How to Wait Out a Relearn Without Wasting a Visit
Why this matters
A relearn is dead time that is not wasted time, and most shops handle it as if those were the same thing. The tech either stands in a mechanical room watching a valve find its stop, burning the back half of the day, or leaves at minute four, gets a callback that evening, and pays for a second visit to observe something that would have finished itself. Both outcomes come from the same gap: no record of what is being waited on, so nobody can decide whether waiting is cheaper than returning.
The fix is not a longer wait or a shorter one. It is a small written artifact, built field by field, that turns "let us give it a few minutes" into a decision anyone in the shop can act on, including the customer.
Before you settle in to watch anything
A relearn is a device deliberately driving itself to a hard stop. It will move without warning, and it may move several times. Before your hands go near an actuator, damper, valve, door, belt or any rotating load, isolate the energy source, lock and tag it, and block or bleed the stored energy, which is the duty at 29 CFR 1910.147. If you need readings at a panel or on a branch circuit while you wait, de-energize and lock or tag before contact under 29 CFR 1910.333(b)(2), and prove dead using the live-dead-live sequence in NFPA 70E-2021, 120.5. Observing a relearn is a watch-only activity. If you cannot get the information without a hand inside, the system comes down first.
Build the record, one field at a time
Everything below goes on the ticket as you determine it, in this order. Each field exists because the field before it is useless without it.
Field 1: The interruption, in three facts
Duration, count, and character. Forty minutes out, three interruptions counting the two fast reclosures, clean loss rather than a sag. You get these from the customer and from whatever the site logged, not from the equipment. Without them you cannot judge whether the delay you are about to accept is plausible, and a wait you cannot justify is a wait you will abandon at the worst moment.
Field 2: The clock start
The moment supply became continuously present. Not the first flicker, not your arrival. If supply dropped again at any point, the clock start moves to the last restoration. If anyone reset anything, including you, the clock start moves to that reset. This single field is where most correctly-executed waits go wrong, because the tech measures from the customer's story rather than from the last real event.
Field 3: The published delays, with the source named
Write down what the service literature actually says and where you read it: hold-off interval, stage count and inter-stage interval, whether a position relearn is documented, whether the control adapts across cycles. "The literature is silent" is a legitimate entry and it changes the next field.
Field 4: The window, computed and written as a clock time
Two times the longest published delay, floor of 15 minutes, counted from Field 2. Write the expiry as an actual time of day, not as a duration. A duration invites re-estimation every few minutes. A time of day on the ticket is a fact the customer can read over your shoulder, and it ends the negotiation.
Field 5: The exclusion check, initialed
Before you accept any wait, confirm none of these is present: a protective device has operated, a hard lockout code is displayed, the same failure has repeated on successive attempts, a physical measurement is wrong on an attempt, or anything is discolored, hot, wet or burnt. Any one of these ends the wait immediately and converts the visit to a diagnosis. This field is what keeps the method from becoming an excuse to stall on a real fault.
Field 6: The hold decision, from a stated rule
State the rule and apply it, do not improvise it. A workable default, per call: stay on site when the remaining window is 20 minutes or less AND your next commitment has no hard start time; release the call when either condition fails. The Boolean is AND, deliberately, because a 10-minute wait still costs you a missed appointment if the next customer is holding a fixed slot.
The economics behind the number are worth carrying in your head. A return visit costs the round-trip windshield time plus a fresh diagnostic setup. If a typical return in your service area runs about 40 minutes of drive alone, a 15-minute on-site hold costs roughly 0.4 of that drive time, so returning is on the order of 2.7 times more expensive in travel before anyone has looked at anything. Be careful what that ratio proves, though: on drive time alone the wait only stops paying at about 40 minutes, not 20. The 20-minute ceiling is set by the second cost, which is that the tech standing there is billable somewhere else. Treat 20 as a default tuned to how full your day is, and raise it on a day with a hole in the schedule.
Field 7: The observation script, if you release
If Field 6 says release, the customer becomes your instrument, so give them something a non-tech can actually do. Three items, no more:
- What "finished" looks like, in plain words: air coming from the vents, water running hot at the tap, the pump noise settling to steady rather than cycling.
- The time it should happen by, copied from Field 4.
- Exactly one thing that means call immediately rather than wait: a breaker trips, something smells burnt, water appears, or the equipment tries and quits repeatedly.
