Tp 01900001 e

7/28/2019 Tp 01900001 e

1/16

A Practical Approach to Arc Flash Hazard Analysis and ReductionWhite Paper

7/28/2019 Tp 01900001 e

2/16

A Practical Approach to Arc Flash Hazard Analysis and Reduction

By:

H. Wallace Tinsley III

Member, IEEEEaton Electrical130 Commonwealth DriveWarrendale, PA 15086

Michael Hodder

Member, IEEE

Eaton Electrical4120B Sladeview CresMississauga, ON L5L 5Z3

Presented at the 2004 IEEE IAS Pulp and Paper Industry Conference in Victoria, BC: IEEE 2004 - Personal use of this material is permitted.

TABLE OF CONTENTS

Abstract ....................................................................

I. Introduction ..........................................................A. Standards

B. Arc Flash Analysis

II. Generalization of Arc Flash ..............................A. Fault Magnitudes

B. Constant Energy

C. Overcurrent Device Responses

III. Time, Current, & Energy Relationship ...........A. Relationship Equations

B. Sotware Application

IV. System Models & Analysis ..............................A. Data Collection

B. Unbalanced Faults

C. Analysis Philosophy

V. Considerations & Solutions ...............................

VI. Conclusions ........................................................

Acknowledgement .................................................

References ...............................................................

12

3

6

11

121213

13

7/28/2019 Tp 01900001 e

3/16

www.eatonelectrical.com/arcash


ABSTRACT

Recent eorts to quantiy the dangersassociated with potential arc fash hazardsrely on overcurrent protection to remove agiven ault condition. The eectiveness ovarious devices is determined by a clearingtime related to the maximum availableault current or each system location. Asindustrial and commercial acilities begin to

embrace arc fash labeling procedures andbegin to recognize arc fash prevention asa part o a complete saety program, thecurrent method o calculation will allow themto quantiy the incident energy (cal/cm2)associated with a maximum, three-phaseault condition. Most aults producecurrent magnitudes less than the three-phase maximum. This paper will consider

ault current magnitudes less than thato the maximum, threephase condition

and discuss the resulting calculations orincident energy across the range o cur-rent magnitudes. Under these additionalscenarios, the perormance o variousovercurrent protection devices will bedemonstrated. Associated considerations

or design, modeling, and maintenancewill be presented.

7/28/2019 Tp 01900001 e

4/16



Index Terms Constant Energy C-Line, Arc Flash Hazard,Unbalanced Faults, Worst-Case Scena

I. INTRODUCTION

Extensive research and experimentation have led to

the development o detailed calculation methods or

determining the magnitude and intensity o Arc Flash

Hazards. These methods have been presented in IEEEStandard 1584 2002 and the NFPA 70E 2000

Edition. Recommended practice now requires that

the incident energy due to an Arc Flash be quantifed

at each system location potentially accessed by

authorized personnel while the equipment is

energized.

This requirement suggests the need or a thoughtul

understanding o the power system and the meth-

ods o calculation. This paper provides a ramework

o considerations on which to base the methods o

calculation presented in the most current standards.

These considerations include worstcase scenarios,

data collection or analysis, design concerns, and

maintenance.

A. Standards

NFPA 70E-2000 Edition, Table 3-3.9.1 requires acil-

ity personnel to wear Personal Protective Equipment

(PPE) when perorming various tasks in locations

susceptible to potential Arc Flash Hazards.[1] These

requirements are mandated on the basis o feld

experience and are categorized by associated voltage

levels. The Hazard/Risk Category is determined by the

nature o the work to be completed, the operating

voltage, and the available short circuit current or thatgeneral location in the electrical distribution system.

The Hazard/Risk Category reers to the appropriate

protective clothing and personal protective equip-

ment (PPE) to be utilized.

In 2002, the IEEE reported the results o extensive

laboratory experiments and calculations. IEEE

Standard 1584 2002 describes the procedures

and provides direction or an accurate means o

determining a sae Arc Flash Boundary and

associated Hazard Level.

The basis or this method is experimental data

recorded rom simulated arcs corresponding to bolt-

ed, three-phase ault current magnitudes measured

at the terminals o an experimental enclosure.

B. Arc Flash Analysis

The Arc Flash analysis requires the completion o a

Short Circuit Study and a Coordination Study. Theresults o the Arc Flash calculations are based on the

calculated values o ault current magnitudes ound

in the short circuit study and the associated

clearing times o overcurrent protection devices

as determined by the coordination study.

The goal o this type o analysis is to determine the

incident energy potentially present during an arc

ash event. The magnitude o the incident energy is

calculated on the basis o the available ault current,

the clearing time o associated system protection,

and the physical parameters o the system location.

