Reliable energy allocation for the Dutch energy market
The energy market is changing rapidly. Due to the growth of solar panels, electric cars, and other decentralized energy sources, it is becoming increasingly difficult to allocate network costs to users fairly and accurately. Traditional allocation methods, which are based on annual statistical estimates, are becoming less and less suitable for this.
Together with NEDU, the cooperation platform of the Dutch energy sector, CQM developed a new sampling method for dynamic allocation. This provides more frequent and accurate insight into actual energy consumption, enabling energy suppliers to better substantiate their costs and reducing the margin of uncertainty.
Realtime
insight
Less
insecurity
Better
cost allocation
From annual estimate to dynamic allocation
Tevfik Prins from BudgetThuis and also the project manager from Nedu for this allocation project:
"Private and small business customers receive an annual total invoice from their energy supplier. Alongside the actual costs of the past year, that invoice always includes an advance payment amount for the coming year—an estimate of the costs that turns out to be too high or too low in retrospect, and which must be settled later. This advance payment therefore contains a margin of uncertainty related to the allocation of network costs. We are not satisfied with that allocation, which is why we asked CQM the question: how can this distribution be done more reliably? Although I must say that the question was not very clear at first. We approached CQM stating that we wanted to do ‘something’ about improving the distribution of customers’ energy costs. By asking the right questions and being flexible and adaptable, they ultimately helped us find a fantastic sampling method—or algorithm—for allocation. One that is dynamic and provides daily insight rather than after the end of a full year. As a result, energy suppliers have better insight into their costs and can therefore choose their rates for customers more effectively."
More accurate allocation of network costs
"What currently appears on invoices sent to customers is accurate. However, at NEDU, we believe people would benefit from more specific invoices. The current method of allocating energy costs is based on a rough estimate/uncertainty margin. This is an estimate made in advance and not based on real-time insight. As a result, it is not possible to view or invoice based on time units, meaning we only gain insight after a full year. Partly due to the arrival of new energy sources such as electric cars and solar panels, this estimate margin continues to rise. Consequently, there is unnecessary risk in the (un)predictability of energy costs, which can be reduced, meaning consumers need to pay less in advance. With the CQM sample, we believe that this margin can be halved. This is because we will then have real-time insight into energy parameters, allowing us to match our customers' energy costs as closely as possible to reality. In this way, cost allocation can be made in an even more accurate manner, meaning customers only pay for actual consumption instead of the estimate/uncertainty margin. This is also possible once a month instead of annually.”
Prepared for a changing energy market
"In addition to the concrete results of reducing that margin of uncertainty and thus better cost estimation, this offers many benefits for customers, even though it is not yet visible that there are gains to be made. However, precisely due to the advent of air conditioners, heat pumps, solar panels, and other energy sources, it is of interest to consumers and small business customers to know exactly what they are purchasing energy for. Especially if you extend this to future innovation: why should you always pay the same energy costs for your washing machine, when you might run two loads one week and five the next?"
"You think you are working on question A, but ultimately you end up with something you didn't have in mind. CQM helped us first clarify the issue and then develop the right solution."
From clarification of the question to a working solution
“We are very satisfied with the sample that was delivered. In particular, we are very pleased with the flexibility CQM provided throughout this entire process. You think you are working on question A at the beginning, but ultimately you end up with something you hadn't envisioned. Helping to clarify the issue, CQM's advisory role in general, and the actual delivery of an algorithm (sample) that will ensure groundbreaking results, make them a great partner to work with. Their added value also lies in their ability to quickly translate unfamiliar subject matter into deep, thorough analyses. It took them very little effort to turn our bottlenecks into concrete improvements. That is why I foresee a long collaboration: also to conduct a validity test later on to see what the sample has yielded. Of course, it is necessary to test whether customers are waiting for this and whether we can achieve great results by lowering the estimation/uncertainty margin.”
Basis for further development
“As I said, we want to be able to fine-tune that allocation even further by having continuous real-time insight into energy parameters. Like in the washing machine example. However, this naturally involves a degree of privacy sensitivity. It is up to the market to convince customers that reading and using specific parameters helps them reduce costs. To make energy affordable and keep it that way. Energy costs are rising annually in the Netherlands, and with this improvement, we can significantly reduce that. So if we implement this, the entire Netherlands will benefit.”
Translating complex data into reliable allocation
"It was a challenging project, which ultimately means that NEDU (and all energy suppliers) can (and can) provide more customized solutions to customers after the implementation of the sample. Of course, it is also a particularly interesting project for us as data scientists. Because we have analyzed large amounts of data to distill a sample design from quarter-hour values of many connections (anonymized households) that provides the required reliability to the allocation.
At CQM we believe in working together with the customer as one team to achieve the best results. We consulted weekly with the very committed members of the NEDU working group and it is therefore not surprising that this led to the right answer to the right question in a relatively short period of time. And finally, every customer question broadens your view. It takes a lot to ensure that you can do laundry at any time and receive an invoice every now and then. The energy industry is continuously working in an innovative and progressive way with the increasingly complex puzzle of changing energy supply and demand. Dynamic allocation of grid use is one of the pieces.”
Jan TelmanSenior Consultant
Also better substantiate complex decisions with data?
CQM helps organizations with statistical models, algorithms, and decision support for complex issues. Contact Jan Telman and discover what data science can mean for your organization.