There are two traditional ways that organisations monitor their performance. One is to scrutinise trends through time and the other is to compare buildings, processes and vehicles with each other. Now specifically in the case of the fuel used for space heating, it has long been recognised that medium-term changes in the weather distort any trends, and just as importantly, systematic differences in the weather between geographical regions make it impossible to compare a building in one region meaningfully with a building in another even if the two were identical.
The classic fix for this problem is to ‘normalise’ actual annual measured consumption by reference to the weather expressed as annual total degree days. The method consists of first separating consumption into fixed and weather-related portions, with the weather-related part then adjusted in proportion to the ratio between a ‘standard’ number of degree days and the actual measured value. Recombining the fixed consumption with the adjusted value of the weather-related part yields the normalised consumption, that is, the amount the building would have used if it had been exposed to a standard weather year. This figure is more meaningful when compared with the corresponding normalised consumptions of buildings in other regions or with published ‘yardstick’ figures for the class of building in question; and the trend through time of normalised consumption more accurately reflects reality.
I have always been lukewarm about the concept of normalising consumption, not because it is inherently wrong, but because it betokens a very passive view of energy reporting. The weather isn’t just a nuisance factor that distorts your nice clean annual trends: it explains variation in fuel use. Degree-day figures should not just be used to remove that variation. Used correctly, they enable you to calculate expected consumption and are thus fundamental to the detection of hidden unexpected waste which is signalled by a divergence between actual and expected quantities.To focus on normalisation is to miss a fundamental opportunity.