The food thermometer: choosing it, where the probe goes, and when to calibrate it

Everything this course has said comes down to a number: four degrees in the fridge, minus eighteen in the freezer, seventy-four in the centre of a piece of chicken, sixty or above in the hot cabinet. Those numbers are the line between food that is safe and food that makes someone ill — and every one of them reaches you through a single intermediary nobody thinks about: the small device that displays the figure. A thermometer that is two degrees out does not merely give you a wrong number; it gives you a reassuring wrong number, and it turns a temperature log from evidence of compliance into a sheet of paper documenting something that never happened. This lesson is about the instrument itself: how to choose it, where to put its probe, and how to know it is still telling the truth.

What the thermometer actually measures#

The only number that matters in chicken is the internal temperature, and the Saudi Ministry of Municipalities and Housing restaurant requirements define it precisely: the temperature at the centre, which is the thickest part of the food item. That definition alone removes half the measurement errors in a working kitchen, because most of them are not faults in the device but in where the probe was put. On a grill the surface runs tens of degrees ahead of the centre, bone conducts heat faster than the meat around it, and the base of a metal tray is hotter than anything sitting on it. A probe touching any of those three gives you a perfectly accurate reading of something you did not want to know.

1. Insert the probe into the thickest point of the piece and push it to at least the middle of that thickness.
2. Stay clear of bone, cartilage and fat: bone heats faster and will read higher than the meat around it.
3. On a whole bird take three readings — the thickest part of the breast, the inner thigh, and the inner wing near the body. The lowest reading is your verdict, not the highest.
4. Never let the probe tip touch the tray, the rack or the wall of the pot.
5. Wait for the display to settle; a number still climbing is not a reading.
6. Check more than one piece from the batch, especially the smallest and the largest — an oven does not cook them equally.
7. Clean and sanitise the probe between pieces, and above all between raw and cooked, or the probe itself becomes the vehicle for cross-contamination.

Four types, each with a job the others cannot do#

The question is not which thermometer is best but which thermometer is for which job. A kitchen that buys one device and uses it for everything ends up with meaningless readings half the time; an organised kitchen keeps three inexpensive devices, each with a fixed place and purpose.

• Fast digital instant-read with a thin probe: the cooking and verification device. It settles in seconds, its slim probe works on thin cuts such as a flattened breast or an escalope, and it sanitises easily. This is the instrument that judges seventy and seventy-four.
• Bimetallic dial: its sensor is not in the tip but spread along part of the stem, so it needs roughly five centimetres of immersion to read at all. Excellent in thick pieces and deep pots, and useless on a two-centimetre cut, where it is really measuring the air and the tray alongside the food.
• Infrared (non-contact): reads surface temperature only. Ideal for sweeping an incoming delivery quickly or checking the surface of a display tray without touching the food, and never acceptable as a verdict on doneness, since a piece of chicken can be hot on the outside and raw in the middle.
• Ambient thermometers and data loggers: these live inside the fridge, the freezer and the hot cabinet, watching the air rather than the food. The restaurant requirements accept either of two arrangements — a digital display fitted to the equipment, or a thermometer placed inside it carrying a label reading "for temperature checking only".

Infrared guns are tempting because they are fast, clean and touch nothing, which is exactly why they are the most misused device in the kitchen. Use one for receiving and for quick sweeps, then confirm anything critical with an internal probe. The rule: if it does not go into the centre of the food, it does not get to decide the food is safe.

What the rules actually require#

On the Saudi side, the restaurant requirements are explicit about when you measure and what you record: frozen food must not arrive above minus eighteen degrees, chilled food must not arrive above four, and a delivery outside the range is refused. High-risk foods are to be temperature-checked on arrival with a non-contact meter or a clean, sanitised probe, and the check is recorded with the documentation kept. Storage limits are set out too: chilled between zero and four degrees, frozen below minus eighteen, and ambient-temperature stores between fifteen and twenty-five.

Note what the document does not say, though: it sets no specific interval for calibrating the instruments. The requirement lands on measuring and recording, while the accuracy of the tool is left to you and to the food-safety system you run. This is where the US Food Code is worth borrowing as a detailed technical reference — many HACCP plans in the region already lean on it — because it states four requirements plainly enough to become written policy in your kitchen.

• Accuracy: food temperature measuring devices scaled in Celsius must be accurate to ±1 °C within their intended range of use (4-203.11).
• Ambient air and water devices: a wider tolerance of ±1.5 °C (4-203.12) — which is precisely why a fridge thermometer cannot rule on whether chicken is cooked.
• Availability: a food thermometer must be provided and readily accessible, together with a small-diameter probe designed for thin foods (4-302.12).
• Calibration: food temperature measuring devices are to be calibrated in accordance with the manufacturer's specifications, as often as necessary to ensure accuracy (4-502.11).

