COMPARISON OF ARC FLASH CALCULATION METHODS

By Neil Van Geem, PE

General

During recent years the electrical industry has focused attention on the electrical hazard of
arc-flash and the danger of its causing severe burns to electrical workers who are in the
vicinity of energized electrical equipment during an arc. Over this time various methods have
been developed to estimate the effects of an electrical arc’s incident energy causing
temperature rise on the human body as a result of the arc and on mitigating techniques such
as levels of fire resistant clothing and work location relative to the arc. This paper reports on
the results of a very limited comparison of four methods available for incident energy
calculation. The methods include those developed and reported with the IEEE Standard
1584, “IEEE Guide for Performing Arc-Flash Hazard Calculations,” with NFPA 70E,
“Standard for Electrical Safety Requirements for Employee Workplaces – 2000 Edition,” with
a 1981 IEEE paper, “The Other Electrical Hazard: Electrical Arc Blast Burns,” by Ralph Lee,
a program of heat flux calculation from Duke Power, and with the ARCPRO program by
Kinetrics, Inc. Toronto, Ontario. All calculation methods caution that the testing and
calculations are based on selected conditions. These are methods for predicting or
estimating arc flash hazards and thus actual cases experienced in the field can be expected
to vary from these values.

Comparison of Equation Design

The IEEE Standard, Duke Heat Flux, and NFPA 70E use equations developed empirically
from tests performed with arcs, while the Lee paper and ARCPRO use equations based on
theoretical analysis and verified by comparison with some measured results.

IEEE 1584 – Calculations consider three-phase arcs in enclosures and in air. Published
input ranges are:

• Voltage of 208 to 15,000 V

• Bolted fault current of 0.700 to 106 kA
• Grounding variations
• Equipment enclosures of commonly available sizes
• Gaps between conductors of 13mm to 152 mm (0.5 to 6 inches)

The equations were developed from curve fitting of results of values measured from
extensive testing performed by the standard’s working group. Some general conclusions
resulting from their testing are:

• System X/R ratio, system frequency, and electrode material had little or no effect
• Incident energy depends primarily on arc current. Bus gap (arc length) is a small
factor
• Calculations use Lee’s equation for voltages above 15 KV
• Calculations use bolted fault current input

NFPA 70E – Calculations consider three-phase arcs in enclosures or in air. Input ranges are
similar to IEEE 1584 including use of bolted fault current input. Tested values were limited to
a distance from the arc of greater than 18” only and for bolted fault currents for the range
from 16 KA to 50 KA and for system voltages rated 600 V and below. The equations were
developed from curve fitting results of values measured on limited testing (compared to IEEE
1584). Some general conclusions resulting from the testing which developed the equations
and as reported in an IEEE paper, “Predicting Incident Energy to Better Manage the Electric
Arc Hazard on 600-V Power Distribution Systems” of the IEEE Transactions on Industry
Applications Vol. 36, No. 1, January/February 2000 were:
Effects of arcs in an enclosure are 1.5 to 2.5 that of the same arc in open air up to currents
of 30 KA and 2.5 to 2.8 for currents of 30 to 50 KA

• Only three phase arcs were considered

• Tests matched Lee’s equation calculations for open air arcs between 16 and 35KA
when calculating burn boundary distance
• Poor correlation was found between equations for this testing range and for
ARCPRO or the Duke Heat Flux calculation program.

The testing performed to develop the empirical formulas used the same test set-up as was
later used for IEEE 1584. The working group for IEEE 1584 built on the results of the testing
used in NFPA 70E and conducted many more tests to develop their empirical equations.

Lee’s Calculation – Calculations are theory based and are developed to consider maximum
arc power. It considers three-phase arcs in open air and calculations use bolted fault current
input. IEEE 1584 recommends this calculation method for medium voltage arcs (Above 15
KV) in open air at substations and for transmission and distribution systems. The IEEE 1584
software defaults to this formula for cases with voltages over 15 KV.

ARCPRO – Considers single-phase arcs in air using theoretical equations. The calculations
use arc current rather than bolted fault current input.

