Technical reference
Two methods suit different instruments. A probe can use the point method, at the and . A sensor that cannot be immersed is compared with a reference instrument placed beside it.
A ±1 °C error matters most where meet growth. At the cold end, FDA's cumulative guidance falls from 7 days up to 5 °C to 1 day at 6 to 10 °C (FDA Table A-2), so a probe reading low can hide a real change in risk.
Parameters
| Reference points | 0 °C ; 100 °C at sea level |
|---|---|
| Tolerance in this example | ±1 °C; your law or plan sets yours |
| Reading resolution | 0.1 °C |
| Comparison method | Let both settle (10 to 30 minutes), then compare; re-check two or three times |
| Failed instrument | Out of use and labelled at once; readings since the last pass are reviewed |
Records that prove it
- Instrument id
- Date and operator
- Reference point or reference instrument
- Reading and signed difference
- Result and action taken
- Decision on readings since the last pass
Reference temperatures
Values from , a code for mass catering. Your own law or plan may set different ones.
CCP or prerequisite?
That depends on the for your process. The number, if it has one, belongs to your plan.
Law and practice
| Control | Australia | New Zealand | Singapore | Malaysia | In practice |
|---|---|---|---|---|---|
| Thermometer accuracy | Accurate to ±1 °C, readily accessible (3.2.2 cl 22). | to schedule (template plan). | A working, maintained thermometer in every chiller, freezer and cold room (reg 12); no accuracy figure read. | ±1 °C (reg 25). | ±1 °C checked at ice and is close to universal. Frequency runs from twice a month to quarterly, and some add a yearly external . At least one plan raises calibration to an OPRP. |
Law as read in the authority library between 14 October and 4 November 2026. In practice: food safety programmes read across hotels, clubs, venues, healthcare, manufacturing and airline catering, with no business named. A common practice isn't proof that a is safe or lawful where you are.