ECpE Department

Distribution System Analysis

Dr. Zhaoyu Wang

1113 Coover Hall, Ames, IA
[email protected]
Distribution System Analysis

Acknowledgement:
The slides are developed based in part on Electric Power and Energy
Distribution Systems, Models, Methods and Applications, Subrahmanyan S.
Venkata, Anil Pahwa, IEEE Press & Wiley, 2022

ECpE Department
1. Introduction
• The objective of distribution system analysis is to determine the state of
the system including voltages, real and reactive power flow on lines, and
losses in the system.
• This requires modeling all the components in the system such as lines and
transformers and their interconnections based on the topology. In
addition, models for loads and sources connected to the system are
needed.
• Since distribution system is connected to transmission systems, which are
connected to large generators, we model the whole upper level system at
the point of coupling as an equivalent source.
• Distributed energy resources (DERs) directly connected to the distribution
system are modeled based on their characteristics.

ECpE
Department
2. Modeling of Source Impedance
• A method that is valid for both radial and loop systems is to derive the sequence
impedance values from the results of a fault study.
• The procedure involves considering different types of faults at the bus of interest
and using the equivalent circuit for each fault to determine the source impedance
for positive, negative, and zero sequences, as given below for bus m.

Where Subscripts 1,2,and 0 represents

, system nominal voltage (line neutral);positive, negative and zero
sequence quantities
, three-phase fault current at bus
, line-to-line fault current at bus
, SLG fault current at bus , fault impedance
• This method requires the presence of all three phases, as well as line-to-line and
single-line-to-ground faults, so it can only be used on buses with all three phases.

ECpE
Department
2. Modeling of Source Impedance
• Knowing the sequence values, the corresponding values in the phase domain can be obtained
by the proper symmetrical component similarity transformation.
• However, the obtained result will be an approximation because all the lines must be fully
transposed to balance the three phases for decoupling of the sequence impedances.
• This assumption works well in transmission systems but not for distribution systems. So, if we
have to represent the entire transmission system at the substation, this approach will work
well.
• For loop systems, if matrix of the entire system in phase domain is known, the diagonal
submatrix corresponding to bus is the source impedance matrix. This is also called the
driving point impedance matrix.

Note: the driving point impedance obtained based on the positive ‐sequence network topology of the system
should not be used for fault calculations because it does not account for transformer connections that are
important for the zero‐sequence network.

ECpE
Department
3. Load Models
• Under steady state, complex power S at any location in a distribution system varies
with voltage and can be described as a function of voltage V at that point, that is

• Different modeling approaches are used to describe the relationship of the above
equation.

3.1 Load Model I

• It is usual to represent both P and Q as a general polynomial function of V.

• Such a representation is valid for individual type of loads or aggregate (composite)

type of loads.
• Different conditions for the coefficients give different models. For example:
i. If only and are nonzero, and all the coefficients are zero, then,
and (Constant Power Load)

ECpE
Department
3. Load Models
ii. If only and are nonzero, and all the coefficients are zero, then,
and (Constant Current Load)
iii. If only and are nonzero, and all the coefficients are zero, then,
and (Constant Impedance Load)
• If a load is a combination of the three above mentioned types, we can combine them to find
a composite expression. For examples, tests and subsequent regression analysis on the data
show the following models for common load types:
(a) Air‐conditioning load demand (per‐unit values):

(b) Fluorescent lighting:

(c) Induction motor:

ECpE
Department
3. Load Models
3.2 Load Model II
• Composite loads, which are assumed to be mixtures of the types as discussed, can be
represented as:

where both k and l vary between 0 and 3, Vn is the initial or base value of voltage, Pn is the initial
or base value of real power, and Qn is the initial or base value of reactive power. When using
these models, we should be aware of the range of V for the models to be valid.
Some examples of using Load Model II:
• (a) If k = 1 and l = 0, it implies that the load is a constant current type with unity power factor.
• (b) If k = l = 2, the load is a constant impedance type.
• (c) If k = 2.5 and l = 2.7, it represents an aluminum reduction plant. This is a simple model and
can be determined empirically from measurements.

ECpE
Department
3. Load Models
3.3 Load Model III
• If loads (or demands) are sensitive to frequency, the frequency effects should be included. In
that case, voltage and frequency dependence are described by the following relationships:

where is the base frequency, and and are constant exponents.