Associated with this calculation is the determination

o an approach distance within which the incidentenergy level is above 1.2 cal/cm2. Appropriate Per-

sonal Protection Equipment (PPE) shall be used when

working on or near energized equipment within the

ash protection boundary.[1]

The results o the approach boundary and incident

energy calculations may be displayed in labels

on equipment enclosures to inorm and direct

acility personnel with respect to the potential arc

ash hazard.

2

7/28/2019 Tp 01900001 e

5/16



II. GENERALIZATION OF ARC FLASH

For proper evaluation o a power system with respect

to potential Arc Flash hazards, accurate generalization

o these hazards is imperative to describe the

worst-case scenario. To understand the worst-case

conditions, one must relate potential ault magni-

tudes to the clearing time associated with various

overcurrent devices.

Figures 1 & 2

7/28/2019 Tp 01900001 e

6/16



A. Fault Magnitudes

IEEE Standard 1584 2002 cautions, it is important

to determine the available shortcircuit current or

modes o operation that provide both the maximum

and the minimum available short-circuit currents.[2]

The importance o this statement is demonstrated

when an o-peak maintenance scenario is compared

with the ull load operating condition. A hazardous

arc ash condition may arise rom various causes,

and oten occurs during maintenance. Maintenance

tasks are oten perormed at times when the acility

and/or its processes are not ully operational.

Although the power system is energized, some o

the contributing motor loads may be shut down.

Thereore, during maintenance operations, when

the propensity or arc ash conditions is high, the

available ault current may be signifcantly lowerthan the calculated maximum.

To demonstrate the eect o various scenarios, we

have modeled a sample system that represents three

acilities that are supplied by a single utility substa-

tion. A portion o this system is shown in Figures 1

and 2. In the Figure 1, all contributing motors are in

service and the utility contribution is at a maximum.

In Figure 2, the low voltage motors have been

reduced to 10% o ull-load and the medium voltage

motors have been removed. The same is true or the

adjacent neighbors and the total ault contribution at

a typical 480V substation is signifcantly decreased.The magnitude o the available bolted ault current

is decreased rom Figure 1 to Figure 2 by approxi-

mately 30%. The arcing current is also reduced by

approximately 30% between the two fgures. For the

calculation o incident energy, we should consider the

range defned by this minimum calculation and this

maximum calculation or any given location.

B. Constant Energy

By the method presented in IEEE Standard 1584

2002, incident energy (E) is calculated or specifc

system locations. This calculated value o energy is

determined by the physical environment at the given

location and the duration o a previously calculated

magnitude o arcing ault current. The duration o the

ault condition is dependent on the clearing time o

the upstream overcurrent protection. This clearing

time is determined by the actual magnitude o

arcing ault current or a given occurrence.

For a given location, there exists a series o potential

arcing ault current magnitudes and theoretical

clearing times or which incident energy remains

constant. Several o these series are shown in Figure

3. On a log-log plot, these combinations o constant

energy points with respect to time and current

appear as a linear line segments. For a typical low-

voltage, grounded, enclosed substation, these

selected lines correspond to the PPE classes outlinedin IEEE Standard 1584 2002. The lowest line shown

in Figure 3 represents a constant energy o 1.2cal/cm2.

This corresponds to the upper limit o PPE Class 0.

The uppermost line represents the maximum value

o 40 cal/cm2 or which PPE Class 4 provides sufcient

protection. Above this line, no PPE class has been

Figure 3

4

7/28/2019 Tp 01900001 e

7/16



C. Overcurrent Device Responses

For the majority o system locations that are

protected by a use, the minimum available arcingault current is the basis o the worst-case calculation

or incident energy. (See Figure 4.)

For a system location protected by a circuit breaker,

the worstcase calculations vary with the regions o

the clearing characteristic. When the considered range

o ault current magnitudes alls completely within

any region o the timecurrent curve (TCC) across

which the time remains constant, the maximum

available ault current will result in the calculation o

the worst-case incident energy. Such regions include

defnite-time relays and defnite-time delay regions o

electronic trip unit characteristics. (See Figure 5.) For

regions o the TCC where the tripping characteristicis inverse or based on the I2t or I4t model, the

lower arcing ault values will correspond to longer

clearing times; resulting in the worst-case scenario.

(See Figure 5.)

Figure 4

Figure 5

7/28/2019 Tp 01900001 e

8/16



III. TIME, CURRENT, & ENERGY RELATIONSHIP

A. Relationship Equations

In order to demonstrate the worst-case arc-ash

scenario across a given range o arcing ault currents,constant-energy lines can be plotted on the TCC plot

in conjunction with tripping characteristics o various

devices.

For voltage levels less than 15kV, the IEEE 1584

2002 presents the equation or incident energy as

shown in Equation (1) [2]. The values or the variables

shown in this equation are presented in Table 1. In

equation (1), the units o energy are Joules/cm2.

