🔥 AI for Arc Flash Analysis: Automating Incident Energy Calculations and Safety Documentation

Why Arc Flash Analysis Benefits from AI Automation

An arc flash hazard analysis is one of the most safety-critical deliverables in electrical engineering. IEEE 1584-2018 provides the calculation methodology; NFPA 70E translates the results into workplace safety requirements; and OSHA 29 CFR 1910.333 mandates that workers be protected from electrical arc flash hazards. The deliverable — a study report with incident energy values, arc flash boundaries, and PPE requirements for every piece of electrical equipment — must be accurate, complete, and regularly updated when the power system changes.

The calculation engine (ETAP, SKM Power*Tools, or DAPPER) handles the IEEE 1584 math. But the downstream work — interpreting results, generating PPE labels, flagging outliers, drafting study reports, and tracking remediation actions — is where AI creates significant time savings without compromising the engineering judgment that must remain with the licensed professional.

This article covers how to integrate AI into each phase of an arc flash study, from importing software outputs to generating compliant PPE labels and writing the study report narrative.

IEEE 1584-2018: The Calculation Framework

The 2018 revision of IEEE 1584 introduced a significantly updated empirical model for arc flash incident energy calculations. The key improvements over the 2002 edition include a more accurate model for systems below 600V, a new electrode configuration factor (the model now accounts for vertical, horizontal, and vertical-in-a-box conductor orientations), and a broader range of validated test data covering 208V to 15,000V systems.

The fundamental outputs of an IEEE 1584-2018 calculation for each bus are:

• Arcing current (Ia): the reduced fault current that flows during an arcing fault, calculated from the bolted fault current using empirical factors. The model calculates both the maximum and minimum arcing current to find the worst-case incident energy.
• Incident energy (IE): the thermal energy at a specified working distance, expressed in calories per square centimeter (cal/cm²). A typical working distance for 600V equipment is 18 inches (457 mm).
• Arc flash protection boundary (AFPB): the distance at which a person would receive 1.2 cal/cm² of incident energy (the onset of a second-degree burn) if an arc flash occurred.
• PPE category: NFPA 70E Table 130.5(G) assigns PPE categories 1 through 4 based on incident energy levels. Category 1: 1.2–4 cal/cm², Category 2: 4–8 cal/cm², Category 3: 8–25 cal/cm², Category 4: 25–40 cal/cm². Equipment with IE > 40 cal/cm² is classified as Danger and requires engineering remediation before work can be performed.

Understanding this framework is essential context before using AI to interpret study results or generate labels. The AI output must always reflect the correct PPE category thresholds from the current edition of NFPA 70E.

Importing ETAP/SKM Results and AI-Assisted QA

Modern arc flash study software exports results in CSV or Excel format, with one row per bus: bus ID, voltage, bolted fault current, arcing fault current, incident energy, arc flash protection boundary, and PPE category. This tabular data is ideal for AI processing.

Import the export file into Python and use AI to accelerate the QA pass:

import pandas as pd

df = pd.read_excel('arc_flash_results.xlsx')

# Flag high-energy buses
df['Hazard_Flag'] = df['IE_cal_cm2'].apply(
    lambda ie: 'DANGER (>40)' if ie > 40
               else 'High (>25)' if ie > 25
               else 'Elevated (>8)' if ie > 8
               else 'Standard'
)

# Generate PPE category label text
def ppe_label(row):
    if row['IE_cal_cm2'] > 40:
        return f"DANGER — IE={row['IE_cal_cm2']:.1f} cal/cm² — DO NOT WORK UNTIL MITIGATED"
    cats = {4: (25, 40), 3: (8, 25), 2: (4, 8), 1: (1.2, 4)}
    for cat, (lo, hi) in cats.items():
        if lo < row['IE_cal_cm2'] <= hi:
            return (f"{row['Bus_ID']} — {row['Voltage_V']}V — "
                    f"IE={row['IE_cal_cm2']:.1f} cal/cm² @ {row['Working_Dist_in']}" — "
                    f"AFPB={row['AFPB_in']:.0f}" — PPE Cat {cat}")
    return f"{row['Bus_ID']} — IE={row['IE_cal_cm2']:.1f} cal/cm² — Check"

df['PPE_Label_Text'] = df.apply(ppe_label, axis=1)
df.to_excel('Arc_Flash_QA_Report.xlsx', index=False)
print(f"DANGER buses: {(df['Hazard_Flag'] == 'DANGER (>40)').sum()}")
print(f"High IE buses (>25): {(df['Hazard_Flag'].str.startswith('High')).sum()}")

With this script, you generate a QA report in seconds. AI then drafts the summary narrative: "Summarize these arc flash results: 45 buses studied, 2 with IE > 40 cal/cm² (DANGER), 5 with IE 25–40 cal/cm² (Category 4), 12 with IE 8–25 cal/cm² (Category 3). Voltage range 208V to 15kV. Draft a study findings paragraph for an engineering report per NFPA 70E and IEEE 1584-2018."