• Instead of determining the exponent for this model, it is often a practice to determine the
four sensitivity coefficients. Knowing these coefficients, the new values for ΔP and ΔQ can be
determined from Δf and ΔV. All the changes are assumed to be small, thus:

ECpE
Department
3. Load Models
3.3 Load Model III
• The table shows the sensitivities coefficients
based on the results of a data survey
conducted by Electric Power Research
Institute (EPRI). The values have been
normalized to apparent power, S.
Table: Suggested values of sensitivities of real power and
• Then, P = Pn + ΔP and Q = Qn + ΔQ for reactive power to voltage and frequency changes based on
load survey results.
f = fn + Δf and V = Vn + ΔV
3.4 Load Model IV
• The model is particularly suitable for modeling uncertainties in aggregate loads at a node knowing the
demand profiles for a day, a season, or a year.
• Assuming Gaussian distribution for load demand:
, and
where, 𝑃_𝑝 is the power value at which the probability of load exceeding that value is p%, 𝑘_ 𝑝 is the
coefficient related to p, and σ is the standard deviation of the load.
• Typically, p of 10% is used in voltage‐drop calculations, and p of 50% or mean values of load are used for
loss calculations. Smaller values of p are used for overload and determination of emergency conditions.
10

ECpE
Department
4. Distributed Energy Resources
(DERs)
• Traditional distribution systems were not designed to accommodate active generation and
storage. However, with decreasing cost and advances in technology, such devices are being
deployed in distribution systems.
• DERs are defined as sources of electric power that are not directly connected to the bulk
power system but are connected to the distribution system, limited in size to 10 MVA or less.
• DER includes generators of different types and energy storage devices with the ability to
inject power into the system.
• Generators include rotating induction or synchronous rotating machines driven by burning
diesel, natural gas, bio gas, propane, or by wind or water flow. The static types of DERs
include solar photovoltaic (PV) and batteries.
• The majority of DERs are connected to the system through a power electronic interface:
• For the resources that generate AC power, the converter changes it to DC power, and an
inverter changes it back to AC power at the system frequency.
• For the resources that produce DC power, the inverter changes it to AC power at the
system frequency.
• The DERs that use a power electronics controller with embedded inverters are called
inverter‐based resources (IBRs).
11

ECpE
Department
4. Distributed Energy Resources
(DERs)
Impacts of DERs on Voltage Regulation
• Addition of DERs can impact voltage in
distribution systems.
• Previously, DERs were not permitted to actively
regulate voltage. They operated at a fixed power
factor.
• IEEE 1547 standard revised in 2018 permits DERs
to regulate voltage by injecting reactive power.
• DERs are divided into Category A and Category B.
• Category A for the systems with lower
penetration of DERs.
• Category B for the systems with higher
penetration of DERs or system with frequent
large variations in power output.
• The required reactive power capabilities of the
DERs for Category A and Category B are shown in
the Figure. These requirements apply to DERs for
continuous operation when the voltage is 0.88
and 1.1 times the nominal voltage.
Fig: Minimum reactive power capability of
Category A and B DER. 12

ECpE
Department
4. Distributed Energy Resources
(DERs)
Control Function Requirements for DERs
• Although constant power factor mode with unity power factor setting is the default
mode for DER operation unless otherwise specified by the distribution system
operator, the distribution system operator has to approve active participation of
DERs in voltage regulation.
• The 1547 standard requires DERs to have the ability to control voltage, reactive
power, and real power within the operating region. These control function
requirements are specified in the Table.

Table: Voltage, reactive power, and real power control function requirements for DER

13

ECpE
Department
4. Distributed Energy Resources
(DERs)
Control Function Requirements for DERs
• Control modes are illustrated in the figures as per the IEEE Standard 1547.

Fig: Volt-VAr characteristic for DER control

14

ECpE
Department
4. Distributed Energy Resources
(DERs)
Control Function Requirements for DERs
• Control modes are illustrated in the figures as per the IEEE Standard 1547.

Fig: Watt-VAr characteristic for DER control

15

ECpE
Department
4. Distributed Energy Resources
(DERs)
Control Function Requirements for DERs
• Control modes are illustrated in the figures as per the IEEE Standard 1547.

