When asked how we handle different days (holidays, weekends, etc) I was told my answer was useful.
So, I thought I better write it up. If you agree it's useful, please do feel free to circulate.
Here is how we treat different days treatment-of-occupancy-and-holidays as a two pager...
In short it says:
Different days have different consumption patterns (based on holidays, meter reading failures, weekends, opening hours etc). This will make a nonsense of energy management statistics that do not handle these issues correctly. So a) this is how we do it, and b) if you are considering buying investing or otherwise paying for tools to help manage your buildings, you may want to check these issues.
To make the blindingly obvious yet more obvious, imagine you compare this month with the same month last year to assess performance.
Imagine there are a different ratio of weekends to week days - does this screw up your calculation, how about the Easter holiday shutdown ( a movable feast) or what about a meter reading failure - how can you know your calculations make sense ?
In the simplest case there may be more days (my wife was born on 29th February - so I only have a quarter of the number of birthdays to forget - so far I have a 50% record in eight years - not shabby IMHO:)
Related articles
Reference
So you have smart-meter data, but is it clean?
A household-name client of ours is very big in Energy Performance Contracting, they, like us face a potentially explosive market demand, and they, like us, understand that the fastest way you can get a take on energy saving opportunity in a group of buildings is using pattern recognition. This can make it viable to "get a grip" on thousands of buildings before pricing services or exposure to expensive in-depth expert surveys on site.
There is one major problem -
tl;dr version
http://kwiqly.com/resources/meter-data-for-energy-management.pdf is a non-technical description of the first thing we do with client energy data to clean it up. It might make you think, or raise problems you have not considered. Please feel do free to comment or circulate the link, stick it online somewhere or whatever (if you do please do also link back in case we make revisions) .
There is one major problem -
tl;dr version
http://kwiqly.com/resources/meter-data-for-energy-management.pdf is a non-technical description of the first thing we do with client energy data to clean it up. It might make you think, or raise problems you have not considered. Please feel do free to comment or circulate the link, stick it online somewhere or whatever (if you do please do also link back in case we make revisions) .
Subscribe to:
Posts (Atom)