Where the sensor cannot come back out.
A store designed to stay sealed for decades cannot have its thermometers taken away for recalibration. That is the whole problem, and it is the one this programme was funded to work on.
What fails
A platinum resistance thermometer works because the relationship between its resistance and its temperature was established once, by calibration. Put the platinum in a neutron flux and the flux rewrites it. Neutron capture changes the isotopic composition of the metal, so the sensor that comes out of a long exposure is not chemically the sensor that went in, and the calibration it was issued with no longer describes it. Thermocouples suffer the equivalent through transmutation and through thermal ageing of the junction.
The failure is silent, which is what makes it costly. The instrument continues to report a plausible temperature even after the sensor’s calibration has changed. The error may only become apparent later, when another measurement disagrees, if it is detected at all.
Why recalibration is not the answer
In decommissioning, retrieving a probe means putting a person, or a machine and the people who maintain it, into a place chosen specifically because nobody should go there. In waste storage the instrument may be inside a barrier designed to remain intact for longer than the careers of everyone who built it. Recalibration on a schedule assumes the sensor is reachable. In this sector that assumption is the thing that has failed.
What a noise measurement changes
Johnson noise is generated by thermal agitation of charge carriers in any resistor, and the Nyquist relation ties its power to temperature, resistance and bandwidth. Resistance is in that relation, so the obvious objection is that we have simply moved the problem. We have not, because we measure the resistance from the same block of data as the noise, in the published work 6.5536 seconds of it, instead of reading it off a certificate.
Over six and a half seconds a sensor is fixed. Transmutation works over years. So the resistance we put into the calculation is always the one the sensor has now, and we leave the sensor free to change composition, oxidise or degrade without carrying the temperature reading with it. Nothing about it is assumed, specified or certified.
The second thing that matters here is electrical. Reactor halls and decommissioning sites are electromagnetically hostile, and interference has defeated every previous attempt at a practical noise thermometer, because the signal we are measuring is random and small and interference is neither.
- Sensor resistance and Johnson noise are measured from the same block of data, so the resistance is measured rather than assumed
- 6.5536 s
- Radiated RF immunity, tested at an accredited laboratory
- 10 V/m, 80–1000 MHz · 3 V/m, 1.0–2.7 GHz
- Passed the more demanding industrial category
- Traceable to electrical standards, not a chain of temperature comparisons
- 1.380649 × 10⁻²³ J K⁻¹
Where this has got to
The nuclear case is not something we retrofitted onto the research. It is where the funding came from and where the papers point. Innovate UK part-funded our initial phase as a feasibility study under the Development of the Civil Nuclear Supply Chain competition. In 2015 we published with NPL on a noise thermometer built for long-term measurement where sensors cannot be retrieved, and in 2020 the nuclear application was set out with NPL again in EPJ Web of Conferences. The ThermoSI consortium names nuclear decommissioning and waste storage among the industrial cases its high-temperature work package is aimed at.
What we still need is on the sensor side, and we are working on it in the open. ThermoSI work package 2 covers sensor elements usable to about 1,200 °C including their susceptibility to ionising radiation. That is the specific piece between the thermometer we have, which tolerates a hostile environment, and a probe assembly qualified for deployment measured in decades.
In June 2023 we showed the instrument to the sector directly, at the Remote Monitoring of Sensitive Sites demonstration day at Harwell, with NPL and the University of Manchester. It was presented there as a thermometer for harsh environments up to 600 °C, which is the envelope we are working to rather than a measurement taken on the day.
- Presented with NPL and Manchester at Harwell
- 14 June 2023
- Initial phase part-funded under the civil nuclear supply chain competition
- ThermoSI work package 2 is taking practical Johnson noise thermometry to 1,200 °C, with a target uncertainty below 3 °C
- <3 °C at 1,200 °C
If this is your measurement
The thermometer is a working second-generation prototype. Where it goes next will be shaped by real applications. If calibration drift is limiting a measurement you are responsible for, we would like to hear from you.