Fig: Volt–Watt characteristic for DER control.

16

ECpE
Department
4. Distributed Energy Resources
(DERs)
Control Function Requirements for DERs
• Standard 1547 provides suggested values for various set points shown in these
figures.
• The DERs will provide the capabilities of mutually exclusive reactive power control
modes listed in this table and will be capable of initiating any of these modes one at
a time.
• DER operator is responsible for implementing changes to settings and mode
selections upon request by the system operator within a specified time. Other
control modes mutually agreeable to the DER operator and the system operator
can also be implemented.
• Irrespective of the type of DER and the selected control mode, DERs supply real
power and either supply or absorb reactive power.
• Thus, while considering DER as a load, one must consider its real power and
reactive power as negative if delivering it, and positive if absorbing it.

17

ECpE
Department
5. Power Flow Studies
• Since a typical distribution system is unbalanced due to both design and operation, it is
imperative that the system's steady‐state performance evaluation or analysis be conducted in
three‐phase domain, or to consider each phase separately
• It is unlike transmission systems, where usually single ‐phase power flow studies are
conducted while considering the system to be balanced.
• To facilitate distribution system analysis, various models shown in the table are used for the
components of the system.

Components Mathematical Model

Substation Infinite source as a reference bus, where voltage magnitude can
be controlled using regulators and/or taps on the transformers
Feeders and Laterals Three‐phase series impedance and shunt admittance matrices
for each line section. Shunt admittance matrix is included for
long underground cable with significant charge currents
Load Complex power for each bus i
DER Complex power for each bus i

18

ECpE
Department
5. Power Flow Studies

Fig: Schematic of a three-phase line connected between buses i and j

5.1 Line Model

• The above figure shows the schematic of a three- phase line connected between
buses i and j.
• For simplicity, only the series impedance of the line () is shown.
• and is the vector of phase a, b, and c voltages at bus i and bus j respectively,
• is the vector of currents flowing in phases a, b, and c voltages from bus i to j of
the line.

19

ECpE
Department
5. Power Flow Studies

Fig: Schematic of a three-phase line connected between buses i and j

5.1 Line Model

• Then, we have the expressions as,

• is a 3 X 3 matrix, which is realized after reducing a fourth-order model of the line

by Kron’s reduction.
• For a two‐phase line, matrix has nonzero entries in the 2 × 2 submatrix
corresponding to the phases of the line.
• For a single‐phase line, it has nonzero entry in the diagonal entry corresponding to
the phase of the line. 20

ECpE
Department
5. Power Flow Studies

Fig: Representation of a three-phase Y-connected Load on bus i

5.2 Load and DER Model

Y-Connected:
• For modeling loads and DER (negative load), we consider that the real and reactive
power for each of the three phases are known separately.
• For a three‐phase Y‐connected load, the complex power drawn is shown in the
Figure. For this load connected at bus i, as shown in the figure, the vector of
complex power is given by:

21

ECpE
Department
5. Power Flow Studies
• The vectors of voltages and currents at bus i are given by,
, and
• Now, consider a matrix operation , which gives the following matrix:

Hence,

22

ECpE
Department
5. Power Flow Studies
And,

• Note that for two‐ or single‐phase loads, , , and must be truncated by removing
the entries for phases that do not exist in the load.

Delta- Connected:
• Consider a -connected load as shown in the figure.
• A relationship between complex powers, voltages,
and currents is,

Fig: Representation of a three‐phase

Δ‐connected load on bus i.
23

ECpE
Department
5. Power Flow Studies
And,

• Following that, we can compute using Kirchhoff's

current law (KCL),

Fig: Representation of a three‐phase

Δ‐connected load on bus i.

24

ECpE
Department
5. Power Flow Studies
5.3 Computing Currents
• Knowing the load or demand vectors at each bus, the corresponding current
vectors can be found using following equation given the voltage vectors at the
buses.

• The above figure shows two three‐phase feeders connected between buses i, j, and
k with respectiveFig:
loads at these
Schematic buses.
showing line sections and connected loads.

25

ECpE
Department
5. Power Flow Studies

Fig: Schematic showing line sections and connected loads.

• , , and are load current vectors at buses i, j, and k, respectively.