AI Prompt Templates for PPE Category Lookup and NFPA 70E Compliance

AI is highly effective for NFPA 70E compliance questions, particularly for engineers who perform arc flash studies infrequently or work across multiple editions of the standard. Use these structured prompts for consistent, traceable results:

• PPE category lookup: "Under NFPA 70E 2024 Table 130.5(G), what PPE category is required for an incident energy of 11.4 cal/cm² at an 18-inch working distance on 480V equipment? List the required minimum PPE clothing and equipment for this category."
• Working distance check: "My arc flash study shows IE = 6.8 cal/cm² at the standard 18-inch working distance on a 480V MCC. The actual task requires working at 12 inches. Recalculate the incident energy at 12 inches using the inverse square relationship and determine the required PPE category." Note: the inverse square relationship (IE_new = IE_ref × (D_ref/D_new)²) is a simplified approximation; IEEE 1584-2018 provides the more accurate method.
• Mitigation strategies: "Bus MSB-1 has an incident energy of 52 cal/cm². List engineering mitigation strategies to reduce this below 40 cal/cm² per NFPA 70E. Consider: adding a zone-selective interlocking (ZSI) scheme, installing an arc flash detection relay (AFDR), adding a maintenance mode setting to the main breaker, or adjusting coordination settings."
• Label content check: "Review this arc flash label text for NFPA 70E compliance: 'Bus P2A — 480V — IE=7.2 cal/cm² — AFPB=24 in — PPE Cat 2.' What additional information does NFPA 70E 130.5(H) require on arc flash warning labels?"

Generating Arc Flash Study Reports with AI

The arc flash study report is a formal engineering deliverable that must document the study methodology, assumptions, results, and recommended actions. AI can draft every section from structured data, saving 4–6 hours of report writing per project.

Typical report sections and corresponding AI prompts:

• Introduction and scope: "Write a scope of work paragraph for an arc flash hazard analysis. Study covers: 45 buses, voltage range 208V–15kV, facility type industrial manufacturing, performed per IEEE 1584-2018 and NFPA 70E 2024. Study software: ETAP version 22.5. Utility fault current: 42 kA at the 15kV bus."
• Assumptions: "List study assumptions for an IEEE 1584-2018 arc flash analysis: utility fault contribution, motor contribution, X/R ratio, working distances by equipment type (480V MCC = 18 in, 15kV switchgear = 36 in), gap distance by electrode configuration, and arc duration based on protective device operating time."
• Findings summary: After running the Python QA script, prompt: "Summarize arc flash study findings. Total buses: 45. Results breakdown: 2 DANGER, 5 Cat 4, 12 Cat 3, 18 Cat 2, 8 Cat 1. List top 3 highest-IE buses with location and recommended mitigation. Write for inclusion in an engineering report."
• Action item table: "Generate an action item table for arc flash mitigation. Columns: Bus ID, Current IE, Target IE, Recommended Action, Priority (High/Medium), Responsible Party, Due Date. Include 5 sample rows for the highest-IE buses."

Maintaining Arc Flash Study Currency

One of the most overlooked aspects of arc flash compliance is keeping the study current. NFPA 70E recommends reviewing and updating the study when changes occur in the power system — new transformer taps, protective device replacements, utility fault current updates, or new loads that change fault contribution. AI can help track these triggers.

Build a change-log in your project management system and prompt AI monthly: "Review these electrical system change orders for the past month [paste list]. Identify any changes that require updating the arc flash study per NFPA 70E 130.5 — including new equipment additions, protective device replacements, transformer tap changes, and utility fault current revisions. Flag each change as 'Study Update Required' or 'No Impact'."

This systematic approach ensures arc flash labels on equipment remain accurate and that the engineering team is never caught with an outdated study during a safety audit or OSHA inspection.