Predictions are only as accurate as the data that goes into the calculation. At measurable.energy we use our own energy model that includes data from all the different appliances we have plugged in and the savings they generate. On top of this layer, we add information from the actual building the system will be installed in, including the different appliance types and quantities. The model is then tweaked to represent the type of installation and the different areas; offices, meeting rooms, heating/cooling, kitchens etc. The prediction changes depending on the predominant mix of appliances. Lower energy use appliances will give a longer payback, while high energy appliances give quicker paybacks. With this approach, we are able to offer a realistic prediction of the energy savings and in many cases exceed the prediction.
The part that is out of our control is the energy behaviour of the occupants. Our predictions factor in that many of the appliances are being left on/standby. This is the most common pattern we see. However, there are examples where the savings have been lower than expected. In these cases our customer success team works to optimise that installation by ensuring higher energy-wasting appliances are plugged in, sockets are relocated or additional periods of inactivity are captured.