...
Once the regression (i.e. expected performance) has been determined, it can then be extended to show a CuSUM representation of the data. This is effectively a cumulative sum of the difference between the expected performance and the actual performance (e.g. 4000 kWh was used, but based on the performance regression line 3500 was expected. This leads to a difference of 500 kWh). This will highlight the step change in performance over time and highlight periods where there is significant performance degradation (i.e. line on the graph goes up), or improvement (i.e. line on the graph goes down). This visual indication trend over time may otherwise be hidden in a large scatter of the data. A The CuSUM technique is a simple but remarkably powerful statistical method, which highlights small differences in energy efficiency performances over time. A typical CuSUM graph follows a trend and shows the random fluctuation of energy consumption and should oscillate around zero (standard or expected consumption). This trend will continue until something happens to alter the pattern of consumption such as the effect of an energy saving measure or, conversely, a worsening in energy efficiency (poor control, housekeeping or maintenance). When looking at CuSUM chart, the changes in direction of the line indicate events that have relevance to the energy consumption pattern.
...