• and are current vectors of currents flowing on lines between buses i and j, and j
and k.
• is vector of currents entering bus i from bus h, and is vector of currents leaving
bus k toward bus l.

26

ECpE
Department
5. Power Flow Studies

Fig: Schematic showing line sections and connected loads.

• Applying KCL at buses, i, j, and k, we get,

27

ECpE
Department
5. Power Flow Studies
• Similar equations can be written for additional sections in the system. If there are
additional feeders splitting off of the main feeder, the KCL can be expanded. For
example, consider a case where two feeders are splitting at bus k as shown in the
Figure:

Fig: Feeders splitting at bus k.

• Applying the KCL at bus k gives:

• Again, if some of the feeders are two or single phase, the equations are modified to
include only the phases that exist for a given line section.

28

ECpE
Department
5. Power Flow Studies
5.4 Power Flow Algorithm
• The Source-Load-Iteration (SLI) method is an
efficient method for radial distribution system
power flow analysis.
• Computation is done in per-unit (pu) for ease of
analysis.
• Conversion of the impedances and loads to pu
values are required using appropriate base
values. Fig: Feeders splitting at bus k.
Step 1:
• The voltage at the substation is fixed or regulated;
therefore, it is known. With substation declared
as bus 1 of the system, we get

29

ECpE
Department
5. Power Flow Studies
Step 2:
• Assume a flat profile for all the other bus voltages and
set initial values of voltage vectors at all the buses equal
to the substation bus, that is

Note: (o) in the superscript is the iteration count, and is

the total number of buses in the system.
Step 3:
• Find the current vectors for all the buses that have loads Fig: Feeders splitting at bus k.
connected to them using:

[ ][ ][ ]
𝒂(𝑚 ) ∗ 𝒂(𝑚) −1 𝑎
𝑰 𝑖 𝑽 𝑖 0 0 𝑺 𝑖
𝒂𝒃𝒄(𝒎)
𝑰 𝑖 = 𝑰 𝒃 (𝑚 )
𝑖
= 0 𝑽
𝒃(𝑚)
𝑖 0
𝑏
𝑺 ;𝑖=2 , 3 , … , 𝑛
𝑖
𝑐(𝑚) c(𝑚) 𝑐
𝑰 𝑖 0 0 𝑽 𝑖 𝑺 𝑖
Note: m in the superscript is the iteration count, and it should not be confused with bus number m used in
the previous section. For DERs, determine and corresponding to the voltage at bus i and the selected
control mode while considering the reactive power capability.
30

ECpE
Department
5. Power Flow Studies
Step 4:
• Find for each feeder section j-k (section between bus j and bus k) starting from the bus at the
edge of the system and sequentially moving toward the source.
• If the system has multiple branches emanating from the main feeder, this must be done for all
branches starting from the bus at the edge.
Step 5:

• Determine the voltage drop in each feeder section by proceeding from source to load using:

• To determine , we start from bus 1 (or source) with known voltages and move toward the
loads to compute the bus voltages sequentially starting with bus 2 ,

• For all other buses,

31

ECpE
Department
5. Power Flow Studies
Step 6:
• Check for the tolerance level, which is the difference between voltages in two
successive iterations for every bus and for all three phases at each bus.

• If all the elements of are less than ε, which is an arbitrarily selected small number
for tolerance, for all the buses, the power flow converges.
• Typically, a value of 0.001 for ε gives good results. If convergence is not achieved,
repeat steps 3–6.

32

ECpE
Department
6. Voltage Regulation
Maintaining proper voltage in the distribution system is very important. There are numerous
methods to improve and control the voltage of primary distribution systems:
1. Use of load tap changing (LTC) transformers.
2. Application of voltage regulators in the distribution substation as well as on the feeders.
3. Application of shunt (or series) capacitors on the feeders or at the distribution substation.
4. Balancing of loads on primary feeders.
5. Increasing feeder conductor size.
6. Increasing primary voltage level.
7. Changing feeder sections from single phase to three phase.
8. Transferring of loads from existing feeders to new feeders.
9. Installation of new substations and primary feeders.
While several of these options are usually considered during the planning stages, LTCs, regulators,
and capacitors provide the best means of achieving good voltage regulation during the
operational stages on a continuous basis.
33

ECpE
Department
6. Voltage Regulation
6.1 Voltage Regulation Definition:
• Voltage regulation is the voltage difference between the two ends of a line defined
as percentage of the receiving end or downstream voltage.
• Mathematically:

where is the magnitude of the sending‐end voltage, and is the magnitude of the
receiving‐end voltage.

