We have calculated the PUs in a lab using a temperature data logger and a spreadsheet, why are they not identical to the Redpost PU Monitor's values?

Q113 - Science

Producing identical values to a Redpost PU Monitor would be an extremely difficult if not impossible task. Here are some of the factors involved:

Alternative formulas being used during calculation
You must make sure that the Cutoff, Z-Value and Base-Value are the same in the calculations. Also the sampling rate of the temperature must be the same and accommodated correctly in the formula. The simplest way to rule out the formula as a cause is to print out a 'full list' from the Redpost RPC-80 unit or Redlink software, this is simply a list of all the readings taken during the run. You can then type these into your spreadsheet by hand and check that the accumulated PUs are the same.

Differences between the two measurement devices
The Redpost PU Monitor is a specialist piece of equipment designed to accurately measure temperature. Your general purpose datalogger will have differences in the accuracy, resolution, calibration and response time, that will lead to deviations in the calculations.

If you are using a thermocouple type datalogger the quoted accuracy might be 0.1% FSD 0.5C in the range 50-70C. This may seem small but it can be interpreted as a maximum error of 1.1C. Due to the exponential nature of PU calculation a 0.1C error in temperature gives a 3.3% error in PU value. So with a 1.1C error this might mean a 36% error in PU before any other sources of error are taken into account. The modern Redpost RPU units are specified to give an absolute worst-case error of 8% PU within a year of calibration.

The response time of the probe is also critical, as if this is not the same as the Redpost probe then this will introduce a disparity, see below regarding sampling intervals.

Variances in the test routine and Sampling Intervals
The beginning and end of the run must be exactly the same as extra PUs will be accumulated during any difference. The timings are critical as say if both instruments record at 10 second intervals, any delay during the start of the run may cause an offset in the timing of maximum 10 seconds. If the temperature is increasing at 1C about every 18 seconds, the two instruments may be offset by 0.5-1C. This may not seem like much but especially towards the end of the run with high temperatures this might mean an error of 5PU per sample and up to 25% over the run.

Conclusion
Overall it would seem that the goal of identical results to the Redpost PU monitor with your datalogger is not achievable given that, due to the nature of PU measurement, tiny temperature errors cause a large error in PU measurement. But by improving the accuracy of your temperature measurement you may get closer towards that goal.

You may find more information on our science pages.

Return to FAQ