When you monitor your organization’s performance, have you ever considered which variables, other than good environmental management, might be relevant to consider?
Data normalization, ie comparing connected variables, is a very useful tool to use. It is a means of comparing efficiency of processes, eg by reporting on kWh of energy used per item produced.
However, we often see data normalization that confuses the picture, rather than clarifying it. For example, gas that’s only being used for heating a building being normalized against production throughput.
A simple way to determine the relevance of a variable is to create a scatter diagram using existing data:
If the points on the scatter diagram are in a tidy line, as above, there is a clear correlation between the two factors. The more scattered they are, and the flatter the line, the less strong the correlation – and the less valuable it is to report the data normalized against that factor.
A more complex way to determine relevance – and predict future consumption – is to consider a range of factors by conducting a full regression analysis.
Different variables are likely to be relevant in different situations. For example, electricity consumption per m2 is more likely to be relevant when it’s mainly used for lighting than when it’s used for production.
Normalization of data is not the be-all and end-all of monitoring. After all, we know that we need to make absolute reductions in some areas too, eg carbon emissions. A combination of absolute and normalized data is useful, so that you can assess the overall position as well as your efficiency.
We can help you to generate meaningful data – call Julia on 07904 389889 to discuss your requirements.