34

ECpE
Department
 6. Voltage Regulation
6.2 Approximate Method for Voltage Regulation:
• From the figure of a feeder on the right,
Fig: A feeder of resistance R and reactance X.

• And the voltage drop on the feeder is,

• A phasor diagram for this equation with the

condition that current lags voltage by an angle
is shown in the figure.
• With as the reference voltage or , and , we get:
Fig: Phasor Diagram

𝑉 𝑠 cos⁡𝛿+ 𝑗𝑉 𝑠 sin⁡𝛿−𝑉 𝑟 =𝐼(𝑅cos⁡∅ +𝑋sin ⁡∅)+ 𝑗𝐼 (𝑋cos⁡∅ − 𝑅sin ⁡∅ )

• Typically, in distribution systems, R is approximately equal to X, and δ is very small.
Hence, we can consider cosδ ≈ 1 and sinδ ≈ 0, which makes the imaginary go to
zero, and the real part giving voltage drop per mile becomes:

35

ECpE
Department
 6. Voltage Regulation
• And, the equation of voltage regulation becomes,

• Now, if we consider a load connected at the receiving end, we can compute the
corresponding current

And,

which gives

• In practice, the approximate voltage drop is an acceptable measure since the error
between the exact and approximate values is negligible.

36

ECpE
Department
 6. Voltage Regulation
• In distribution systems, usually Vs, the voltage at the substation, is controlled and
held constant for varying loading conditions, which implies that the voltage drop
and hence Vr at other buses changes.
• Obtaining Vr at a given bus for the given impedance and load values requires
iterative solution using power flow method
Example:
Consider a 12.47‐kV feeder with three point loads
as shown in Figure 4.13. The loads are
S2 = 2.5 MVA with 0.92 lagging power factor,
S3 = 3.0 MVA with 0.90 lagging power factor, and
S4 = 2.0 MVA with 0.95 lagging power factor. Fig: A single‐phase feeder supplying
three point loads.
The given impedance of the feeder, z, is (0.258 + j
0.6644) Ω/mi or 0.7127∠68.78° Ω/mi. Find the
percent voltage drop at bus 4 of the primary feeder
for the stipulated load conditions using the
approximate method. 37

ECpE
Department
 6. Voltage Regulation
Solution:
• First, compute impedances, Z, of the line sections by multiplying z by l, which is the
length of the feeder sections, or:

• Next, compute the line‐to‐neutral voltage at bus 1 for the given line ‐to ‐line voltage for
the feeder, or:

38

ECpE
Department
 6. Voltage Regulation
• Assume this voltage to be the reference voltage, or V1 = 7200 ∠ 0°. We also assume
that the voltage at each bus remains at 7200 ∠ 0°, and we compute the load currents
for each bus, or:

• Now, compute the currents in the feeder sections using KCL:

39

ECpE
Department
 6. Voltage Regulation
• Compute voltage drops in each feeder section:

• Thus, the total voltage drop in the entire feeder is the sum of the voltage drops in
individual sections, or:

40

ECpE
Department
 6. Voltage Regulation
• Therefore, the percent voltage regulation is

• Verify the accuracy on this result by finding the exact solution for this problem by
determining voltages with power flow using the source ‐to ‐load iterative method. You
will find that the error is minimal in this case due to the power factors of all three
loads being close to each other.

41

ECpE
Department
 6. Voltage Regulation
6.3 Voltage Drop on Radial Feeders with Uniformly Distributed Load

• In the previous section, we considered spot loads connected at the buses.

• In this section, we consider a case where the loads are distributed uniformly across
the length of the feeder in a rectangular service area.
• Also, we generalize it for a three‐phase feeder, assuming that all the three phases
have identical values of resistance and reactance.
• Consider the current density to be k A/mile for the feeder shown in the Figure.
With l as the length of the feeder, the current at the sending end is = kl, and the
current at the receiving end is 0.

