The errors in resistance thermometry fall naturally into four main groups, all of the errors it is relatively easy to identify the causes and to separate the various functional elements in the construction
Immersion errors:
For resistance thermometry the sensing elements are large.
Ice points can be difficult, particularly with stainless steel sheathed PRTs; the combination of the thermally conductive sheath and the self-heating can make it difficult to realise the ice point to better than ±0.01 °C.
It is important that the ice is very fine and well packed. For measurements requiring accuracies better than ±0.01°C a water triple point should be used in preference.
Lag and settling errors:
The errors due to the response time of the meter. Time constants for PRTs vary considerably: 0.2 seconds for film types; 2 to 6 seconds for larger fully and partially supported types; 5 seconds for SPRTs; and 5 to 20 seconds for stainless steel sheathed assemblies
Radiation errors:
The long-standing resistance thermometer sheath provides a transparent ‘ light pipe ‘ along which radiation can carry heat from and to the sensing element.
Because of the radiation the PRT is not only in thermal contact with the medium immediately surrounding the platinum element but is also in radiative contact with whatever it ‘sees’ down the sheath.
Self-heating:
Because a current is passed through the sensing element to measure its resistance, the element dissipates heat, which in turn causes the temperature of the element to increase.
This self-heating error is very simply modelled as the power dissipated divided by the dissipation constant, h. The error in the temperature measurement is
where R(t) is the resistance of the sensing element and I is the sensing current. The dissipation constant h is normally expressed in milliwatts per degree Celsius.
