Last updated: August 2026
Good ChatGPT prompts for electrical engineers do one thing well: they turn a vague request into a structured draft you can check, while keeping you firmly in control of the engineering. Used that way, ChatGPT is a strong aid for load schedules, sizing logic, control sequences, and report drafting. Used carelessly, it will hand you a confident, plausible, and wrong number, which is dangerous in a field where a mistake can start a fire. This guide gives a prompt library for real electrical tasks, each paired with an honest note on where the model must not be trusted. It builds on our broader guide to how to use ChatGPT for engineering, and for the full toolkit, the hub on the best AI tools for electrical engineering.
How to use these prompts
Every prompt below follows the same shape: give ChatGPT a role, the context and inputs, clear constraints, and an instruction to show its assumptions and reasoning. That structure gets you a checkable draft instead of a black-box answer. The one rule that never bends: ChatGPT does not know your adopted code, cannot run a real study, and must never be the authority for safety-critical or code-critical work such as protection settings, arc-flash, or compliance. Use it to draft and explain, then verify every number against the official source.
What makes a good electrical-engineering prompt
The difference between a useless answer and a useful one is almost always the prompt. Four habits carry most of the weight: name the role you want ChatGPT to play, give it the real inputs rather than a vague description, set constraints that forbid guessing, and ask it to label every assumption and show the arithmetic. When it has to expose its reasoning, its mistakes become visible, which is exactly what you need before you verify an AI engineering answer.
The prompt library
Copy these, replace the bracketed placeholders, and treat every output as a first draft.
1. Load schedule and demand factors
You are a senior electrical design engineer. I am building a load schedule for a [FACILITY TYPE] served at [VOLTAGE, PHASE]. Connected loads: [LIST kW or kVA, quantity, type]. Organize the schedule and explain which demand and diversity factors typically apply to each category and why. Label every assumption, show the formulas before the numbers, and list which values depend on the adopted code and must be confirmed against it.
Watch out: demand and diversity factors are code and jurisdiction specific, so confirm each one against the adopted code and the authority having jurisdiction.
2. Voltage drop and conductor sizing logic
Act as a power distribution engineer and tutor. Feeder: length [M or FT], current [A], conductor [Cu or Al], system [V, PHASE], power factor [PF], target max drop [%]. Walk through the voltage-drop method step by step, show the formula, plug in my numbers, and state the result and whether it meets the target. Keep impedance values as labeled assumptions I must replace with datasheet data, and give me a checklist of what to verify against the official ampacity tables.
Watch out: ampacity, derating, and termination-temperature rules come from code tables, and the model may cite wrong values. Never size a conductor on its number alone.
3. Protection and coordination concepts
You are a protection engineer explaining to a competent colleague. Given this radial setup [source, transformer kVA and impedance, main and downstream devices with ratings], explain the principles of selective coordination. Explain time-current-curve logic conceptually, do not output specific device settings, and list the inputs a real coordination and arc-flash study requires that you cannot substitute for.
Watch out: this is life-safety work. Use ChatGPT only to explain concepts, never to set protection or arc-flash values. Those need a proper study reviewed by a qualified engineer.
4. Control sequence and ladder pseudo-code
You are a controls engineer. Draft a first control sequence for [MACHINE or PROCESS]. Inputs: [SENSORS and switches]. Outputs: [ACTUATORS and motors]. Behavior: [sequence, interlocks, faults, E-stop]. Produce a plain-English sequence of operations, vendor-neutral ladder pseudo-code, and a list of interlocks and fault conditions. Treat safety functions as out of scope here and flag that they need a rated safety system, and add a commissioning checklist.
Watch out: generated logic is a draft that can miss edge cases. E-stop and interlocks must live in rated safety hardware, never in AI-drafted standard logic. Simulate before you trust it.

5. Single-line diagram review checklist
Act as an electrical reviewer. I will describe a single-line diagram in text [sources, transformers, main and feeder breakers, major loads, grounding, metering, protection devices]. Restate the topology to confirm understanding, list ambiguities, and give a review checklist of items commonly missing or mislabeled, such as ratings, interrupting capacity notes, grounding, and device IDs. Do not invent components I did not describe, and mark any inference clearly.
Watch out: the model cannot see the drawing and may infer parts that are not there. Use it as a checklist, and confirm everything against the real single-line.
6. Structured datasheet comparison
You are a component selection engineer. Compare these [COMPONENTS, e.g. two breakers or VFDs]. I will paste the specs for each. Build a side-by-side table of the parameters that matter for [MY APPLICATION], flag missing or ambiguous specs, and list the questions to ask the vendor. Use only the specs I paste; if a value is not provided, write not provided rather than filling it in.