Fig: Feeder with load distributed uniformly

in a rectangular service area 42

ECpE
Department
6. Voltage Regulation
• Fig (b) shows the plot of the magnitude of a
current assuming that the current throughout
the feeder has the same phase angle.
• From this figure, we get

or, Fig (a): Feeder with load distributed

uniformly in a rectangular service
area

• Consider a differential element of the feeder

of length at a distance . With ohms/mile as
impedance of the feeder, the differential
voltage across is,

Fig (b): Current on the feeder

43

ECpE
Department
6. Voltage Regulation
• Therefore, the voltage drop on the feeder is

• We can also compute losses on the feeder by

considering differential power loss with
ohms/mile as the resistance of the feeder.
Fig (a): Feeder with load distributed
uniformly in a rectangular service
Then, the power loss per phase is area

Fig (b): Current on the feeder

44

ECpE
Department
6. Voltage Regulation
• The total power loss for the three phases is

• The results shown above are useful for planning studies because while planning we
do not have the system topology and actual load values available. All we have are
projected values.
• These approximations are also useful in optimizing operating scenarios where
precise calculations may slow down the process of obtaining the final solution. In
those cases, approximation can be used to narrow the solution space, and precise
calculations can be implemented on a smaller solution space, thus, speeding the
overall computation in search for the optimal solution.

45

ECpE
Department
6. Voltage Regulation
6.4 Voltage Drop on Radial Feeders Serving a Triangular Area

• This section, we consider a case in which the

service area of a feeder is triangular with load
distributed uniformly in the service area.
• Let be the current density in amperes per
square mile.

• Next, we get an expression for current at

distance . Fig: Feeder serving a triangular area
with fixed load density

Substituting the value of gives

46

ECpE
Department
6. Voltage Regulation
Note that . Substituting this ,

• Again, consider a differential element and write

an expression for the differential voltage drop

• Further, the voltage drop across the feeder is

Fig: Feeder serving a triangular area

or with fixed load density

47

ECpE
Department
6. Voltage Regulation
Similarly, we can find power loss, which is

Then, the power loss per phase is

And,

Fig: Feeder serving a triangular area

with fixed load density

48

ECpE
Department
7. Fault Calculation
• The standard procedure for fault calculation in power systems requires determining
Thevenin's equivalent for positive‐, negative‐, and zero‐sequence networks at the
point of fault.
• The sequence networks are connected to each other based on the type of fault.
• However, decoupling of circuits in the sequence domain works only under the
conditions of symmetry.
• For example, consider a three‐phase feeder with impedance matrix in the phase
domain as shown below,

• The conditions of symmetry require that,

, and

49

ECpE
Department
7. Fault Calculation
• While for transmission lines these are satisfied with transposition, they are not
feasible for distribution feeders. Distribution feeders are rarely transposed and also
can be of two or one phase in addition to three phases.
• Since single‐ and two‐phase feeders do not have models in the sequence domain,
symmetrical component‐based analysis is not applicable to them. Also, for three ‐
phase part of the system, symmetrical components do not provide any advantage
because the sequence impedance matrix is not diagonal.
• Hence, positive‐, negative‐, and zero‐sequence impedances cannot be decoupled
from one another.
• However, we can implement a solution technique in phase domain to determine
fault currents in distribution systems. We initially consider a radial distribution
system connected to the bulk power system with no additional sources in the
system.

50

ECpE
Department
7.1 Prefault System

Fig: An example Distribution System

• A general distribution feeder connected to the substation bus with multiple feeder sections as
shown in the Figure. While the main feeder is a three ‐phase feeder, the laterals can be three,
two, or single phase. Each feeder section is modeled by its impedance matrix.
• For illustration, we are considering delta ‐wye ‐grounded connection for the transformer, but
the method will work for other configurations too.
• The system in the figure represented in the oval represents the entire power system upstream
of the transformer.
• This system can be represented by a Thevenin equivalent impedance matrix and a voltage
source vector for the three phases.
51

ECpE
Department
7.1 Prefault System
• The equivalent voltage source is considered balanced with 1
pu magnitude, i.e.
, where subscript s is the source.
• All the prefault load currents are considered to be 0 with the
assumption that load currents are much smaller compared to
the fault current.
Fig: An example Distribution System