Equation (2) shows Equation (1) algebraically

rearranged in order to calculate values or time with

respect to a given set o parameters. The conversion

actor between Joules and calories has also been

included so that the units o Energy (E) in Equation (2)are cal/cm2.

Equation (1):

Where:

E is energy inJoules

/cm

2

.C is a calculation actor, equal to1.0 or voltages above 1kV and1.5 or voltages at or below 1kV.

K1 is -0.792 or open confgurations (no enclosure) andis -0.555 or closed confgurations (enclosed).

K2 is 0 or ungrounded and HRG systems andis -0.133 or grounded systems.

Ia the magnitude o the arcing ault current (kA)that may be determined according to IEEE1584 2002, equation (1).

G is the gap between conductors (mm).t is the duration o the arc (seconds).x is the distance exponent.D is the distance rom the arc to the worker (mm).

Equation (2):

Where:

t is the duration o the arc (seconds).E is energy in cal/cm

2.

C is a calculation actor, equal to1.0 or voltages above 1kV and1.5 or voltages at or below 1kV.

K1 is -0.792 or open confgurations (no enclosure) andis -0.555 or closed confgurations (enclosed).

K2 is 0 or ungrounded and HRG systems andis -0.133 or grounded systems.

Ia the magnitude o the arcing ault current (kA)that may be determined according to IEEE1584 2002, equation (1).

G is the gap between conductors (mm).x is the distance exponent.D is the distance rom the arc to the worker (mm).

Equation (3) shows the linear relationship between

time and arcing current with respect to a given

energy and specifc system parameters shown in

Table 1. With the aid o curveftting sotware [3], this

relationship was ound consistent or all systemconfgurations considered.

Equation (3):

Where:

t is time in seconds.k is a unique constant based on specifc system

parameters (See Table 1 or a summary and thediscussion below or details.)

Ia is the magnitude o arcing ault current.b is a constant value = -1.081.

E=(4.184)C10[K1 + K2 + 1.081 log [Ia]+ 0.0011G]

[t/0.20][610x

/Dx]

E (0.20) 4.1667 ____________________________________t =

(4.184)C10[K1 + K2 + 1.081 log [Ia]+ 0.0011G]

[610x

/Dx]

t =k (Ia)b

6

7/28/2019 Tp 01900001 e

9/16



Constant k is determined or each system location

according to system parameters and a distance actor

related to the equipment type and the system location

voltage.

This determination was made according to the

ollowing steps:

1. For each system location considered, a fnite

series o time-current ordered pairs (Ia, t) was

ound, or which incident energy remains

constant. (See Equation (1) and Figure 3)

2. This series o ordered pairs (Ia, t) was

provided as input or the curve-ftting

sotware [3].

3. The resulting time versus current plot was

consistently ftted with a curve o the ormshown in Equation (3). The constant b (-1.081)

remained constant regardless o the system

parameters. The constant k was ound to be

unique or each new set o parameters.

The system parameters are shown in Table 1 and

include: system voltage, equipment type, bus gap

(mm), working distance (mm), enclosure confgura-

tion, and grounding. For some typical systemlocations, Table 1 shows the resulting values or

the unique constant k.

With a point defned on a TCC plot by the magnitude

or arcing ault current and the associated clearing

time or a specifc device; it is useul to defne a

corresponding line that represents all combinations

o time and arcing current or which energy remains

constant with respect to the given point. This line on

the TCC plot is called a C-line, and the points (Ia, t)

along this line o constant energy can be defned by

the constant C in Equation (4).

Table 1

7/28/2019 Tp 01900001 e

10/16



Equation (4):

For a given system location (defned by k), C is a

unique constant describing the relationship o a fnite

series o time and current combinations or which

energy remains constant. For increasing energy, C

is also increasing. Using this relationship, any two

regions on a TCC can be compared to determine the

worst-case scenario.

Consider the clearing time or both the maximum and

minimum ault conditions and let the ordered pair,

(Ia1, t1), represent the maximum arcing ault current

and the associated clearing time. Let the ordered pair,

(Ia2, t2), represent the minimum arcing ault current

and the associated clearing time. Compare as ollows:

I C1 > C2, then E1 > E2 and vice versa. The larger value

or C will correspond to the energy (E) greater value.

Using the relationship one can quickly determine

the worstcase condition between any number o sce-

narios (time and arcing current) at a given location.

B. Sotware Application

On a standard time-current curve (TCC), sotware

packages could use a location-specifcC-lineto

provide a visual representation or the severity o

several incident energy calculations within the range

o possible arcing ault conditions at a given location.

