Guru System's Casey Cole explains how to maximise value from data while minimising the associated risks of sharing.
We’re often told that energy data is valuable. Less often discussed is the fact that handling data can be risky. But just as not all data is equally valuable, some types of data are riskier than others. The trick is to maximise value while minimising risk.
Energy performance data from utility meters is almost worthless if it stays locked up in the meter. To realise value from the data, for it to become useful information, it needs to move between stakeholders. This is equally true for electricity, gas and heat networks but I'll frame the rest of this discussion in terms of heat.
For illustration, imagine a domestic customer on a heat network. The minimum useful flow of data would be periodic manual reads either by the customer or a meter reader. In this scenario a snapshot of the total count of kilowatt hours moves from the dwelling to the utility company just a few times a year.
There are three aspects of this utility data that we can improve in order to increase its value: granularity (e.g. shorter intervals between reads), richness (e.g. flow and return temperatures, flow rates), and timeliness (e.g. ability to find out what's happening right now). Improve the data in any of these dimensions and the potential value is multiplied. By connecting directly to the meters over a communication network, you can improve all three at once.
With valuable data freed from the meter, it can now flow between stakeholders, each of whom may have a different purpose for it. Customers may use it to control their spending and minimise wastage. A heat network operator wants it in order to minimise distribution losses and lower the cost of delivered heat. Engineers want to learn from past projects and design better systems in future. Clients procuring heat networks may use data to check they got what they paid for.
But, regardless of data’s value, some of these stakeholders are wary of obtaining and handling data because of the attendant risks. Top of the list is exposure to the Data Protection Act (DPA). Meter readings (when they're accompanied by a meter id) are personal data and so covered by the DPA. Mishandle such data and the Information Commissioner’s Office may slap a fine on you of up to £500k.
Once a stakeholder has got useful data, they may be reluctant to share it with others. For example, network operators may fear the commercial risk that comes from competitors knowing how their systems are performing. Or engineers may be reluctant to admit that a design failed to achieve its performance requirements. As a result, promising data that has been obtained from meters will now stay locked up in silos rather than helping to improve general practice in the market.
Part of the solution to these silos is for stakeholders to become more confident in how they handle data. The first step to achieve this is to put in place key systems and processes. Standards such as ISO 27001 and Cyber Essentials are invaluable for this.
With a framework in place, incoming data can then be categorised according to type and risk (perhaps by using the Open Data Institute’s data spectrum) and the appropriate controls applied. For example, anonymised data aggregated from several networks is far less sensitive than a year’s worth of half hourly data from a vulnerable residential customer.
Another part of the solution is to recognise that risk and value are not interchangeable: some of the most valuable data carries little or no risk at all. For example, as part of a recent project funded by the Department for Energy and Climate Change, the team at Guru Systems anonymised half hourly data from around a thousand homes to produce diversity curves that are essential for engineers to build better heat networks. We then published the curves as open data under a Creative Commons License. While the anonymised data is low risk, we estimate that it may be worth as much as £450m to the nascent heat industry over the next 10 years.
To sum up, data is worthless if it stays locked up in the utility meter. Improving the data’s granularity, richness and timeliness can increase its usefulness but, to get the most out of it, data needs to be able to move between stakeholders. The risks associated with handling and sharing data can make stakeholders keep it to themselves, but by developing systems and processes to handle data better and by recognising that some of the most valuable data is lowest risk, they can maximise value for themselves and the wider market.