• Directly determining the Thevenin equivalent matrix in the phase domain is difficult. An
indirect approach is considered due to the difficulty.
• Utilities usually have information on equivalent positive ‐, negative ‐, and zero ‐sequence
impedance values at different buses in the transmission system.
• The values can also be computed using the approach discussed in Section 2 of the slide
(Modeling of Source Impedance).
• The values are denoted as , , and at bus m.
• The zero‐sequence impedance value only includes the impedance of the transformer due to
the selected transformer connection. 52

ECpE
Department
7.1 Prefault System
• Delta/wye‐grounded transformers create an open
circuit in the zero‐sequence circuit, where current
flows through the transformer only on the wye‐
grounded side.
• The transmission system is assumed to be fully Fig: An example Distribution System
balanced. As a result, the sequence networks are
decoupled.
• The impedance matrix of the system in the sequence domain can be written as
follows:

•Now, we use transformation to convert this matrix to phase domain

53

ECpE
Department
7.1 Prefault System
where a is a complex number equal to 1∠120o
• Let bus i be the candidate bus for faults.
• Since all the feeder sections between the substation
bus (bus m) and the faulted bus are in series, we can
add their impedance matrices to compute the
equivalent impedance matrix.
Fig: An example Distribution System
• All the feeder sections not in the path from the
substation to the faulted bus do not influence the fault
current and thus are not included in the calculations.
• Consider this matrix to be:

54

ECpE
Department
7.1 Prefault System
• Now, this matrix is added to the system impedance matrix to get the impedance
matrix from the equivalent source to the faulted bus,

Note: This summation can only be done if both matrices on the right ‐hand side are in
per unit, or impedances are referred to the low‐voltage side of the transformer due to
different voltage levels of the distribution system and the bulk power system.

55

ECpE
Department
7.2 Three – Phase Fault

Fig: Three – phase fault at bus i

• For a three‐phase fault at bus i, current will flow from the substation to bus i.
• Since there is no additional source, no other currents will flow.
• Also, we assume that the fault impedance is 0.
• If the fault is a high impedance fault, these assumptions will not be fully valid. The
Figure above shows the conditions for a three‐phase fault.
• The fault currents are, .

56

ECpE
Department
7.2 Three – Phase Fault
• The equation from the voltage drop from the source to bus i is,

• For a solidly grounded fault, all the voltages at bus i will be 0.

• Applying this condition and substituting values of the source bus
voltage, we get,
Fig: Three – phase fault at bus i

• In the next step, we get,

57

ECpE
Department
7.2 Three – Phase Fault
• If the system has additional sources, they can be applied one at a time while
removing the equivalent source at the substation.
• Source impedance matrix needs to be known and the equivalent impedance matrix
from each source to bus i must be determined.
• Voltages at the source terminal are required for the calculations.
• For inverter-based sources, their characteristics under faults and operating rules must
be known.
• Some inverter‐based sources are automatically disconnected during fault, and most
are adjusted to produce a smaller current by adjusting the impedance of inverter,
which makes fault calculations challenging.
• However, if we can compute fault currents for all the sources one at a time, all the
fault currents can be added using superposition to compute the cumulative current.
• Since the circuit is linear, superposition can be applied without affecting the results.

58

ECpE
Department
7.3 Double-Line-to-Ground
(DLG) Fault
• For a fault at bus i, current will flow from the substation to
bus i on the two faulted phases and no current on the third
phase. As shown in the Figure, a fault between phases b
and c is considered.
•The currents for this fault are

• The voltage drop equation from the source to bus is:

• Discard the first row and the first column and write the Fig: DLG fault between phases b and c at bus i.
equations in a reduced form, as,

59

ECpE
Department
7.3 Double-Line-to-Ground
(DLG) Fault
•Further,

• The same approach can be used for faults on any

other two phases.
Fig: DLG fault between phases b and c at bus i.

60

ECpE
Department
7.4 Single-Line-to-Ground (SLG)
Fault
• For a fault at bus i, current will flow from the
substation to bus i on the faulted phase and no
current on the other two phases.
• For illustration, we consider a fault on phase a as
shown in the figure.
Fig: SLG fault on phase a at bus i.
•The currents for this fault are

• We can write an equation for voltage drop from the

equivalent source to bus as,

61

ECpE
Department
7.4 Single-Line-to-Ground (SLG)
Fault
• Discard the last two equations and keep only the first
one, or

Therefore,
Fig: SLG fault on phase a at bus i.