Provided with the values shown or k in Table 1, a

C-linecan be generated or each device with respect

to the bus location immediately downstream or withrespect to a selected bus downstream o several

devices. To aid in overcurrent device coordination, the

unique C-linewill visually demonstrate which setting

regions might be adjusted to reduce the arc ash haz-

ard. Figure 6 shows the tripping characteristics o two

devices. The electronic-trip circuit breaker is shown as

the 480V main breaker o a typical unit substation. The

use characteristic is representative o the primary

device on the 13.8kV side o the source transormer.

Ater determining maximum and minimum

magnitudes o the available bolted ault currents

at the substation bus, the corresponding arcing ault

magnitudes can be calculated. These arcing ault

current magnitudes are calculated according to

IEEE Standard 1584 2002 using specifc system

parameters. Given these parameters, the appropriate

value or k may be selected rom Table 1 or alterna-tively, k can be calculated or system parameters not

ound in the table.

In Figure 6, the maximum arcing ault current magni-

tude (Ia1) o 15.7kA is shown. For an arc ash event at

the substation bus, the associated clearing time o the

main breaker will be used to determine the incident

energy or this ordered pair (Ia1, t1). A clearing time o

0.323 seconds is shown or t1.

t

________C =k (Ia)

-1.081

t1

Set C1

=

/k (Ia1)-1.081

t2

and C2

= /k (Ia2)-1.081

Figure 6

8

7/28/2019 Tp 01900001 e

11/16



Similarly, or an arc ash event on the line side o

the main breaker or the secondary terminals o the

substation transormer, the incident energy is deter-

mined by the clearing time o the primary use. Thispoint on the plot or the maximum arcing current and

associated clearing time is shown as (15.7kA, 4.1s).

Figure 6 also shows the minimum magnitude o the

available arcing ault current at the substation bus.

This minimum value o available ault current

(Ia2 = 9.1kA) relates to the systemoperating scenario

when motor contributions are the lowest. The time

required to clear the potential arc ash event rom

the substation bus is 2.7 seconds, and is shown in

the long-delay region o the circuit breaker trip unit.

Likewise, or an event on the line side o the substa-

tion main breaker or the secondary terminals o thetransormer, the time required or the primary use

to clear the ault is ound to be 90 seconds.

Using the points that correspond to the maximum

value o arcing ault current, a uniqueC-lineis drawn

or each protective device characteristic in Figure 7.

From Equation (4), the C-lineor each device is deter-mined with the value o k selected rom Table 1 and

the time-current pairs associated with the maximum

available arcing ault current.

By visual inspection o Figure 7, it is shown that, or

both protective devices, the greatest incident energy

is present under the minimum ault condition. This is

evident because the point on the tripping characteris-

tic o each device that is associated with the minimum

arcing ault magnitude is shown above theC-line

that passes through the similar point associated with

the maximum available ault current. Each time-cur-

rent point on a TCC that is above a givenC-linehasa corresponding value or incident energy (E) that is

greater than the value o incident energy (E) associ-

ated with all points shown on or below thisC-line.

Figure 7 Figure 8

7/28/2019 Tp 01900001 e

12/16



The plotting o the C-Line line may be automated

within a sotware package, but can be quickly plotted,

by hand. For the main substation breaker in this par-

ticular scenario, the ollowing steps are demonstratedin Figure 8:

1. For a typical, solidly grounded, low voltage

switchgear location, select the value or k

rom Table 1: 0.6841.

The tripping characteristic o the main breaker is

plotted in Figure 8. The breakers maximum clearing

time at the maximum arcing ault current o 15.7kA is

shown to be 0.323 seconds.

2. Select (Ia1, t1) to correspond to the maximum

arcing ault current and the associated

clearing time: (15.7kA, 0.323).

3. Calculate C rom Equation (4):

Remember,Cisonlyanenergyspecifcconstant.

4. Select Ia2: 9.1kA

Thisvaluecanbeanyothercurrentvalueontheplot,but itisconvenienttousethecalculatedminimumvalue.

5. Calculate t2 using Equation (4):

6. Connect the two points with a line segment.

ThisistheC-Lineassociatedwiththesubstation mainbreaker.

Using the C-Line in Figure 8, one can be visually

observe that the minimum arcing ault condition has

a higher incident energy that the maximum arcing

condition. Following the determination o the worst-

case scenario, system changes may be recommended

or specifed to reduce the incident energy potentially

present at the substation bus.

Figure 9 shows a change in settings or the main

breaker and a change in use type or the primary

device. In both cases, the original C-lines are still

shown or comparison. For the electronic-trip circuit

breaker, the maximum magnitude or arcing ault

current now corresponds to with the worst-case

scenario. For the primary use, the minimum magni-

tude or arcing ault current remains the worst-case

t ________C=k (Ia)

-1.081

0.323 ________________C=0.6841 (15.7 103)

-1.081

16.213103

C=

Documents

Tp 01900001 e