The ice-point method: the one that does not mislead#

The cheapest and simplest accuracy check is a mixture of ice and water, and its answer is known in advance: zero degrees Celsius. Any deviation from zero is your device's error, measured in degrees and written down. More importantly, the melting point of ice is for practical purposes unaffected by altitude — it is the same in Jeddah, Riyadh and Abha — which is what makes it the standard check we recommend here. There is a practical bonus, too: zero sits inside the range that actually matters for chilled chicken, zero to four degrees, so you are testing the instrument in the region where you use it.

1. Fill a large glass to the top with crushed ice, add cold drinking water until the ice is covered, and stir well.
2. Let the mixture stand for a minute or two so it settles at the melting point.
3. Immerse at least five centimetres of the probe stem in the centre of the slush, touching neither the side nor the bottom of the glass.
4. Wait at least thirty seconds, and read the display with the probe still in the water rather than lifting it out.
5. The correct reading is zero. If the device is adjustable, set it to zero — the calibration nut under the dial on a bimetallic model, or the reset button on a digital one.
6. If it is not adjustable, record the offset and apply it to every reading, or replace the device if the error exceeds one degree.
7. Log the check: date, device number, reading before and after adjustment, and the name of whoever did it.

And why the boiling-water method misleads you in Riyadh#

The other well-known method is boiling water, and the familiar rule is that the device should read one hundred degrees. The problem is that the rule is true only at sea level. Water boils lower as elevation rises, and the practical correction published in US extension guidance is to subtract 1 °F — about 0.6 °C — for every 550 feet, roughly 170 metres, above sea level. Apply that to our own cities: Jeddah is effectively at sea level, so water there boils at one hundred; Riyadh sits at around six hundred metres, so it boils close to ninety-eight; and Abha, at roughly 2,270 metres, brings boiling down to about ninety-two.

The practical consequence is both serious and quiet. If a technician in a Riyadh kitchen takes a thermometer that reads ninety-eight in boiling water, decides it is faulty, and adjusts it up to one hundred, he has not repaired the instrument — he has installed a two-degree error in the dangerous direction. Every reading afterwards will sit two degrees above the truth. A fridge actually running at six will display four; a piece of chicken whose centre is at seventy-two will display seventy-four. A device adjusted in perfect good faith now signs off on safety in exactly the two situations this whole course exists to prevent. So: at altitude, use the ice point; and if you insist on boiling water, work out the boiling point for your own city first and never assume one hundred.

Write the correct boiling point for your city on a label and stick it inside the equipment cupboard door, next to the calibration log. A number that is not written down gets forgotten, and a forgotten number gets replaced by an assumed one hundred.

When to calibrate, and who signs#

Calibration is not an annual event but a short daily habit. Common practice in busy kitchens is to check the device at the start of the shift, before the first reading of the day; in lower-volume operations, at least weekly — with additional checks in specific situations that should never be deferred.

• After the device has been dropped, even if it looks undamaged.
• After a large thermal swing — straight from the freezer into a hot pot, for instance.
• After a battery change on a digital model.
• At any moment of doubt: two divergent readings on the same piece, or a number that disagrees with what you can see and smell.
• On receipt of a new device, before it enters service rather than after.

Documentation is what turns the habit into a system. A separate calibration log with its five columns — date, device number, reading before adjustment, reading after, and name — is what gives the rest of your records meaning, because it demonstrates that the figures in the fridge-temperature log came from a tool somebody verified. A temperature log with no calibration log behind it is a set of pleasant numbers that nobody stands behind.

Mistakes that repeat in nearly every kitchen#

• Reading the number before it settles, so what gets logged is a figure still on its way to the truth.
• Using a bimetallic dial on a thin cut, where it measures what is around the food rather than what is inside it.
• Trusting the fridge's built-in display with no independent verification thermometer inside: that display reads one point, and the bottom rear shelf may be two degrees warmer.
• Measuring one piece and judging a whole tray by it.
• Keeping the ambient thermometer in the fridge door — the warmest and most variable spot in the unit.
• Slipping the probe back into an apron pocket without sanitising it after raw chicken.
• Relying on meat colour or clear juices instead of the number, a judgement that is unreliable in both directions.

Nothing in this lesson costs meaningful money: a glass of ice, thirty seconds, and a line in a notebook. In return you get the one thing every other rule of chicken safety rests on — that the number you can see is the number actually inside the food. Instruments do not fail suddenly or announce their error; they drift slowly while continuing to display entirely plausible figures. That short check at the start of the shift is the difference between knowing and assuming.