What is selective coordination in electrical protection and when is it required by the NEC?

Selective coordination means that when a fault occurs, only the protective device immediately upstream of the fault operates — all other upstream devices remain closed, keeping the rest of the system energized. The NEC mandates selective coordination for emergency systems (NEC 700.32), legally required standby systems (NEC 701.27), and critical operations power systems (NEC 708.54). It also applies to healthcare facilities under NEC 517.26. Selective coordination must be demonstrated over the entire range of fault currents from the minimum to the maximum available, which is a more stringent standard than simple non-simultaneous tripping.

How do I prompt AI to interpret ETAP or SKM protection coordination results?

The most effective approach is to export your bus fault summary and device duty tables from ETAP or SKM as CSV, then paste the data into your AI prompt with a structured request. Include: the system voltage, the number of buses, the pass/fail threshold (device interrupting rating vs available fault current), and the desired output format (executive summary, action item table, or report paragraph). Always ask AI to cite the relevant NEC article — for example, NEC 110.10 for equipment interrupting ratings — in its output so the language is immediately suitable for your engineering report.

What is the IEC 60255 standard inverse curve equation and how is it used in relay coordination?

The IEC 60255-151 standard inverse (SI) overcurrent relay curve is defined by: t = (TMS × 0.14) / ((I/Is)^0.02 − 1), where t is the trip time in seconds, TMS is the time multiplier setting (0.05–1.0), I is the fault current, and Is is the pickup current setting. This curve produces longer trip times at lower fault currents and shorter times at higher currents. Very inverse (VI) and extremely inverse (EI) curves use different constants (K and α) and are selected when faster operation at high fault currents is required. AI can compute operating times for any fault current, helping engineers verify coordination margins between upstream and downstream relays.

What relay setting QA checks should be automated with Python?

The core QA checks that benefit from Python automation are: verifying that phase overcurrent (51) pickup settings are 1.25–1.5× full-load current; confirming that instantaneous (50) settings are at least 1.2× maximum load current and below 80% of available fault current; checking that time dial settings match the study recommendations within ±20%; verifying that ground fault (51G) pickup is set at 50% or less of the phase pickup; and confirming coordination margin between upstream/downstream devices is at least 0.3 seconds for electromechanical relays. A Python script that reads the relay export CSV and flags any item outside these ranges can be run in seconds and generates a report suitable for the project QA file.

What is the coordination margin requirement between protective devices?

The minimum coordination time interval (CTI) between protective devices depends on device technology. For electromechanical relays, IEEE 242 recommends a minimum of 0.3–0.4 seconds to account for relay overtravel, breaker interrupting time, CT accuracy, and relay timing tolerances. For modern digital relays, the minimum CTI can be reduced to 0.1–0.2 seconds due to better timing accuracy and no overtravel. For current-limiting fuses coordinating with downstream breakers, the coordination can be nearly instantaneous because the fuse clears before the breaker can operate. Always verify the CTI at the maximum available fault current, not just at intermediate levels.

← AI Engineering Studio

AI & Automation·9 min read·June 19, 2026

🔌 AI for Electrical Protection Coordination: Automating Relay Settings and Selectivity Studies

Discover how AI tools accelerate protection coordination studies — from interpreting ETAP and SKM outputs to automating relay setting QA, generating time-current curve reports, and prompting AI for IEC 60255 and IEEE 242 compliant protection calculations.

Why Protection Coordination Is Ripe for AI Assistance

Protection coordination is the discipline of ensuring that when a fault occurs in a power system, the protective device closest to the fault clears it first — and only that device. Getting this right requires iterating through time-current curves (TCCs), comparing relay pickup settings against load currents and available fault currents, and documenting every device pair with coordination margins. On a medium-sized industrial project, this can mean dozens of breaker-relay pairs and hours of manual analysis in ETAP or SKM Power*Tools.

AI assistance compresses the most time-consuming parts of this work: interpreting software outputs, grouping failures by type, drafting report narratives, and performing QA checks on relay setting exports. It does not replace the engineer's judgment on TCC interpretation or device selection, but it dramatically accelerates the documentation and QA layers that consume a disproportionate share of coordination study time.