Published Input Ranges

• Arc current of 0.2 KA to 100 KA (verified for 3.5 KA to 21.5 KA)

• Arc duration of 0.05 to no limit in cycles (verified from 4 to 30)
• Arc gap of 1 to 20 inches (verified from 1 to12 inches)
• Source voltage (open circuit voltage across the gap in V) of “any that will sustain the
arc”
• Electrode material choice of copper or stainless steel
• Distance from arc of 0.4 to 400 inches (verified from 8 to 24)
• In an appendix it is recommended that the results of the calculation be multiplied by
1.5 to convert to an arc in a box. It further states that this gives “an extremely
preliminary approximation.”

Duke Heat Flux Calculator - This calculation is based on empirical values developed from
measurements. It is made available to the public at no charge. It considers a single- phase
arc in air, and uses arc current input.

The paper used for the NFPA 70E calculation method (See reference above) compared the
measured three phase results with the calculated results given for single phase arcs. It
reported that, “Three-phase test values of maximum incident energy for the open arcs were
from 2.5 to 3 times the values predicted by the single-phase models. Three-phase test
values of maximum incident energy for the arcs in the cubic box were 5.2 to 12.2 times the
values predicted by the single-phase models.”
Calculation Comparison

Case 1
Consider a three phase arc at a 480 volt panel at a distance of 18” from the worker, with
clearing time of 0.04 sec (2.4 cycles) given by the instantaneous operation of a 400 A
MCCB, a 1” arc gap, and with bolted fault current of 13,725 amperes:

Incident Energy for arc in air

*NFPA 70E E = 0.7985 cal/cm2 Clothing Class - 0

IEEE 1584 E = 0.6 cal/cm2 Clothing Class – 0
*Lee Method E = 0.6454 cal/cm 2 Clothing Class - 0
~+ARCPRO E = 0.17 cal/cm 2 Clothing Class - 0
~Duke Heat Flux E = 0.167 cal/cm 2 Clothing Class - 0

Incident Energy for the same arc in an enclosure (box)

*NFPA 70E E = 1.749 cal/cm 2 Clothing Class - 1

IEEE 1584 E = 1.1285 cal/cm 2 Clothing Class - 1
#*Lee Method E = 1.1295 cal/cm 2 Clothing Class - 1
++ARCPRO E = 0.51 cal/cm 2 Clothing Class - 0
#~Duke Heat Flux E = 0.29 cal/cm 2 Clothing Class - 0

Case 2
Consider the same arc conditions as for Case 1 except a bolted fault current of 18.775 KA.

Incident Energy for arc in air

*NFPA 70E E = 0.9626 cal/cm 2 Clothing Class - 0

IEEE 1584 E = 0.8 cal/cm 2 Clothing Class - 0
*Lee Method E = 0.88 cal/cm 2 Clothing Class - 0
~+ARCPRO E = 0.34 cal/cm 2 Clothing Class - 0
~Duke Heat Flux E = 0.23 cal/cm 2 Clothing Class - 0

Incident Energy for same arc in an enclosure

*NFPA 70E E = 1.62 cal/cm 2 Clothing Class - 1

IEEE 1584 E = 1.5 cal/cm 2 Clothing Class - 1
#*Lee E = 1.32 cal/cm 2 Clothing Class - 1
++ARCPRO E = 1.02 cal/cm 2 Clothing Class - 0
#~Duke Heat Flux E = 0.40 cal/cm 2 Clothing Class - 0
Case 3
Consider a three phase arc at a 480 volt Main Switchboard with a bolted fault current of 30.1
KA, a 0.5 sec. (30 cycles) clearing time, a 1.25” arc gap, and a 24” working distance.

For arc in air

*NFPA 70E E = 11.01 cal/cm 2 Clothing Class - 3

IEEE 1584 E = 8.7 cal/cm 2 Clothing Class - 3
#*Lee E = 9.94 cal/cm 2 Clothing Class - 3
~+ARCPRO E = 5.27 cal/cm 2 Clothing Class - 2
~Duke Heat Flux E = 3.28 cal/cm 2 Clothing Class - 1

For same arc in an enclosure

*NFPA 70E E = 19.2 cal/cm 2 Clothing Class - 3

IEEE 1584 E = 17.2 cal/cm 2 Clothing Class - 3
#*Lee E = 14.9 cal/cm 2 Clothing Class - 3
++ARCPRO E = 15.81 cal/cm 2 Clothing Class - 3
~Duke Heat Flux E = 5.74 cal/cm 2 Clothing Class - 2

Case 4
Consider a three phase arc at a 120/208 volt panel at the terminals of a 150 KV transformer
with a bolted fault current of 9.1 KA, an 18” working distance, a 1” arc gap, and a 0.1 sec. (6
cycle) clearing time.

