CEATI 11th Annual Grounding & Lightning Conference

Houston, TX, USA October 7-9, 2019

Accurate Calculation of the Split Factor

A Numerical Study
R. Andolfato – SINT Ingegneria – Italy
D. Cuccarollo – SINT Ingegneria - Italy
 Introduction
The simulation of a grounding system requires the
knowledge of grid layout, soil model and current to
earth. This presentation is related to the evaluation
of the current to earth.

Because alternative return paths as overhead earth

wires and cable screens or armors, the current to
earth can be substantially different to the fault
current.
 Introduction
The split factor is the ratio:
Current to earth / Fault current

The fault current (the single phase to earth

fault current) can be calculated with
traditional methods or is often a data provided
by the distributor or the network owner.
 Introduction
The calculation of the current to earth on the
opposite requires a specific tool.

An accurate calculation of the current to

earth can lead to a substantial economy in the
grounding system sizing. On the other side, a
wrong current to earth can lead to oversized
or unsafe systems.
 Introduction
The problem is obviously not new.

The relevance of the topic is increasing

with the short circuit fault current values
and so with the power systems expansion.
 Introduction
The evaluation of the split factor has been widely
treated in literature and in the past some analytical
methods based on simplified assumption have been
proposed (fault current independent on grounding
impedance, uniform characteristics along lines,
ungrounded ground wires, infinite line length …).

Unfortunately, these assumptions are not always

realistic.
 Introduction
The issue today can be easily solved with a rigorous
numerical simulation.

Modern tools can consider complex scenarios

without particular assumptions or constraints.

The true limit in simulation is the knowledge of the

network parameters and the data entry process.
But some tools can do that quite easy.
 Introduction
It’s possible to consider in a very accurate way:
- Radial or meshed power networks
- Terminals and substations grounding systems
- Overhead lines with their ground wires and tower footings
- Cables with their screens or armors and junction chambers
grounding
- Heterogeneous line and cable sections
- Transposition effects
- …
 Introduction
All this is made possible by tools based on
the PCM (Phase Components Method).
This method is general and versatile and
can be applied with unbalanced and
unsymmetrical multi-phase and multi-
conductor systems and of course with
systems that involve currents to earth.
 Introduction
The module NETSTM of the program XGSLabTM is based on the
PCM and is able to perform an accurate evaluation of the fault
current distribution and then of the split factor.

In the following:
- A short comparison symmetrical Vs phase components
methods
- A numerical study for the calculation of the split factor in a
typical case
 XGSLab
XGSLabTM includes the following modules:

GSA, GSA_FD, XGSA_FD, XGSA_TD and NETS

The first 4 module are based on the electromagnetic field theory

and on the so-called PEEC (Partial Element Equivalent Circuit)
method and are used mainly for grounding, electromagnetic
interferences, cathodic protections and lightning.

The module NETS is based on the circuit theory and in particular on

the PCM.
 Network Analysis
The analysis of electrical networks can be done essentially using the
following two methods:
- Symmetrical Components Method (based on the Kirkhhoff laws
and the Fortescue technique)
- Phase Components Method (based only on the Kirkhhoff laws)

The two methods are not alternative the one to the other and
cannot be directly compared. It is however important to know
advantages, disadvantages and limits of both methods as below
described.
SYMMETRICAL COMPONENTS METHOD
Network Analysis PHASE COMPONENTS METHOD

Advantages: Advantages:
- Is a powerful analytical tool and also conceptually useful - Based only physic laws and then rigorous but
- In case of symmetrical networks the calculation of unbalanced conditions is simple also simple and intuitive
because the system is converted in balanced systems easy to solve - For general applications with balanced or
- Reduces the size of the linear system involved with the problem and requires limited unbalanced systems and with symmetrical and
computing power unsymmetrical systems (for instance overhead
- Used by many years in industry and engineering for short circuit and load flow lines with overhead earth wires or
analysis and so widely validated underground cables with screens or armors
with their grounding systems)
- Can be used with problems that involves
currents to earth (fault currents distribution)
Disadvantages: Disadvantages:
- Can be uses only for symmetrical systems or when the assumption of symmetric - Increases the size of the linear system
three-phase system is acceptable. It could be anyway used with quasi symmetrical involved with the problem and requires more
systems like the common transmission power lines (overhead lines and cables). On memory resources and computing power
the opposite, usually this method cannot be used with distribution lines, with many
single-phase loads or generators
- Can be used for short circuit analysis, power flow analysis, transient stability and so
on, but not for fault current distribution when grounding systems are involved
 Network Analysis
Taking into account the memory resources
and computing power of modern PC, the
main disadvantage of the PCM is today
overcomed.
The PCM is often a feasible approach and in
any case sometimes the only applicable
method.
 Network Analysis
The PCM (as for the SCM), requires a complete
description of the multi-phase and multi-
conductors network.
The network is represented using multi-port cells
connected using multi-port buses.
A multi-port cell can represents a network
component. With a convenient set of components
is possible to represent any practical condition.
 NETS Components
NETS includes mainly the following components:
- Feeders
- Ideal Voltage or Current Sources
- Transverse Impedances (load, fault)
- Longitudinal Impedances (fault, switch)
- Transformers
- Overhead Lines
- Overhead or Underground Cables
 Feeders
The Feeder can represents a generator or a Z1 U1, I1
network. In a three-phases network the
E1

