You Can't Reproduce a Weather-Driven Fault On Site: Decision Tree
Why this matters
You know the fault is weather-driven. The customer is clear it happens in the rain, the freeze, the heat. But the weather is not here now, and you cannot wait a week on the clock for the next front. You have two honest paths: recreate the condition yourself so the fault appears on demand, or instrument the system and let it catch the fault when the real weather returns. Picking the right one, and doing the simulation safely, is what turns a "come back when it is raining" into an answer today. This tree sorts that decision.
Safety first
Simulating weather means creating a condition, and the condition can be a hazard. Do not create one you cannot control. Never introduce water near energized equipment without de-energizing and verifying dead first. Never force heat, pressure, or load past a rating to chase a fault, and never defeat a safety or interlock to reproduce one. If the only way to recreate the condition is unsafe, that is your answer: stop simulating and switch to instrument-and-wait.
Start here: can you safely simulate the trigger?
Name the specific condition, then ask whether you can produce a controlled, safe version of it on site right now. That single question routes the whole job.
If you can simulate it
Recreate the condition in a controlled way and watch for the fault:
- Rain or wet: apply water deliberately to the suspected intrusion path, low volume, from the outside working in, so you find the entry point. De-energize anything the water could reach first.
- Freeze: chill the suspected spot with a controlled cold source rather than waiting for winter, and watch for the bind, leak, or dropout as it cools.
- Heat or sun: apply controlled heat to a suspected component or run the system hard to build its own heat load, and measure as the temperature climbs toward the limit.
- Humidity: raise local moisture around a suspected enclosure and watch for a connection to bridge or a control to act up.
- Wind or pressure: reproduce the pressure effect with a fan or by changing the building pressure, and re-check draft, drainage, or the mount that moves.
The rule for any simulation: change one condition at a time, hold it steady, and measure the system while it is under the condition. A fault that appears when you apply the condition and clears when you remove it, repeatably, is a caught fault, not a guess.
If you cannot simulate it
When the condition is unsafe to create, too large to reproduce, or you tried and the fault would not come, switch to capture and wait for real weather:
- Leave a recorder on the values most likely to move when it faults, so the event is logged with a timestamp you can line up against the weather.
- Set a customer-triggered capture: show them one simple thing to note or photograph the moment it happens, including the weather at that time.
- Set a defined recheck window tied to the forecast, so you return when the condition is actually present rather than hoping.
Simulate or wait: how to choose
| Factor | Lean toward simulating | Lean toward instrument-and-wait |
|---|---|---|
| Can you produce the condition safely? | Yes, controlled and reversible | No, or only by creating a hazard |
| How localized is the trigger? | A specific spot you can target | Whole-system or building-scale |
| How soon is the real condition due? | Not soon, or unpredictable | A front is forecast shortly |
| Confidence in the suspected trigger | High, you know what to apply | Low, you are still confirming the condition |
When both are viable, simulate first, because it gives you the fault under your control today. Fall back to waiting when simulation is unsafe or came up empty.
Recap
- Never create a condition you cannot control; if the only simulation is unsafe, switch to waiting.
- Ask first whether you can safely produce the trigger on site.
- If yes, recreate it one variable at a time and measure under the condition.
- If no, instrument for capture or set a forecast-tied recheck, and let the real weather bring the fault.
References
- Trade-standard practice for controlled fault reproduction and data logging
- Manufacturer operating-condition and safe-test specifications
- See related: The Fault Only Happens in Certain Weather (decision tree); How to Test Under the Conditions the Fault Actually Happens In