This article covers how to integrate AI into a short-circuit and protection coordination workflow, with real prompt templates, Python automation patterns, and the key standards references that must underpin every deliverable.

Short-Circuit Study Inputs and AI-Assisted Triage

A protection coordination study begins with short-circuit calculations. Using ETAP or SKM Power*Tools, engineers calculate three-phase (3Ø) and single-line-to-ground (1Ø-G) fault duties at each bus under worst-case scenarios — typically maximum utility contribution for device interrupting duty and minimum contribution for relay pickup coordination. The software exports bus fault currents, device ratings, and optionally TCC plots.

Once you have the export, AI accelerates the triage step. Paste the bus-by-bus results table into your AI assistant and prompt: "Group all buses where calculated fault current (Isc) exceeds the device interrupting rating. Identify device type, voltage class, and percent margin. List by severity — most exceeded first — and suggest corrective actions: rating increase, current-limiting fuse, or upstream impedance addition."

A typical output summary might be: "Bus P1-BUS at 208V: Isc = 14.5 kA, Panel P1 rating = 22 kA. Margin = 7.5 kA (34%). PASS. Bus SWGR-A at 480V: Isc = 42 kA, Main Breaker rating = 65 kA. PASS. Bus MCC-3 at 480V: Isc = 38 kA, motor control center rating = 30 kA. FAIL — upgrade to 42 kAIC bus or add current-limiting fuses upstream."

This triage output feeds directly into your action item log and report draft. AI also helps draft the utility fault assumption paragraph: "Summarize these utility fault contribution assumptions for a protection study report: available fault current 42 kA at the 480V bus, X/R ratio 15, and both maximum and minimum contribution scenarios considered. Include a note about motor contribution."

Relay Setting Calculations and IEC/IEEE Curve Equations

Relay pickup settings for overcurrent elements follow specific mathematical curves defined by standards. The two dominant frameworks are:

IEC 60255-151 defines standard inverse (SI), very inverse (VI), and extremely inverse (EI) curve equations: t = (TMS × K) / ((I/Is)^α − 1) where TMS is the time multiplier setting, Is is the pickup current, and K and α are curve-specific constants. For the standard inverse curve, K = 0.14, α = 0.02.
IEEE C37.112 defines the moderately inverse, very inverse, and extremely inverse curves used by North American relay manufacturers. These use a different equation form: t = TD × [A / ((M^p) − 1) + B] where TD is the time dial, M is the multiple of pickup, and A, B, p are curve constants.

AI can implement either curve family in Python. Prompt: "Write a Python function that computes operating time for an IEC 60255 Standard Inverse overcurrent relay. Inputs: TMS (time multiplier setting), Is (pickup current in amps), I (fault current in amps). Output: trip time in seconds. Include a check for I > Is."

def iec_standard_inverse(tms, Is, I):
    """IEC 60255-151 Standard Inverse curve."""
    if I <= Is:
        return float('inf')  # relay does not operate
    M = I / Is
    K, alpha = 0.14, 0.02
    return tms * K / ((M ** alpha) - 1)

# Example: TMS=0.3, pickup=120A, fault=600A
t = iec_standard_inverse(0.3, 120, 600)
print(f"Trip time: {t:.3f} s")  # → approximately 0.54 s

With this function, you can automate coordination margin checks. For each upstream/downstream device pair, calculate operating times at the fault current and verify the minimum coordination margin — typically 0.3 seconds for electromechanical relays and 0.1–0.2 seconds for digital relays. If the gap is below the threshold, AI can suggest time dial adjustments.

AI-Assisted ETAP and SKM Output Interpretation

ETAP and SKM Power*Tools generate voluminous outputs: bus fault summary tables, device duty reports, TCC plot files, and protective device coordination logs. Interpreting these and translating them into client-ready report language is where AI saves the most time in a coordination study.