Do not ask them to read a code, count a sequence, or judge a sound level. Anything requiring interpretation comes back as a wrong report, and a wrong report is worse than no report because you will plan the return around it.
Field 8: The disposition rule, written before you leave
Both branches, on the ticket, so dispatch does not have to call you. If it completes by the expiry time, the ticket closes as a recovery hold with no repair. If it has not completed by the expiry plus a 10-minute grace, it is a confirmed fault and dispatch books the return with the diagnostic time already allocated rather than as another short look. Setting this before you leave is what stops a second visit from being another 15 minutes of standing around.
The filled-in record: a heating system after a long outage
Here is the artifact completed on a real-shaped call, so you can see what a good one looks like.
Field 1. Power out about 40 minutes. Three interruptions counting two fast flickers before the sustained loss. Clean loss, no reported dimming or flickering lights while power was present.
Field 2. Continuous restoration at 2:10 pm, per the customer's phone timestamp on the message she sent her manager. No further interruptions. No resets performed by anyone, confirmed by asking directly rather than by asking whether they "did anything."
Field 3. Service literature documents a mixing-device position relearn on power-up, with full travel of the device given as the driver, and no fixed interval published. No staging documented. No hold-off published.
Field 4. Literature is silent on interval, so the floor governs. Window is 15 minutes from 2:10 pm, expiring at 2:25 pm, written on the ticket.
Field 5. Exclusion check at 2:16 pm: no breaker tripped, no lockout code, one attempt in progress rather than repeated attempts, nothing hot, wet, discolored or burnt. Initialed. The device is audibly driving and stalling briefly at travel limits, which matches the documented relearn rather than contradicting it.
Field 6. Time is 2:16 pm. Remaining window is 9 minutes, which is at or below 20. Next commitment is a flexible afternoon slot with no fixed start. Both conditions pass, so the decision is stay.
Field 7. Not used. The tech stayed.
Field 8. At 2:23 pm, inside the window, the device settles at a controlled position and the system produces heat at the terminal units. Ticket closes as a recovery hold, no parts, with the interruption timeline and the window on the record.
Total cost of the correct call: 7 minutes on site from the exclusion check at 2:16 pm to completion at 2:23 pm, and no return. Change one field and it inverts. Had the next commitment been a fixed-time appointment, Field 6 fails the AND, the tech releases at 2:16 pm with the script from Field 7, and the customer confirms heat by phone before 2:35 pm. Same outcome, no return, and the fixed appointment is not blown. Had the exclusion check at 2:16 pm found a tripped protective device instead, Field 5 ends the wait on the spot and the visit becomes a diagnosis regardless of how much window is left.
The failure mode this record prevents is specific. Without Field 2 written down, a tech who arrives at 2:16 pm typically dates the clock from the customer's "the power went out around 1:30," concludes 46 minutes have passed, decides the window is long gone, and condemns a mixing device that was 7 minutes from finishing its documented routine. The arithmetic is fine. The start time is wrong, and no amount of careful measurement downstream recovers from that.
What the record does after you leave
The hold record is worth more to the shop than to the tech who wrote it, and only if it survives the visit. Two uses.
Dispatch reads Field 8 rather than calling the tech. A return booked off "did not complete by 2:35 pm" is scheduled as a full diagnostic, with the interruption timeline already attached, so the second tech starts where the first one stopped instead of re-interviewing the customer.
The office reads Field 3 across many tickets. A shop that logs published delays over a season learns which equipment on its route genuinely takes long enough to matter, and that is what lets a dispatcher tell a caller "give it 15 minutes and call me back" on the phone, before a truck rolls at all. The visit you never make is the cheapest one on the board, and that phone script is only trustworthy if it came out of records rather than out of somebody's optimism.
References
- 29 CFR 1910.147, OSHA control of hazardous energy, for isolating and blocking stored energy on equipment that may move during a relearn
- 29 CFR 1910.333(b)(2), OSHA electrical safe work practices, for de-energizing and locking or tagging before work on electric circuit parts
- NFPA 70E-2021, 120.5, the live-dead-live process for establishing an electrically safe work condition
- Manufacturer service literature for documented relearn routines, staging intervals and restart delays
- See related: What a System Does While It Is Recovering From an Outage; How to Tell a Reset State From a Real Fault After a Power Interruption