A thermometer’s accuracy is vital if the temperature readings it produces are to be relied upon. For this reason calibration, or the regular checking of a thermometer’s accuracy, should always be incorporated into any relevant business safety system.
The type of calibration check varies, depending upon factors such as the size of the business and how vital the readings being taken are to the performance of the business.
The Cheap Option
In its simplest form, the calibration of a digital thermometer can be done by measuring known temperatures such as water at melting and boiling points, 0°C and 100°C respectively.
To do this you would place some crushed ice and a little water in a container and insert the thermometer’s probe. Allow time for the instrument to stabilise and the reading the thermometer should give should be 0°C ±0.5°C.
Repeat the process using boiling water and this time a reading of 100°C ±0.5°C should be achieved. It should be born in mind, however, that the temperature of the water will drop quickly and therefore the reading should be taken as soon as possible. Care should be taken to avoid scalding.
A better and more reliable calibration system involves the use of at least two thermometers. One of the instruments should be a reference thermometer, which should be tested annually by a UKAS accredited laboratory to ensure it is reading to accurately.
It is recommended that this instrument be kept specifically for calibration testing and not for daily or regular process testing. Other thermometers can then be checked against the tested instrument and the temperature readings compared. Ideally, this should always be done as soon as possible after the instrument has been certified.
The two or more probes should be placed together (an elastic band can hold the probes together securely) and placed in the water. Again testing can be done in melting ice and hot water so that the instrument’s accuracy is being checked over a temperature range. The probes should be allowed to stand for at least 30 seconds, and then steadily stirred in the water for at least another 30 seconds. The stirring should ensure that the water is the same, uniform temperature throughout, and will improve the speed of response of the readings.
If the two instruments have a claimed accuracy of 0.5°C the temperature readings of both should read the same ±1°C. This tolerance allows each of the instruments to be at the extreme of their accuracy tolerances, i.e. if one is reading 0.5°C high and the other 0.5°C low, there will be an apparent difference of 1.0°C.
Testing accuracy can be improved further using a comparator combined with a certified reference thermometer which has an accuracy of ±0.2°C. Alternatively systems can be checked by immersion in a stirred water bath or even a mini calibration bath and comparing the temperature against another thermometer that you know is accurate i.e. has a current UKAS certificate.
Where an instrument has a socket for a probe, calibrated thermometer test caps are available. These may be used on a regular basis to monitor the performance of the instrument. Test caps are available for Thermocouple, Thermistor and PT100 instruments and in a range of temperatures. The caps are simply inserted into the probe socket in place of the normal probe and the instrument should give a reading equivalent to the test cap’s rating.
Thermometer Checkers, which also act as probes at three different temperatures, and are also available and are used in the same way as test caps. However, as with test caps, the checker will only show the accuracy of the instrument. It will not test the accuracy of the probe and therefore to ensure confidence in a measurement, the combined system of instrument and probe should be calibrated.