Calculating Arc Flash Hazard Levels

Discover common mistakes in calculating arc flash hazard levels and how to
avoid them.
By Peter R. Walsh, PE, Ferraz Shawmut Inc., Concord, Mass. -- Consulting-Specifying Engineer,
12/1/2008 http://www.csemag.com/article/CA6623727.html

IEEE Standard 1584-2002, Guide for Performing Arc Flash Hazard Calculations, is the most
widely used method of calculating arc flash hazard levels, and a realistic available fault current
value provides critical input for proper evaluation. The analysis method requires a second
calculation at 0.85 of the originally calculated arc fault current. This calculation is designed to
transform the given available fault current and other parameters to a calculated arc flash current
value.
Relying on the IEEE equations to compensate for an inaccurate available fault current can yield
unacceptable results. IEEE 1584 wasnt designed with safety factors to accommodate all bolted
fault current inaccuracies. This article focuses on some specific pitfalls in calculating the arc
fault current for up to 1 kV.
IEEE 1584 BACKGROUND
Arc flash energy can inflict injury on nearby workersand greater potential energy yields
greater hazards. Engineers and facility operators are now determining the correct arc flash
boundaries and personal protective equipment (PPE) requirements to protect workers from arc
flash dangers. The OSHA Code of Federal Regulations is mandating adequate protection
required by law. NFPA 70E: Standard for Electrical Safety in the Workplace was developed by
consensus to explain how to comply with the OSHA laws.
NFPA 70E-2004 requires arc flash hazard analyses by either calculation or its table method.
The 2009 edition will require visible posting of analysis results on the equipment. A calculation
method was refined in the late 1990s and formally documented in IEEE 1584-2002, Guide for
Performing Arc Flash Hazard Calculations.
Other calculation methods are used, but IEEE 1584 is the most widely applied and accepted,
and can be more easily defended in a court situation.
IEEE 1584s calculation method predicts the arc flash dangers in terms of an arc flash
protection boundary and the PPE level needed for worker safety. Engineers often use industry
software packages to calculate arc flash hazards; however, without proper training, engineers
can easily make erroneous conclusions. Common misapplications come from making
assumptions similar to short circuit analyses, which arent valid with arc flash analyses.
THE ROLE OF ARC FLASH CURRENT
To following equation demonstrates the concept of arc flash energy:
E = I2 * R * t
Where:
E = Energy released from arc flash in joules

I = Current through the arc flash in amps

R = Resistance of the arc flash in ohms
t = Duration of the arc flash in seconds
While this equation explains the energy concept, the variables are complex to use. The arc
current value isnt the available bolted fault current; its a smaller value because of the series
arc flash resistance in the circuit. From this equation one can conclude that the arc current
value greatly affects the resulting energy. The arc flash duration is also directly proportional to
the energy released.
The upstream protective device operation controls the arc flash duration. A fuse or properly
maintained overcurrent protective device has a predictable time to open the circuit with a
specific arc current value. Thus, arc current impacts the released energy in two ways: directly
through the current itself, and then through interacting with the overcurrent protective device to
change the duration. This dual role of the arc current can disqualify some typical assumptions
made with bolted fault-interrupting current analyses.
OVERVIEW OF THE CALCULATION PROCEDURE
To determine the required level of PPE, first calculate the heat energy density at a standard
distance. This energy density can then be adjusted for the distance to the worker and the
different channeling effects of an arc flash occurring in open air, as compared to those of an arc
flash in a box.
The IEEE 1584 arch flash calculation includes nine steps:
Step 1: Collect the system and installation data
Step 2: Determine the system modes of operation
Step 3: Determine the bolted fault currents
Step 4: Determine the arc fault currents
Step 5: Find the protective device characteristics and the duration of the arcs
Step 6: Document system voltages and classes of equipment
Step 7: Select the working distances
Step 8: Determine the incident energy for all equipment
Step 9: Determine the flash-protection boundary for all equipment.
Commercially available software programs typically use this nine-step procedure. However, as
mentioned above, without proper training, incorrect assumptions for data collection can be
made. Incorrect data in the software program could yield inaccurate results.

For example, if the collected system data in Step 1 included only an estimate of available fault
current from the utility that was actually the utility maximum value, that value could be
misleading. An additional calculation with the minimum available fault current is required. The
worst-case energy release can occur at either the minimum or maximum current value.
DEVELOPMENT OF THE ARC FLASH CURRENT EQUATION
The IEEE 1584 developers used an empirical calculation method instead of a theoretically
based equation for the kV and below analysis.1 This empirical method was derived by taking
data from laboratory-controlled conditions and altering many variables. The effects were
examined on the arc current and the resulting released arc energy. The calculation considered
open-circuit voltage, system grounding, bolted fault current, X/R ratio, gap between electrodes
and box, and box size. The equations were developed using statistical analysis programs,
including regression and curve-fitting analyses.
Upon completion, some variables were found to be more significant than others. Arc current
depends primarily on the available bolted fault current, and arc time is proportional to the energy
released from the arc fault.
The arc current can be found from equation 36 in IEEE 1584-2002,
as shown in Figure 1. Used in its stated range, it has an R-square
value of 98.3% (see Figure 2 on left). This means its a predictor of
the arc current value under standard laboratory conditions if the
bolted fault current, system voltage, configuration, and distance
between conductors are entered into the formula correctly.
The first use for the arc fault current is to calculate the heat density
released by the arc flash for a standardized time. The equations developed in IEEE 1584
depend on this current value for subsequent steps. The heat energy density value will be used
with the duration of the arc flash to find the resulting energy released.
Second, the arc fault current indicates duration by using the arc current going through the
overcurrent protection device. This device opening time is often nonlinear, so a small change in
current can result in a major duration variance.
Figure 3, on left shows a device with a steeply sloped timecurrent curve. A minor change from 3 kA to 2.5 kA in arc
current through the circuit breaker could result in the time
duration increasing from 0.02 s to 8 s, a factor of 400. For
this reason, the IEEE 1584 procedures require two separate
calculations. Although the equations are accurate, variables
in the arc flash event create a range of possible values.
The initial IEEE equation was modified to give lower arc fault
currents in 95% of the situations, to be safe when the actual
arc flash draws less current than the average. The final
equation was developed from laboratory data, where the
exact available fault current was known.
The resulting IEEE 1584 procedure for arc fault current determination uses an accurate bolted
fault current for the first calculation and requires a second calculation using a 0.85 factor of the