For arc in air

*NFPA 70E E = 1.75 cal/ cm 2 Clothing Class - 1

IEEE 1584 E = 0.69 cal/cm 2 Clothing Class - 0
*Lee E = 0.464 cal/cm 2 Clothing Class - 0
~+ARCPRO E = 0.34 cal/cm 2 Clothing Class - 0
~Duke Heat Flux E = 0.272 cal/cm 2 Clothing Class - 0

For same arc in enclosure

*NFPA 70E E = 5.3 cal/cm 2 Clothing Class - 2

IEEE 1584 E = 1.26 cal/cm 2 Clothing Class - 2
#*Lee E = 0.81 cal/cm 2 Clothing Class - 0
++ARCPRO E =1.02 cal/cm 2 Clothing Class - 0
#~Duke Heat Flux E = 0.476 cal/cm 2 Clothing Class - 0
Case 5
Consider a medium voltage arc in air near a substation at 24.94 KV, a 60 inch (1524 mm)
work distance, an arc duration time of 6 cycles (0.1 sec.), an arc gap of 6,” and abolted fault
current of 10,000 amperes.

NFPA 70E Not confirmed for voltage above 600 V

IEEE 1584 E = 5.5 cal/cm 2 Clothing Class - 2
*Lee E = 5.497 cal/cm 2 Clothing Class - 2
(This verifies that Lee’sEquation is used in IEEE 1584 for medium voltages)
+ARCPRO E = 0.34 cal/cm 2 Clothing Class - 0
~Duke Heat Flux E = 0.112 cal/cm 2 Clothing Class - 0

Case 6 Consider a medium voltage three phase arc remote from a substation and having the
same conditions as Case 5Except a fault current of 3.55 KA

NFPA 70E Not confirmed for voltage above 600 V

IEEE 1584 E = 2.0 cal/cm 2 Clothing Class - 1
*Lee E = 1.95 cal/cm 2 Clothing Class - 1
+ARCPRO E = 0.0 cal/cm 2 Clothing Class - 0
~Duke Heat Flux E = 0.112 cal/cm 2 Clothing Class - 0

Case 7
Consider a medium voltage, three phase arc with same conditions as Case 6 except with
working distance of 48”

NFPA 70E Not confirmed for voltage above 600 V

IEEE 1584 E = 3.0 cal/cm 2 Clothing Class - 1
*Lee E = 3.05 cal/cm 2 Clothing Class - 1
+ARCPRO E = 0.17 cal/cm 2 Clothing Class - 0
~Duke Heat Flux E = 0.175 cal/cm 2 Clothing Class - 0

Case 8
Consider a medium voltage, three phase arc with same conditions as Case 6 except with
work distance of 36”

NFPA 70E Not confirmed for voltage above 600 V

IEEE 1584 E = 5.4 cal/cm 2 Clothing Class - 2
*Lee E = 5.4 cal/cm 2 Clothing Class - 2
+ARCPRO E = 0.17 cal/cm 2 Clothing Class - 0
~Duke Heat Flux Clothing Class - 0 Clothing Class - 0
Case 9
Consider a medium voltage, three phase arc with same conditions as Case 5 except with
voltage of 12.5 KV

NFPA 70E Not confirmed for voltage above 600 V

IEEE 1584 E = 0.2 cal/cm 2 Clothing Class - 0
*Lee E = 2.75 cal/cm 2 Clothing Class - 1
(IEEE 1584 does not use Lee equation for voltage under 15 KV)
+ARCPRO E = 0.34 cal/cm 2 Clothing Class - 0
~Duke Heat Flux E = 0.32 cal/cm 2 Clothing Class - 0

Case 10
Consider a medium voltage, three phase arc in air for an ungrounded 2.4 KV system, with a
bolted fault current of 3.28 KA, 24” working distance, 6” arc gap, and clearing time of 0.1333
sec. (8 cycles).

