schemes of the whole bank.

At the same time, there is a maximum number of capacitors that may be connected in
parallel. It is based on considerations of possible discharge transient from a large number
of capacitors following the failure of one of them. This discharge may cause a rupture of
the failed unit with possible damage to the rest of the bank. To prevent it, the maximum
reactive power of one series section should not be higher than 4,650 kvar at a rated
voltage and 60 Hz frequency. Refer to IEEE Std. C37.99-1990 “IEEE Guide for
Protection of Shunt Capacitor Banks [1]1.

Example 1 shows, in a simplified form, a calculation of capacitive power required to

improve a power factor up to a certain level and a selection of the number of capacitor
units that should be connected together to provide this power.

Example 1. Design of 230 kV shunt capacitor bank:

For a 230 kV system with a power flow, shown in Figure 6, design a shunt capacitor bank
that should be installed at substation to increase cosϕ up to at least 0.9 (lagging). The 3-
phase bank should be built using capacitor units rated 13.28 kV, 200 kvar. Phases should
be connected in a grounded wye.

Fig. 6 - Power Flow for Example 1

____________________
1
IEEE C.37.99 – 1990, Copyright © 1990, IEEE. All rights reserved

© Boris Shvartsberg, 2022 7

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

Where:

Sin is Incoming power, MVA Sout is Outgoing power, MVA

Sload is Load, MVA SCB is Shunt capacitor bank

Solution:

1. Find the existing power factor cosϕ1:

tanϕ1 = Qin / Pin = 150/200 = 0.75 → ϕ1 = 370 → cosϕ1 = 0.8

2. Draw power triangle:

Fig. 7 - Power Triangle for Example 1

3. Find new angle ϕ2 corresponding cosϕ2 = 0,9:

ϕ2 = 25.8o

4. Find a new incoming reactive power Q’ in:

Q’in = Pin x tanϕ2 = 200 x tan 25.8o = 200 x 0.484 = 96.86 mvar

It should be noted, that a value of the active power Pin didn’t change with a
change of cosϕ, because it depended on a load demand that stayed the same.

5. Show Q’in, S’in and ϕ2 on a power triangle (see Figure 7). A corresponding
apparent power S’in may be written as:

© Boris Shvartsberg, 2022 8

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

S’in = Pin + j Q’in = (200 + j96.86) mVA

6. Find a size of shunt capacitor bank:

QSCB = Q in - Q’ in = 150 – 96.86 = 53.14 mvar

7. Assuming for a capacitor bank connection option shown in Figure 5 and using
Equation (6) find:

7.1. Number of series sections in each phase:

NSER = 230 / (31/2 x 13.28) = 10.0

7.2. Number of parallel capacitor units in each series section:

NPAR = 53.14 / (3 x 10 x 0.2) = 8.85

The next higher number of 9 should be selected to meet a requirement of

improving cosϕ up to at least 0.9.

From Table 2 of this IEEE Guide [1], the minimum number of parallel capacitors in each
out of 10 series sections, for a grounded wye connection of the whole bank, is 10. Thus,
NPAR = 10 should be used.

The total reactive power QPAR of each series section, consisting of 10 parallel capacitors,
may be found as follows:

QPAR = QU x NPAR = 200 x 10 = 2000 kvar

That is less than a maximum of 4650 kvar, recommended by IEEE Std. C37.99-1990
“IEEE Guide for Protection of Shunt Capacitor Banks [1]. Thus, a selected number of 10
parallel capacitors in each series section is acceptable.

The total capacitive power of the whole bank is:

QC’ = 0.2 x 10 x 10 x 3 = 60 mvar

The new reactive power, flowing into the substation, is:

Qin” = Qin - QC’ = 150 - 60 = 90 mvar

And the new power factor cosϕ3 may be found as follows:

tan ϕ3 = Qin” / Pin = 90/200 = 0.45, ϕ3 = 24.22o, cosϕ3 = 0.912,

© Boris Shvartsberg, 2022 9

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

Thus, a designed three-phase capacitor bank provides the required improvement of cosϕ
of the incoming power. Each phase of this bank consists of 10 series sections with 10
capacitors, connected in parallel in each section, with connections made, as shown in
Figure. 5.

