In this post we take a look at the data we’ll be using and publishing as part of the Energy Sparks project.
Energy Sparks will help schools become more energy efficient by improving access to their energy usage data. In order to present and analyse data on energy consumption we also need to use a variety of additional data sources.
The energy usage data alone won’t give us a clear picture because a school may also generate some of its own power. Similarly, variations in weather conditions might change the amount of energy being used. It’s reasonable to consume more energy when its colder outside, for example.
We’ve identified the following list of datasets that we will ideally need for each school:
- Gas and electricity consumption data
- Electricity generated from solar panels installed at the school
- Calendar information that describes the daily schedule for each school and term times throughout the year
- Weather observations that include the outdoor temperature
- Information about the school itself, e.g. address, floorspace size, number of pupils, etc.
Not all of this information may be available for all schools, but we’re aiming to be inclusive as possible. However the key dataset, the consumption data, is available for the majority of schools in our area.
The following sections include more detail on how some of the different datasets will be collected, published and used.
Like a number of local authorities, B&NES offers a scheme in which it negotiates energy contracts on behalf of local schools. Schools that join the programme benefit from a better deal than they may be able to arrange individually.
For several years now, schools in the area have been fitted with smart meters that monitor their electricity or gas usage (or both). Some schools may actually have several meters fitted across the school property. The smart meters are read automatically by the utility companies via GSM. A problem with early smart meter installations was that the meters were changed every time there was a change in supplier. But, as part of the UK’s smart meter roll-out, there is a move towards having a single meter for each property. Data access will eventually be managed via an intermediary.
Today there are several energy management platforms that are used by local authorities. B&NES currently uses Systems Link. The platform aggregates data from multiple utility companies and then provides a portal for the council and authorised users to access the data. These platforms provide a number of ways to report and analyse the data on-line, but are less useful for casual users such as teachers or pupils.
Energy usage data is currently shared data. Energy Sparks will help make this data open and more accessible to a wider variety of users. Our plan is to openly publish local school energy usage data via the Bath: Hacked data store. You can read more about the process and explore some sample data.
Bath and West Community Energy are a local, community-owned energy company. They run a number of local solar installations, including a number at our local schools.
This data from these installations is harvested automatically and aggregated using a platform called Solarlog. There is a portal that provides some charts that visualise the energy being generated. Access to the raw data would help improve the information about how much energy school’s are really consuming. Energy that is used from their solar cells won’t be reflected in the usage data from their smart meters.
The data is currently public data, but is not available in a machine-readable format. Energy Sparks will help publish this data as open data, again via the Bath: Hacked data store.
Local weather observations, and specifically temperature measurements, can help inform the analysis of energy data. Weather normalisation of energy consumption can make it easier to perform comparison of energy consumption across different periods or locations.
There are several ways to get local temperature readings:
- From the Met Office Hourly Site Specific Observations. However, as we’ve documented here, there is only data from a limited number of stations. The closest to us is in Bristol.
- From local weather stations, e.g. via the Met Office WOW site, Weather Underground, or Open Weather Map API
There is also the Degree Days service which offers downloads of normalised weather data for different locations. Unfortunately there are issues with many of these, including lack of clarity around licensing, ease of access, etc. We are likely to use data from this weather station via Weather Underground.
Our goal is to ensure that all of the data we’re using in Energy Sparks is licensed as open data. Any data that we generate will also be similarly licensed.