NFPA 70E Not confirmed for voltage above 600 V

IEEE 1584 E = 0.5 cal/cm 2 Clothing Class - 0
(IEEE 1584 does not use Lee equation for voltage under 15 KV)
*Lee E = 1.44 cal/cm 2 Clothing Class - 1`
+ARCPRO E = 0.68 cal/cm 2 Clothing Class - 0
~Duke Heat Flux Clothing Class - 0 Clothing Class - 0

Notes:

* Values were hand calculated using author’s published formulas.

+ The program computed values were converted to three phase conditions using the mid
range of the author’s recommended adjustment factors (1.7). Further, arc current was
assumed to be 0.5 of bolted fault current.

++ ARCPRO computed values were converted to three phase in a box using the mid range
of the author’s recommended adjustment factors (5.1). Further, arc current was assumed to
be 0.5 of bolted fault current.

# The calculated incident energy for an arc in air was multiplied by 1.75 to estimate the value
for an arc in a box.

~ Arc current was assumed to be 0.5 of bolted fault current, and the calculated value of
incident energy was multiplied by 1.7 to estimate the value for a three phase arc.

Clothing Class is determined from hazard risk category table in NFPA 70E.
Calculations for Varying Voltage
Because of varying values for incident energy in the above cases, separate comparisons
were made with all arc conditions remaining the same while varying the voltage supply. The
set conditions were for a three phase arc in air with a bolted fault current of 16 KA, a working
distance of 36 inches (914 mm), an arc gap of 2 inches (50.8 mm), and a clearing time of 12
cycles (0.2 seconds).

For varying voltage the incident energy, E, was calculated for the two methods which
calculate three phase arcs, IEEE, and Lee methods. It should be noted that the NFPA
method was not confirmed for voltage above 600 V or fault currents below 16,000 amps and
thus its results are included in the following cases only for the two cases of voltage of 600 or
less even though the method calculates for three phase arcs. The results, in calories per
square centimeters, are as follows:

35 KV
IEEE 68.6 Clothing Class - X
Lee 68.6 Clothing Class - X

24.94 KV
IEEE 48.9 Clothing Class - X
Lee 48.9 Clothing Class - X

13.8 KV
IEEE 1.6 Clothing Class - 1
Lee 27.7 Clothing Class - 4

12.5 KV
IEEE 1.6 Clothing Class - 1
Lee 27.7 Clothing Class - 3

4.16 KV
IEEE 1.4 Clothing Class - 1
Lee 8.16 Clothing Class - 3

2.4 KV
IEEE 1.4 Clothing Class - 1
Lee 4.64 Clothing Class - 1

1.0 KV
IEEE 1.1 Clothing Class - 0
Lee 1.92 Clothing Class - 1

0.6 KV
IEEE 0.9 Clothing Class - 0
Lee 0.96 Clothing Class - 0
NFPA 0.556 Clothing Class - 0
CONCLUSIONS

Based on all the comparison cases tabulated above and their varying computed values for
incident energy, it is easily seen why selecting an appropriate method of estimating incident
energy is confusing and difficult. Even though using a comparison of Clothing Class narrows
the gap between predictions somewhat, there is still disparity between the predicted arc
incident energy. The favored choice may be to select the IEEE Standards method primarily
because it is based on empirical equations developed through multiple tests of varying fault
cases. It calculates for three phase faults (which are the most prevalent form of faults for
voltages 1 KV and under) in both open air and in an enclosure.

The NFPA method uses empirical equations developed in much the same manner as the
IEEE but based on fewer test cases, and is limited in both voltage and fault current ranges.

ARCPRO and the Duke Heat Flux programs are limited to single phase arcs in air with
recommended adjustment factors to estimate three phase arcs and arcs in enclosures.

The Lee method, though accepted by IEEE for supply voltages above 15 KV, makes
assumptions about the arc current magnitude and may be overly conservative.

It would be more reassuring to see better correlation of answers between the calculation
methods studied, but it should be remembered that all these methods were developed to
give the electric industry an estimate or prediction of incident energy for a worker exposed to
an electric arc. We, who are concerned about electric safety, are left to making the best
choice from among different methods of calculation giving different answers.
Copyright © Associated Training Corporation. Any unauthorized duplication or distribution
prohibited.