Location of shunt capacitors in the power system is based on special optimization studies.
They may be installed at the transmission and distribution substations as well as on
distribution line poles. After optimal location for shunt capacitors is identified, substation
engineer should select the proper equipment to switch and protect capacitor banks. Very
often, there are several capacitor banks installed at the station and they may be switched
in and out separately to obtain a desired level of compensation of reactive power in the
system. This switching is usually done by properly rated circuit breakers or circuit
switchers, depending on particular application and system requirements. Circuit breaker
is rated to interrupt both load and fault current while circuit switcher is mostly used to
switch load current only. These switching devices may be operated locally and remotely,
giving system personnel an opportunity to disconnect capacitors from the substation bus
when a situation in the system temporarily does not require additional reactive power
compensation, and put them back in service when it is necessary. Examples of circuit
breaker and circuit switcher are shown in Figures 8 and 9 respectively.

Fig. 8 - 230 kV 2000 A Circuit Breaker

© Boris Shvartsberg, 2022 10

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

Fig. 9 - 26 kV 1200A Circuit Switcher for Capacitor Banks Shunt Reactors

So far, we’ve considered the application of shunt capacitors to compensate inductive

reactive power, but in some power systems a reactive component Q of an apparent power
S is mostly capacitive, i.e. a power factor is leading. Usually it occurs when transmission
lines in power system are underground cables rather than overhead lines. Equation for the
apparent power for capacitive systems will look as follows:

S = P - jQ (9)

Power triangle for a capacitive load is shown in Figure 10.

Fig. 10 - Power Triangle for a Capacitive Load

© Boris Shvartsberg, 2022 11

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

To reduce a value of a capacitive reactive power carried by transmission line in this

system we need to use a shunt reactor as a local source of inductive reactive power. The
results achieved are as follows:

• Without a shunt reactor, apparent power carried by the line SL = PL - jQL and
power factor cosϕ = PL /SL
• With a reactor, line apparent power, SL1 = PL - j(QL – QR ) < SL , and cosϕ1
= PL / SL1 > cosϕ
• Ultimately, power losses ∆P and voltage drop ∆V will be reduced after shunt
reactor is installed, i.e. ∆P1 < ∆P, and ∆V1 < ∆V

The reactive power rating of a shunt reactor may be found similarly to the selection of a
shunt capacitor in example 1 (see Figure 11).

To improve a power factor from cosϕ1 to cosϕ2, the required reactor rating QR is:

QR = Qin - Qin’ (10)

Fig. 11 - Calculation of Shunt Reactor Rating

An optimum location of shunt reactor in the system is defined by special studies. Usually
it is installed at transmission substations. Construction of a shunt reactor is similar to the
arrangement of a power transformer. The difference is that transformers have at least two
windings, while a reactor has only one. The devices used to switch shunt reactor are
circuit breakers and circuit switchers. The example of 230 kV shunt reactor is shown in
Figure 12.

© Boris Shvartsberg, 2022 12

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

Fig. 12 - 230 kV Shunt Reactor Voltage Regulation

Previously we’ve discussed how to reduce power losses and voltage drops in power
systems using compensation of reactive power with either shunt capacitors (for inductive
load), or shunt reactors (for capacitive load). However, it is not the only way for utility
company to meet their goal - to provide customer with electricity at the standard voltage
level (208/120 V, for example). Another option to compensate for voltage drops in power
lines and transformers is to use a voltage regulation. Let’s discuss the main methods used
by power companies to do it.

One of the most common means of voltage regulation is having transformer windings
with taps to change the turns ratio. A high voltage (26 – 230 kV) winding of a substation
transformer usually has no-load taps that may be changed only manually when the
transformer is deenergized. Because of this necessity to take transformer out of service,
this method of voltage regulation is used infrequently, usually when there is a substantial
load growth or seasonal load fluctuations.

Another way to regulate voltage is to have a load tap changer (LTC) on a low voltage (4 -
26 kV) winding of substation transformer, which is used for frequent changes in the
transformer turns ratio while it is carrying the load. LTC may be operated manually or
automatically, locally or remotely. For automatic operation, LTC has a control device
that monitors voltage and changes taps to keep it within certain limits. The example of a
transformer with LTC is shown in Figure 13.

© Boris Shvartsberg, 2022 13

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

Fig. 13 - 26 kV LTC on 230/26 kV 55 MVA Transformer

The 3rd option for voltage regulation is an application of voltage regulators, which are
installed on each substation feeder to change their voltage under load. The example of a
voltage regulator is shown in Figure 14.