Watch out: asked to recall specs from memory, the model invents numbers. Paste the real datasheet text and forbid it from guessing.
7. Standards orientation, not clause quoting
Act as an engineering mentor. For [TASK], map the standards and code areas that usually apply, what each generally governs, and exactly which official documents and tables I should open for authoritative values. Do not quote specific clause numbers or numeric limits as fact; speak at the level of which standard type applies, and stress that the locally adopted edition and any amendments govern.
Watch out: the model may cite wrong clauses, outdated editions, or the wrong jurisdiction. Use it only to find which document to open, never to quote a compliant value.
8. Troubleshooting hypothesis tree
Act as a troubleshooting partner for a [SYSTEM, e.g. motor drive or distribution board]. Symptom: [DESCRIBE]. Already checked: [LIST]. Readings: [VALUES]. Give a ranked hypothesis tree from most to least likely, and for each the next safe diagnostic test and the reading that would confirm or rule it out. Prioritize de-energized checks, remind me of lockout and PPE, and ask for any missing measurement before concluding.
Watch out: it cannot measure and can anchor on the wrong cause. Follow site safety and lockout procedures, and confirm each step with real readings.
For the writing-heavy tasks, prompt 9 is really a workflow: draft a specification or test report from verified inputs, then tighten the language. A dedicated paraphrasing and clarity tool such as QuillBot is handy for that final polish once the engineering content is fixed, so the numbers stay yours and only the prose gets cleaned up.

Where ChatGPT is not safe for electrical work
The prompts above are deliberately framed to draft and explain, not to decide. Some electrical tasks should never rest on a model’s output, because being wrong is a safety event, not an inconvenience.
- Protection settings, coordination, and arc-flash. These are life-safety calculations that require a proper study and a qualified review, never a chatbot number.
- Code compliance. ChatGPT does not know your adopted edition, so treat any clause or limit it offers as a lead to verify, not a fact, the same way it can get units wrong.
- Safety functions. E-stop, guarding, and interlocks belong in rated safety systems, not in AI-drafted logic.
- Any final number. Recompute anything that will be built, energized, or certified, in line with the broader limitations of AI in engineering and what AI genuinely can and cannot calculate.
Broader surveys suggest engineers are adopting these tools for drafting and explanation while staying wary of raw accuracy: in one 2025 developer survey, trust in AI answer accuracy fell to about 29% even as use kept climbing (Stack Overflow, a developer population rather than electrical engineers), and a civil-engineering body has described the wider sector as slow to adopt AI (ASCE). The same discipline that keeps that trust low is what makes these prompts safe, and it is why our sibling guide of ChatGPT prompts for civil engineers takes the same verify-everything stance.
Frequently asked questions
Can ChatGPT do electrical load calculations accurately?
It can lay out the method and organize a load schedule, but the demand factors and code-driven values it recalls may be wrong or outdated. Use it to structure and explain the calculation, then confirm every factor and result against the adopted code before relying on it.
Is it safe to use ChatGPT for protection settings or arc-flash?
No. Protection coordination and arc-flash are life-safety calculations that require a proper study and a qualified engineer’s review. ChatGPT can explain the concepts, but it must never produce the settings you apply.
Can ChatGPT write PLC or ladder logic I can trust on a real machine?
It can draft a first control sequence and vendor-neutral pseudo-code, which is useful for structure. It is not validated code, it can miss edge cases, and safety functions such as E-stop must live in a rated safety system. Simulate and commission before trusting any of it.
Does ChatGPT know the NEC, IEC, or IEEE codes and the right clause?
Not dependably. It does not know which edition your jurisdiction adopted and can cite the wrong clause or a superseded value. Use it to orient which document applies, then read the authoritative clause from the adopted code itself.
How do I write a good ChatGPT prompt for an electrical problem?
Give it a clear role, the real inputs instead of a vague description, constraints that forbid guessing, and an instruction to label assumptions and show its reasoning. That structure turns a black-box answer into a checkable draft you can verify.
Some links above are affiliate links, meaning we may earn a commission at no extra cost to you if you sign up. We only mention tools we consider genuinely useful, and this does not change our editorial assessments.
Sources
- Developer trust in AI accuracy, 2025 (developers, not electrical engineers): Stack Overflow
- Wider engineering sector slow to adopt AI, December 2025: ASCE
About the author: this guide was written and edited by the CognitiveFuture editorial team, which researches how AI tools fit real professional workflows. We cite primary sources for the claims we make and keep our guidance current. We do not test products ourselves; our assessments synthesize primary reporting and practitioner experience. This is general information, not engineering or safety advice, so verify all work against the adopted code and have it reviewed by a qualified engineer.