Running calculation is an extension of the running average methodology. In order to follow a building, I have been using it to check if consumptions were ok. I also used this method in other contexts and found it very efficient.
It is easy, often well known, but not so much used. I wanted to write this document to show how much information can be found in a dataset if data are transformed using that simple method.
It is very helpful with data that don't follow physical rules, like the water consumption in a building (the aim being for example to find leaks).