bolted fault current. Finally, the calculation uses the worst case for the total energy released.
Sometimes the lower bolted current has much more energy released and has a higher hazard
level.
Data input from the research labs was checked using the nine IEEE 1584 procedures against
the resulting recommended PPE levels. The initially proposed equations didnt have enough
safety factors built into them at 1 kV or less.
The final incident energy equation has a calculation factor of 1.5 that makes the total process
safer, as described in Figure 4.
This results in PPE that has sufficient protection for 95% of incidents and, when added to the
0.85 procedure, results in a 95% confidence level. Table 1 illustrates the achievement of the
95% confidence through the selection of a 1.5 calculation factor. This is only valid using the
correct values of the available bolted fault current. Further information on the variability of the
arc flash safety PPE has been investigated.2,3
COMMON ERRORS IN DETERMING HAZARD LEVELS
The first common error in using available fault currents to determine hazard levels is assuming a
high fault current. It can be confusing when someone trained to determine available bolted fault
currents becomes responsible for determining arc flash hazard levels. If the person is trained
only to calculate the available fault currents to determine interrupting rating requirements,
rounding up the available fault currents can be misconstrued as a conservative procedure.
Utility companies usually give out the highest available fault current from their connections,
because they assume customers are calculating their interrupting rating requirements.
However, calculating arc flash hazard levels requires both the minimum and maximum available
fault current. A person trained only in calculating interrupting rating requirements finds this
concern with minimum ratings counterintuitive.
If the arc flash hazard calculations are only performed with the highest possible available fault
current, the resulting hazard calculations could be too low. Therefore, a conservative
assumption for an interrupting rating calculation could be a dangerous assumption for an arc
flash hazard analysis.
The second common error using available fault current values is to assume that accurate values
arent necessary, because the 0.85 procedure will compensate for approximate values. The
0.85 factor was developed to achieve safe results only with the actual bolted fault current
known.
The arc fault current equation was developed empirically from data when the actual fault current
was known in the laboratory. The 0.85 multiplier procedure predicts the minimum arc fault
current 95% of the time. If the actual available fault current is lower, the arc current will be lower.
This could result in an arc current duration 400 times longer with significantly more arc flash
energy released.
In short, at least four calculations are required. The first two involve the highest available fault
current and the lowest available fault current. These are called scenarios. The second two
calculations are then performed with 0.85 times the bolted fault current of each scenario. The
most hazardous values are used for future steps, ensuring that such conditions are identified.

SUGGESTED PROCEDURE
Obtain the best minimum and maximum available fault current values from the utility. Be
prepared to explain why the minimum and maximum values are needed. If the utility is
uncooperative, use engineering judgment to determine the minimum and maximum values.
These two values will form the basis of at least two scenarios for the calculations.
Consider other scenarios as well. Some examples are large motor loads (both running and off),
on-site generation used in sole-source and parallel with the utility configurations, and tie circuit
breakers in every allowable condition. The number of scenarios required increases rapidly with
the systems greater complexity.
Calculate each scenario at the full arc fault current value, and then again using a 0.85 factor, to
determine the most hazardous condition. Use the worst case of all the scenarios, unless
maintenance practices ensure that some specific scenarios wont occur.
SUMMARY
IEEE 1584 is the standard most recognized by codes and regulations for calculating arc flash
dangers and PPE for up to 1 kV. For that reason, it has been incorporated into standard
industry software programs. Engineers, familiar with calculating bolted available fault currents
for interrupting ratings, tend to make similar assumptions when calculating arc fault currents. Be
aware, however, that some invalid assumptions can cause errors.
PPE Level

Two High

One High

Same

One Low

Two Low

Calculation Factor
1.00

10

129

25

1.25

30

113

21

1.50

49

106

1.75

75

86

1.90

82

79

Table 1: This table shows calculated versus actual PPE required for LV data from IEEE 1584.

The proper determination of fault currents is critical. The IEEE equations incorporate some
general safety factors, but using inadequate bolted fault current data can provide unsafe results.

Author Information
Walsh is a senior field engineer for Ferraz Shawmut Inc., Concord, Mass. He is a member of the National
Electrical Code (NEC) CMP No. 4 for the 2008 code cycle. He also is a member of IEEE, NFPA, and the
International Assn. of Electrical Inspectors (IAEI).

REFERENCES

1. Wilkins, R., M. Allison, and M. Lang. Improved Method for Arc Flash Hazard Analysis.
Presented at IEEE Industrial and Commercial Power Systems Technical Conference,
May 2004.
2. Lang, M., and K. Jones. An Evaluation of Additional Test Configurations for Future Arc
Flash Models. Presented at IEEE Electrical Safety Workshop, Calgary, Canada,

February 2007.
3. Neal, T., and M. Lang, Impact of Arc Flash Test Conditions on Arc Ratings of PPE.
Presented at IEEE Electrical Safety Workshop, Dallas, March 2008.