A gas detector that fails its bump test at seven in the morning, with a confined-space entry scheduled for eight, is a problem you can manage. A detector nobody bump tested, carried into that same entry, is a problem you find out about later - or never. Portable gas detection only works as a system: calibrated, tested, charged, and traceable to the wearer. The hardware is the easy half. This guide covers the record-keeping half - how a small safety team tracks a fleet of monitors without a full-time administrator.
What you will learn
- Bump tests and calibration, briefly
- Why detector fleets drift off the record
- The record each unit needs
- Running a shared fleet on checkouts
- Labels, stations and calibration gas
- Running this in software
- FAQ
Bump tests and calibration, briefly
The two checks get conflated, and the register should keep them apart. A bump test exposes the detector briefly to test gas to confirm the sensors react and the alarms - audible, visual, vibrating - actually fire. It verifies function, not accuracy, and most manufacturers recommend one before each day’s use. Calibration exposes the sensors to a known gas concentration and adjusts them so the displayed readings are accurate; it happens at the interval the manufacturer specifies and produces a record worth keeping formally.
Automated docking stations exist in the market that run both and log the results. Plenty of smaller fleets do it manually with a cylinder, a regulator and discipline - which is exactly where a per-unit record matters most, because no machine is keeping the log for you.
Why detector fleets drift off the record
Detector fleets are usually pooled, and pooled equipment drifts in familiar ways:
- Grab-and-go beats sign-out. Under shift pressure, whoever needs a monitor takes the nearest charged one. Within a month, “unit 7” is whichever unit is in slot 7.
- Sensors expire quietly. A detector can look healthy, charge happily and still carry a sensor past its service life. Without a due date on record, the first warning is a failed calibration - or a failed bump test on a bad morning.
- Units wander to the edges. One lives in a supervisor’s drawer “for visitors”, one rides in a van, one went out with a contractor. The pool shrinks until someone orders more.
- The log and the unit separate. Bump tests get recorded on a sheet by the charging rack; calibration certificates live in email. Reconstructing one unit’s history means a hunt across three places.
A brewery cellar, a machine-shop welding bay and a food plant’s CIP area all share the same failure mode: the detectors work, but nobody can prove which one was where.
The record each unit needs
| Field | Why it matters |
|---|---|
| Unit ID | Big, visible and quotable - “monitor GD-04” beats reading a serial through a rubber boot |
| Make and model | Test gas, sensor spares and calibration procedure are all model-specific |
| Serial number | Ties the unit to manufacturer records, warranty and safety notices |
| Sensor configuration | A four-gas and an O2-only unit look similar racked; the record says which is which |
| Calibration due date | The field that decides whether the unit may be issued today |
| Bump test log | Daily proof of function - the record an incident investigation asks for first |
| Sensor / battery changes | Service life runs per sensor, not per unit; replacements reset the clock |
| Current holder + status | Who has it now, and whether it is in service, in calibration or retired |
Attach calibration certificates to the unit’s record as documents. The point of the register is that any unit’s full history - tests, calibrations, sensor swaps, holders - reads as one audit trail instead of a paper chase.
Running a shared fleet on checkouts
The fix for grab-and-go is to make the sign-out cheaper than the grab:
- Issue at the rack. A unit leaving the charging station gets checked out to a named person - scan, confirm, go. Open-ended assignments suit personally issued units; same-day returns suit pool units.
- Return closes the loop. Back on charge, checked back in, condition noted. Units that fail to come back show on an overdue list while the trail is hours old, not weeks.
- Quarantine failures explicitly. A unit that fails a bump test gets a status change and comes out of the pool - not a sticky note on the rack.
- Count the fleet on a schedule. A periodic physical inventory count of units against the register catches the drawer-dwellers and van-riders. The same count can sweep the rest of the issued safety kit - respirators and other PPE drift in exactly the same way.
The payoff is bigger than tidiness. When an alarm event, near miss or exposure question surfaces, “who carried GD-04 on Tuesday” has a documented answer.
Labels, stations and calibration gas
Label the unit where the label survives and never obstructs: on the back of the housing, away from sensor ports, the display and the alarm openings. Laminated polyester stock handles the knocks; print the unit ID beneath the QR code for gloved-hands moments. Label the charging station slots too, so “back in its slot” and “checked in” stay the same event.
Track calibration gas cylinders as consumables with their own records: gas mixture, expiry date stamped on the cylinder, pressure remaining, and a reorder point. An expired or empty cylinder silently stops your whole testing routine - the fleet is only as current as the gas you test it with.
Tip: rack detectors by ID, charge-side out, and photograph the full rack. Taped to the wall or attached to the station’s record, the photo turns the end-of-shift check into a five-second glance for gaps.
Running this in software
Spreadsheets cope with ten detectors until the first audit question that crosses columns: which units are calibration-current, in service and actually on site today? The sheet knows what was typed, not what happened - and the person typing was never standing at the rack.
AMPthilly keeps the record on the unit. Each detector gets a profile with serial, sensor configuration as custom fields, calibration due dates, attached certificates and a complete history of checkouts, returns and status changes. Scanning the unit’s QR label with a phone camera opens the profile in the browser - no app - so issue and return happen at the rack in seconds, and a failed bump test becomes a service ticket photographed on the spot. Calibration gas fits as a consumable with reorder points on the Starter plan. The free plan covers 3 users and 25 assets, which is a whole small detector fleet - start there and see /features/ for the rest.
FAQ
What is the difference between a bump test and calibration? A bump test confirms sensors and alarms respond; calibration adjusts sensors against a known concentration for accuracy. Frequent and fast versus periodic and formally recorded.
How often should gas detectors be bump tested? Most manufacturers recommend before each day’s use, with calibration at the stated interval. The hard part is proof - log tests against the unit.
How do I track who has which gas detector? Check-out at the rack to a named person, check-in on return. The register then answers “who carried which unit” with dates.
What records should I keep for a gas detector fleet? Per unit: ID, model, serial, sensor configuration, calibration due date, bump test log, sensor changes, holder and status - plus certificates attached.
Does calibration gas expire? Yes - cylinders carry expiry dates and degrade, especially reactive mixtures. Track cylinders as consumables with expiry and a reorder point.
The takeaway
A gas detector fleet is a system of proofs: proof the unit works today, proof its readings can be trusted, proof of who carried it. Separate bump tests from calibration in the record, check units out and in at the rack, quarantine failures by status, and keep the gas cylinders on the register beside the units. None of it is complicated - it just has to happen at the rack, every shift, with the record attached to the unit rather than to a sheet in the office.