Fig. 14 - 4 kV Voltage Regulators

© Boris Shvartsberg, 2022 14

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

It needs to be noted, that voltage regulation should not be done to raise voltage to a
standard level for the very last customer on the line, but rather to keep voltage for all the
customers inside required boundaries, for example, ±5%, which for 120V will amount to
a 114V – 126V range. It is very important to keep voltage inside these boundaries
because they are specified for all the appliances to have them functioning the way they
are intended to. For example, if voltage in your house is too low, lights will dim making
reading difficult. If voltage is too high, lights will be very bright, blinding the reader, and
requiring frequent replacement of the lighting bulbs. For explanation of this principle of
voltage regulation, let’s use Figure 15, showing distribution line supplying power from
substation A to several customers B – F.

Fig. 15 - Example of Distribution Line.

Voltage regulation is done at the substation either by changing transformer taps (under
load or with no load) for all the substation feeders or by using voltage regulators just on
the line A – F. In any case, voltage level at the substation should be installed high
enough to compensate the voltage drop in the line under maximum load conditions.
However, if regulation is based on keeping the voltage for customer F at standard level
(120 V, for example), the voltage at substation A will be set so high that customer B will
see voltage higher that allowed 5% over standard, for example 130 V. So, to make all the
customers satisfied, the voltage level at substation should be set to have voltage at
customer B no higher than 126 V and at customer F no lower than 114 V. If this can’t be
accomplished by regulating voltage at the substation, additional capacitor banks may be
placed between substation and customer F.

To understand how transformer taps may be selected, we’ll use Example 2.

Example 2. Selection of no-load taps on the high side of 26/4 kV transformer

A high voltage winding of a single phase 26.4/4.16 kV transformer has five no-load taps
to keep secondary voltage constant by changing the turns ratio correspondingly to the
changes in a primary voltage within ±5% in 2.5% increments (see Figure 16). Load flow
studies show that for a maximum summer load, the expected voltage on 26.4 kV side will
be 24.8 kV. Knowing that the lowest allowed limit for a secondary voltage is 4.08 kV
select a corresponding primary no-load tap for the summer season.

© Boris Shvartsberg, 2022 15

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

Fig. 16 - 26.4/4.16 kV Distribution Transformer No-load Taps

Solution:

• We start with calculating primary transformer voltages for all five (5) taps. The
results, which are shown in Table 1, will be used for all further iterations.

Table 1 - High Voltage Values for Taps 1 – 5 for Example 2

HV Changes in Primary Primary Secondary

Tap Voltage, % Voltage, V Voltage, V
1–6 -5 25,080 4,160
1–5 - 2.5 25,740 4,160
1–4 0 26,400 4,160
1–3 +2.5 27,060 4,160
1–2 +5.0 27,720 4,160

• For a middle high voltage tap 1 – 4 transformation factor:

N(0) = 26,400/4,160 = 6.346

During summer load maximum,

V1 = 24.8 kV and V2 = V1/N(0) = 24.8/6.346 = 3.91 < 4.08 kV.

So, requirement is not met and next lower tap should be selected

© Boris Shvartsberg, 2022 16

 Fundamentals of Reactive Power and Voltage Regulation in Power Systems – E03-011

• For the next lower tap 1 – 5:

N(-2.5%) = 25, 740/4,160 = 6.188,

V’2 = V1/N(-2.5%) = 24.8/6.188 = 4.0 < 4.08 kV

Again, requirement is not met and next lower tap should be selected.

• For the lowest tap 1 – 6,

N(-5%) = 25,080/4,160 = 6.03,

V’’2 = V1/N(-5.0%) = 24.8/6.03 = 4.11 > 4.08 kV,

This meets the requirement.

So, before summer season, the transformer should be deenergized and a tap 1 – 6 should
be installed on a high voltage winding to ensure that secondary voltage meets
requirements during a load maximum.

If the last iteration did not provide a needed result, additional means of voltage regulation
could be used, like voltage regulators installed on distribution feeders, shunt capacitors
etc.

Conclusion
This course provided an overview of fundamentals of reactive power and voltage
regulation to enable you to:

• Understand why reactive power needs to be regulated

• Draw power triangle
• List main equipment for reactive power regulation
• Size and design shunt capacitor bank for a specific power factor improvement
• Describe the mission of voltage regulation and means used to perform it
• For a specific limits of secondary voltage select a no-load taps on primary side of
transformer

References
1. IEEE Std. C37.99-1990 “IEEE Guide for Protection of Shunt Capacitor Banks”

© Boris Shvartsberg, 2022 17