• Again, the same approach can be used for faults on

other phases.

62

ECpE
Department
7.5 Line-to-Line (LL) Fault
• For an LL fault at bus i, current will flow from the
substation to bus i on the faulted phases and no
current on the third phases.
• For illustration, we consider a fault between phase b
and c as shown in the figure. Fig: LL fault between phases b and c at bus i.

• For this fault, the currents will not flow to the ground
but return through the second phase, as,

• Hence, the currents for this fault are:

63

ECpE
Department
7.5 Line-to-Line (LL) Fault
• Also, the voltages at bus for the faulted phases are equal,

• Therefore, the equation for voltage drop from the

equivalent source to bus is Fig: LL fault between phases b and c at bus i.

• Expanding the last two equations gives

And.

64

ECpE
Department
7.5 Line-to-Line (LL) Fault
• Rearranging and expressing in the matrix form gives

• Solving it gives both and as well as and .

Fig: LL fault between phases b and c at bus i.

65

ECpE
Department
7.6 Symmetrical Component-based Fault
Analysis
• Although symmetrical component‐based fault analysis does not work well for
distribution systems, it can be used to get approximate results only for the part of
the system that has all the three phases.
• The first step is to determine the three-phase impedance matrix from the
substation bus (bus m) to the faulted bus (bus i).
• The general equation for this impedance matrix is given by,

• The next step is to transform this matrix to sequence domain, which gives:

66

ECpE
Department
7.6 Symmetrical Component-based Fault
Analysis
• The resulting sequence impedance matrix is not diagonal, but for approximation,
we discard the off-diagonal terms.
• The diagonal entries are considered as the zero-, positive-, and negative-sequence
impedances of the distribution system from the substation bus to the point of fault.
• To determine the pre-fault circuits in the sequence domain, we add the respective
impedances from the equivalent source to the substation bus and the impedances
from the substation bus to the faulted bus.
• All impedances must be converted to per unit before adding them.
• The voltage source in the phase domain is balanced, providing a source only for the
positive sequence.
• Hence, we can determine the three pre-fault sequence domain circuits as shown in
the figure (next slide).

67

ECpE
Department
7.6 Symmetrical Component-based Fault
Analysis

Fig: Pre-fault positive‐, negative‐, and zero ‐sequence equivalent circuits with respect to bus i.

• Note that in the figure,

68

ECpE
Department
7.6.1 Three-phase Fault
• For a three‐phase fault, the system stays
balanced, and the negative‐ and the zero‐
sequence currents are 0.
• To compute the positive‐sequence current,
we create a short circuit across the positive‐
sequence circuit as shown in Figure.
• From this circuit,
Fig: Three‐phase fault in sequence domain.

• Now, we convert them to the phase domain,

69

ECpE
Department
7.6.2 DLG Fault
The conditions for this fault in phase domain are
and .
Transforming these conditions to sequence domain
gives,

And,

Fig: DLG fault in sequence domain.

Simplifying gives,

or,

70

ECpE
Department
7.6.2 DLG Fault
Also,

or,

Therefore,

• We use conditions given by: Fig: DLG fault in sequence domain.

and , to create the circuit in sequence domain shown

in the Figure .

71

ECpE
Department
7.6.2 DLG Fault

and,

The sequence currents and voltages can be

transformed to the phase domain using the relevant
equations.

Fig: DLG fault in sequence domain.

72

ECpE
Department
7.6.3 SLG Fault
• The conditions for this fault in phase domain are
and .
• Transforming these conditions to sequence
domain gives

or,

Also,

• Based on the above two equations, we can create

a circuit shown in the Figure.

Fig: SLG fault in sequence domain.

73

ECpE
Department
7.6.3 SLG Fault
• From this circuit, we get

And,

Fig: SLG fault in sequence domain.

74

ECpE
Department
7.6.4 LL Fault
• The conditions for this fault in the phase domain
are , and .
• Transforming these conditions to sequence
domain gives

Fig: LL fault in sequence domain.

and

Therefore,

and

75

ECpE
Department
7.6.4 LL Fault
or

• Based on the equations,

, and , we can create a circuit as shown in the figure.
Fig: LL fault in sequence domain.

and

76

ECpE
Department
Thank You!

77

ECpE
Department