IEEE 1584-2018 Arc Flash — Required Inputs, Valid Ranges, and Common Mistakes

IEEE 1584-2018 — IEEE Guide for Performing Arc-Flash Hazard Calculations — is the most widely referenced standard for quantifying arc flash incident energy in electrical systems. The 2018 edition replaced the 2002 edition with a fundamentally new empirical model based on over 1,800 laboratory tests, covering a wider range of equipment types, voltages, and electrode configurations.

Despite the standard’s comprehensive methodology, arc flash calculations are only as accurate as their inputs. The most dangerous errors in arc flash studies are not calculation mistakes — they are input mistakes. An incorrect gap distance, a misidentified electrode configuration, or the use of bus fault current instead of arcing current can shift the result by one or two arc flash PPE categories, potentially exposing workers to lethal hazard levels.

This guide covers every required input to an IEEE 1584-2018 calculation, the valid ranges for each parameter, the physical reasoning behind those limits, and the mistakes that experienced engineers still make.

The IEEE 1584-2018 model is empirically validated for systems operating between 208 V and 15,000 V (15 kV), three-phase AC. This is explicitly stated in IEEE 1584-2018, Clause 4.2.

Key considerations for the voltage input:

• Below 208 V: The standard does not provide a validated model. For 120 V single-phase systems, some jurisdictions use the simplified Lee method or consider the arc flash risk negligible. However, arcs can sustain at voltages as low as 100 V DC, so dismissing low-voltage arc flash entirely is not recommended.
• Above 15 kV: The empirical model does not apply. For systems above 15 kV, the Lee method (theoretical model based on the maximum power transfer theorem) must be used. The Lee method is significantly more conservative and typically produces higher incident energy values.
• DC systems: IEEE 1584-2018 does not cover DC arc flash. For DC systems (battery rooms, solar PV, traction), use the method in IEEE 1584 Annex D or the simplified DC arc flash calculation per NFPA 70E.

The voltage input should be the nominal system voltage (e.g., 480 V, 4,160 V, 13,800 V), not the actual measured voltage. The model was developed using nominal voltages, and using measured values (which may be slightly higher or lower) introduces false precision without improving accuracy.

Valid voltage input examples:
  208 V   (US 3-phase low voltage)
  240 V   (AU/NZ 3-phase low voltage)
  400 V   (IEC 3-phase low voltage)
  415 V   (AU/NZ/UK 3-phase)
  480 V   (US 3-phase)
  4,160 V (US medium voltage)
  6,600 V (AU/NZ medium voltage)
  11,000 V (IEC/AU/NZ/UK medium voltage)
  13,800 V (US medium voltage)

Out of range (use Lee method):
  22,000 V, 33,000 V, 66,000 V, 132,000 V

The single most significant improvement in IEEE 1584-2018 over the 2002 edition is the introduction of five distinct electrode configurations. The 2002 edition used a single generic model; the 2018 edition recognises that the geometry of the arc — how it forms relative to the equipment enclosure — fundamentally changes the incident energy at a given working distance.

The five configurations, defined in IEEE 1584-2018, Clause 4.3 and Table 1, are:

Why this matters: For the same fault current, voltage, and gap distance, the VCB configuration can produce incident energy 2–3 times higher than VOA. The enclosure acts as a focusing element, directing the arc plasma jet outward like a nozzle. An engineer who uses VOA (open air) for equipment that is actually VCB (enclosed) will dramatically underestimate the hazard.

The selection of electrode configuration requires engineering judgement. The standard provides guidance in Table 1, but the engineer must understand the physical bus arrangement inside the equipment to make the correct selection. When in doubt, the more conservative configuration (typically VCB or HCB) should be used.

The gap distance (G) is the distance between conductors where the arc initiates, measured in millimetres. It is not the working distance. It is not the bus bar spacing. It is the expected gap across which the arc will form during a fault event.

The valid range per IEEE 1584-2018, Clause 4.2 is:

IEEE 1584-2018, Table 1 provides typical gap distances by equipment type:

Common mistakes with gap distance:

• Using phase-to-phase spacing instead of arc gap: The bus bar centre-to-centre distance is not the same as the arc gap. The arc gap is the clearance between the edges of adjacent conductors, not between centres. For rectangular bus bars, subtract the bar width from the centre-to-centre distance.
• Using the minimum clearance from the equipment standard: Dielectric clearances (e.g., per IEC 61439 or UL 891) are minimum insulation distances, not arc gap distances. The actual gap may be larger than the standard minimum.
• Ignoring the valid range: If the actual gap exceeds the maximum validated value (76.2 mm for LV, 254 mm for MV), the model results are extrapolated and may not be accurate. Use Table 1 defaults when actual measurements are unavailable.