The key workflow is to export results to CSV or paste tabular data into the AI context window, then prompt for structured interpretation. Example prompts that generate high-quality, report-ready output:

"Here is my ETAP bus fault summary [paste table]. Write a one-paragraph executive summary for an engineering report. State: number of buses studied, maximum available fault current, lowest device margin, number of PASS/FAIL items, and overall compliance statement per NEC 110.10."
"Identify all coordination pairs in this TCC log where the upstream and downstream device operating times overlap at any fault current below the maximum available. List the pair IDs, the fault current at which overlap occurs, and recommended remediation."
"Draft meeting notes for a protection coordination review meeting. Open items: Bus MCC-3 interrupting rating failure, Relay R2 time dial mismatch. Format as: Item, Owner, Due Date, Priority."

For relay settings specifically, the AI assists when you export relay setting files from SEL AcSELerator, ETAP, or manufacturer configuration software. The QA workflow compares exported settings against the study recommendations line by line. A Python automation script can load the export CSV and flag discrepancies:

import pandas as pd

df = pd.read_csv('relay_settings_export.csv')

def qa_check(row):
    notes = []
    if row['Function'] == '51':  # Phase overcurrent
        if row['Pickup_Exported'] < 1.2 * row['Pickup_Study']:
            notes.append('Pickup below study recommendation')
        if abs(row['TD_Exported'] - row['TD_Study']) > 0.2 * row['TD_Study']:
            notes.append('Time dial mismatch > 20%')
    if row['Function'] == '50':  # Instantaneous
        if row['Inst_Setting'] < 1.2 * row['Load_Current']:
            notes.append('Inst may trip on load inrush')
    return 'Pass' if not notes else 'Check: ' + ', '.join(notes)

df['QA_Result'] = df.apply(qa_check, axis=1)
df.to_excel('Relay_QA_Report.xlsx', index=False)

Coordination Report Generation with AI

The final step in a protection coordination study is the deliverable: a report that includes study assumptions, short-circuit summary tables, TCC screenshots with device identification, coordination analysis for critical pairs, and an action item list. AI can draft every narrative section of this report from structured data.

For each TCC plot pair, prompt: "Describe the coordination between a 1200A main breaker (LSIG trip unit, Long = 0.5 pu, Short = 8× at 0.1 s) and a 225A downstream panelboard circuit breaker (Ir = 1.0, tr = 12×). State whether they coordinate for fault currents from 1× to 20× rated current, identify the coordination range, and flag any overlap."

The QA checklist for an AI-assisted coordination report should verify: all buses have an interrupting duty check; all critical upstream/downstream pairs are in the TCC log with a coordination gap or overlap note; utility fault assumptions are documented; motor contribution is acknowledged; and the report includes a compliance statement referencing NEC 240, NEC 700/701/708 for emergency systems (which require selective coordination), IEEE 242 (Buff Book), and applicable manufacturer data.

For projects with emergency or legally required standby systems, NEC 700.32 and 701.27 mandate selective coordination — meaning the coordination margin must be demonstrated over the full fault current range, not just at the available short-circuit current. AI can generate the compliance language and flag any device pairs where this cannot be demonstrated.

Building a Protection Coordination Knowledge Base with AI

Over time, the most valuable AI application in protection coordination is building a project knowledge base: a library of past solutions, device swap decisions, and curve setting combinations that can be queried for new projects. A Retrieval-Augmented Generation (RAG) system trained on past coordination study reports can answer questions like: "What time dial setting did we use for the 51-element on the 480V feeder in the pharmaceutical facility project?" or "Which breaker did we upgrade when the available fault current exceeded the panel rating on a 208V system?"

Combined with AI-generated QA automation, this makes protection coordination a discipline where repetitive verification work is handled by code, and the engineer focuses on the judgment-intensive steps: TCC interpretation, margin decisions for borderline cases, and selectivity tradeoffs in complex radial or looped systems.

Topics covered

protection coordination AIrelay settings automationETAP AISKM Power*Tools AItime-current curvesshort circuit study AIselective coordination NECIEC 60255 relay curvesIEEE 242 protectionovercurrent relay settingsAI electrical protectionrelay coordination Pythonarc flash coordinationOCPD selectivityprotective relay QAelectrical power system AINEC 700 selective coordinationpower systems engineering AI

🛠️ Related Free Tools

Put this knowledge to work on your iPhone

Browse our full catalog of professional iOS apps — from electrical code tools to AI builders.

Browse All 95+ Apps