longitudinal impedances can be calculated using E2 Z2 U2, I2

the formulas:
Z  Z  cos   j sin   En
Zn
Un, In

V2 1
cos  
Z 
1  X R
2 Ze

Scc
Ie

The short circuit power Scc depends on the rated

voltage, for instance 400 GVA at 400 kV, 180 GVA at
230 kV, 80 GVA at 132 kV. The ratio X/R is in the
range 10 – 30 for transmission networks and 3 – 10
for distribution networks.
 Voltage and Current Sources
The Ideal Voltage or U1, I1 U1, I1
Current Source can be
E1 I1

useful to represent E2 U2, I2 I2 U2, I2

ideal conditions with

impressed voltages or
currents. En Un, In In Un, In

Ideal Current Sources

can be used to
impress single phase
to earth fault Ie Ie

currents.

Ideal Voltage Source Ideal Current Source

 Impedances
The Transverse Z1 U1, I1

Impedance can be U1 in, I1 in Z1 U1 out, I1 out

useful to represent Z2 U2, I2

Z2
loads or faults.
U2 in, I2 in U2 out, I2 out

The Longitudinal
Impedance can be Zn Un, In
Un in, In in Zn Un out, In out
useful to represent
phases interruption or Ze Ie in Ie out

switches. Ie

Transverse Longitudinal
Impedance Impedance
 Transformers
The Transformer can represents
single-phase or three-phases
U1 in, I1 in U1 out, I1 out

transformers. U2 in, I2 in U2 out, I2 out

The three-phase transformers can

have the usual connections (Yy, U3 in, I3 in U3 out, I3 out

Yd, Yz, Dy, Dd, Dz) and angle Un in, In in Un out, In out

group (0 to 11).
Ie out
The transformer parameters are
Ie in

calculated on the base of the data

in the library.
 Overhead Lines
The Line can represent a Reference

chain of spans of an arbitrary

long overhead line. x (+)

Each conductor can

represents a phase conductor
(single or bundled) or an h

overhead earth wire.

The line self and mutual Soil surface

parameters are calculated on

the base of the conductors
data and location.
 Cables
The Cable can represents an arbitrary long Reference

overhead or underground cable line. Soil surface

Each single cable can includes core, screen and x (+) h

armor. The cable self and mutual parameters

are calculated on the base of the cables data
and location.

Also pipe-type
cables can be
represented.
 Application 1

Main data:
- Rated voltage of the system: 20 kV (distribution system)
- Single-phase to earth fault current at the start cable: 4354 A
- Resistance to earth at the start cable station: 0.313 Ω
- Resistance to earth at the end cable station: 0.17 Ω
- Cable length: 5000 m
- Cables: single core with screen and armor (solid bonding system)
 Application 1
Scenario 1
n. 3 single core cables

Split factor
Sf = 1090/4354 = 25%
 Application 1
Scenario 2
n. 3 single core cables
+
n. 3 single core cables

Uncoupled (far)
and in parallel

Split factor
Sf = 1400/4354 = 32%

Result is unexpected.
 Application 1
Scenario 3
n. 3 single core cables
+
n. 3 single core cables

Coupled (close)
and in parallel

Split factor
Sf = 880/4354 = 20%

The effect of coupling between close cables is relevant.

 Application 1
Currents balance. Phase Current «If»
The balance depends on
mutual coupling between
core and screen / armor
and on the impedances of Induced EMF «E = M*If»
screen / armor, and
grounding systems. «If = Ie+Id» Screen / Armor Current «Id» Current to Earth «Ie»
Resistance to Earth «Rd» Resistance to Earth «Re»

Scenario 1 Id M  Re Example
S f  1  1 Re = Rd = 1 Ω
n. 3 single core cables If Z s / a  Re  Rd M = j9 Ω
Scenario 2 M* = j12 Ω
I M  2 Re Zs/a = 0.5+j12 Ω
n. 3+3 single core uncoupled cables S f  1  d  1 
If Z s / a  2 Re  2 Rd
Scenario 3 Scenario 1: Sf = 0.27
I 2 M * 2 Re Scenario 2: Sf = 0.30
n. 3+3 single core coupled cables S f  1 d  1 Scenario 3: Sf = 0.23
If Z s / a  2 Re  2 Rd
 Application 2
Complex Scenario with: 400 kV 230 kV
1 Feeder 400 kV
1 Tr 400/132 kV
1 Feeder 230 kV
1 Tr 230/132 kV
1 Overhead line 132 kV
2 Underground cables 132 132 kV
kV (with transpositions)

Split factor NETS

Sf = 334/2240 = 14.9%
Fault
Split factor OpenDSS®
Sf = 14.7%
 Conclusion
Modern programs based on the phase components
method like XGSLabTM – NETS are able to represent
in a very realistic way complex scenarios with full
meshed, unsymmetrical, unbalanced multi-
conductor systems that involve currents to earth.

It’s than possible an accurate calculation of the

fault currents distribution and in particular of the
split factor related to a specific grounding system.
 Conclusion
It’s then possible the calculation of the right value
of the expected current to earth in a grounding
system and in this way it’s possible to avoid
expensive over-sizing or dangerous under-sizing.

The examples confirm the importance of a realistic

simulation of the actual conditions because the
calculation is often non trivial and results are
sometime unexpected.
Thanks for your attention.