When performing an arc flash study for existing equipment, the best practice is to measure the actual gap distance with the equipment de-energised. When designing new equipment or performing studies without site access, use the Table 1 default values.

The bolted fault current (I_bf) is the prospective symmetrical RMS short-circuit current at the point of the arc flash study, assuming zero arc impedance (i.e., a solid bolted connection between phases). This value comes from a separate short-circuit study — it is not calculated by IEEE 1584 itself.

The validated range per IEEE 1584-2018, Clause 4.2 is:

The IEEE 1584 model uses the bolted fault current to calculate the arcing current (I_arc), which is always less than the bolted fault current because the arc itself has significant impedance. The relationship between bolted and arcing current depends on voltage, gap distance, and electrode configuration.

IEEE 1584-2018 arcing current model (simplified):
  I_arc = f(I_bf, V_oc, G, electrode_config)

Where:
  I_arc  = arcing current (kA)
  I_bf   = bolted fault current (kA)
  V_oc   = system voltage (V)
  G      = gap distance (mm)

Typical ratio I_arc / I_bf:
  At 480 V:    0.40 to 0.65  (arc impedance is significant)
  At 4,160 V:  0.85 to 0.95  (arc impedance is small relative to system)
  At 13,800 V: 0.90 to 0.98  (arc impedance is negligible)

Critical point — bus fault current vs arcing current: The most dangerous input mistake in arc flash studies is using the bus fault current to determine the protective device clearing time, rather than the arcing current. The protective device (circuit breaker or fuse) responds to the current flowing through it, which during an arc flash event is the arcing current, not the bolted fault current.

Because the arcing current is lower than the bolted fault current, the protective device may operate in a slower region of its time-current curve. A relay set to trip instantaneously at 10 kA will clear a 12 kA bolted fault in 3 cycles, but a 7 kA arcing current may fall below the instantaneous pickup and clear in 30 cycles on the time-delay element — ten times longer. This tenfold increase in clearing time produces a tenfold increase in incident energy.

IEEE 1584-2018 requires that the arcing current be calculated at both the full value and a reduced value (to account for arc current variability), and the clearing time must be checked at both levels. The higher incident energy result governs.

The working distance (D) is the distance from the arc source to the worker’s face and torso. It is measured in millimetres and directly affects the calculated incident energy. The inverse relationship between distance and incident energy means that small changes in working distance produce large changes in results.

The 2018 edition also introduced enclosure dimensions (height, width, depth) as inputs to the model. The enclosure size affects how the arc energy is focused toward the worker. Larger enclosures allow more energy dissipation within the box; smaller enclosures concentrate the energy outward.

Default enclosure dimensions are provided in IEEE 1584-2018, Table 8. If the actual enclosure dimensions are known, they should be used instead of the defaults, as the enclosure correction factor (CF) can shift the incident energy by ±20%.

For open-air configurations (VOA, HOA), no enclosure dimensions are required, and the enclosure correction factor is 1.0.

The arc duration (T_arc) is the total time the arc persists before being extinguished by the protective device. It is the product of the protective device’s clearing time at the arcing current level. This is arguably the most influential input in the entire calculation — incident energy is directly proportional to arc duration.

Incident energy (simplified relationship):
  E = K × I_arc^x × T_arc × distance_factor

Doubling the arc duration doubles the incident energy.
Halving the arc duration halves the incident energy.

Determining the arc duration requires:

1. Identify the upstream protective device (circuit breaker, fuse, or relay) that will clear the fault.
2. Obtain the time-current characteristic (TCC) of that device.
3. Read the clearing time at the calculated arcing current — not the bolted fault current.
4. Check both the full arcing current and the reduced arcing current per IEEE 1584-2018 variation methodology.
5. Use the longer clearing time if the reduced arcing current falls in a slower region of the TCC.

Maximum arc duration is typically capped at 2 seconds in most arc flash studies, as beyond this point the equipment is likely to be destroyed and the arc may self-extinguish. However, some practitioners use longer durations for specific scenarios (e.g., utility source impedance limiting the fault current to just above the relay pickup).

Common mistake: Using the circuit breaker’s interrupting time at the bolted fault current level. If the arcing current is significantly lower than the bolted fault current (common at 480 V), the breaker may operate on its long-time delay element instead of the instantaneous element. This can increase clearing time from 30 ms to 3,000 ms — a 100x increase in incident energy.

The following table summarises the most frequent input errors observed in IEEE 1584-2018 arc flash studies, ranked by their impact on incident energy results:

Arc flash studies are life-safety calculations. Every input must be verified, documented, and traceable. The IEEE 1584-2018 standard itself states in Clause 1 that the results are only as accurate as the data used in the calculations. Garbage in, garbage out — and in arc flash work, garbage out means incorrect PPE labels that can get people killed.