A practical guide to integrating AI tools into electrical engineering workflows — from NEC code compliance calculations to protection coordination, arc flash analysis, PV system sizing, and construction phase document management.
AI tools like ChatGPT, Claude, and Copilot are transforming how electrical engineers handle repetitive calculations, code lookups, specification writing, and RFI responses. Engineers using AI report saving 2–4 hours per day on documentation tasks while reducing code-compliance errors. This guide maps every major EE workflow to a specific AI use case and prompt pattern.
NEC Code Calculations: AI-assisted load calculations, conductor sizing, and panel schedule generation with NEC 2023 references. Protection Coordination: relay setting rationale generation and protective device selection documentation. Arc Flash Analysis: PPE category documentation and warning label content. PV System Sizing: solar array calculations and interconnection narrative writing. RFI Management: drafting field RFI responses from specification language and drawing references.
Engineering prompts require specificity that generic prompts lack. The guide covers the CREED framework: Context (project type and voltage level), Role (EE with relevant credentials), Example (reference calculation), Exact output (table, NEC citation, or paragraph), and Disclaimer (verify with authority). This structure produces engineering-grade outputs that can be used as starting drafts, not just brainstorming.
AI cannot replace licensed engineer judgment, stamp calculations, or verify against current adopted code editions (which vary by jurisdiction). Treat AI output as a first-draft tool: verify every NEC section reference, check conductor ampacity tables directly, and confirm AHJ requirements. AI is strongest for writing (narratives, specs, submittals) and weakest for numerical accuracy in complex calculations.
AI can generate the structure and logic of NEC load calculations — demand factors, feeder sizing steps, panel schedule formats — but you must verify every NEC section reference and arithmetic result. AI models are trained on code text and engineering documents, so they understand the methodology, but they can cite wrong section numbers or make arithmetic errors in complex calculations. Best practice: use AI to generate the calculation template and narrative, then verify all values against current NEC tables and your local adopted code edition before stamping.
AI is most useful in protection coordination for generating the written basis-of-design narrative, relay setting justification documentation, and protective device selection rationale — the text portions that typically take hours to write. For the actual coordination curves and time-current characteristic analysis, use dedicated software (SKM Power*Tools, ETAP, EasyPower). Once the study is complete, AI can draft the report sections describing why each overcurrent device was selected and how the coordination intervals were achieved.
ChatGPT-4o and Claude 3.5 Sonnet are the most capable general-purpose AI tools for EE tasks — both can handle complex code lookups and document drafting. For code-specific work, tools trained on NEC and NFPA 70E (like some specialty engineering AI assistants) can be more reliable for citations. GitHub Copilot helps engineers who write Python scripts for load flow analysis or data processing. Microsoft Copilot in Word/Excel is useful for specification writing and panel schedule formatting directly in the documents engineers already use.
Arc flash documentation using AI follows a two-step process: (1) Run the arc flash study in ETAP, SKM, or similar software to get the incident energy values, working distances, and PPE categories for each equipment location. (2) Use AI to generate the warning label content, PPE selection rationale, and Section 130.5 risk assessment documentation using the study output values as input. Prompt example: "Generate NFPA 70E-compliant arc flash warning label text and PPE category justification for a 480V MCC with 12.3 cal/cm² incident energy at 18 inches working distance." AI handles the writing; the study software handles the engineering.
Yes — solar PV interconnection narratives and utility application narratives are excellent AI use cases because they follow a standard structure (system description, inverter specs, interconnection method, protection scheme, metering) that AI can template and fill in. Provide AI with the system parameters: DC array size, inverter model and count, point of common coupling voltage, metering configuration, and applicable IEEE 1547 or utility-specific interconnection requirements. AI generates a draft narrative in minutes that typically needs only minor technical verification before submission.