Utility Week

UTILITY WEEK | 6TH - 12TH MARCH 2020 | 25 Operations & Assets Share big data to meet net zero Using the fundamentals of open, shared data to improve the accuracy of predictions will be central to eff ectively managing power networks as we rapidly adapt to and enable a net zero world. View from the top Patrick Erwin, Northern Powergrid D ata-driven predictions, arti cial intelligence, data science – the value and power of data is increas- ingly apparent and the positive impact of opening up data in other sectors, including transport, is now challenging "business as usual" in the energy sector. With growing concerns over climate breakdown and the enormous complexity of data-driven climate model- ling that goes with it, the value we place on big data is only set to become more noticeable. This is perhaps most pronounced when planning the future of our energy net- works. The e• ective development of power networks is reliant on collaboration between stakeholders. To plan, we must agree on what the most likely future scenarios are, based on all the best data, plans and information available. Without comprehensive data, the agreed con- sensus could be wrong, resulting in misplaced resources – or worse, a network struggling to manage demand. As an industry we've become exceptionally good at managing uctuating energy demand and matching supply to it. Access to stakeholder information ows has been central to our success. However, there are new chal- lenges on the horizon. Transport, heat, industry – these are all sectors rapidly turning to electri cation to decar- bonise, ahead of the UK's net zero 2050 commitment. For example, according to National Grid's Future Energy Scenarios (FES), the number of electric vehicles (EVs) in the UK could reach 36 million by 2040 and contribute an additional 6.5GW in increased peak demand, posing a huge challenge for network operators. To prepare, we must try to nd out which sectors are likely to electrify rst, how fast, and understand the areas of the network in need of additional capacity. That means to prepare the network today, we must predict how many people will be driving an EV in ten years' time, where they will be driving, and what type of EV they will buy. We believe that for energy to work in that world, we need to stop thinking in terms of power and move to a smarter, more exible world in which we manage energy across the system, "match-making" between supply and demand while optimising for cost and minimising carbon. Data and information have proved valuable assets for energy networks, helping us make more informed predictions. In fact, we already have a very good idea of what kind of network constraints to expect over the next 30 years. However, the stakes are high – if consensus with our stakeholders on the most likely future scenarios is built on an incomplete picture, our assump- tions could be wrong, and our network planning won't be as reliable as we hope. That's why we need to overcome a persistent problem in the industry – data siloes. We're seeing the rapid development of charging infrastructure by local authorities, housebuilders, and EV charging companies alike. There are already more than 3,000 charging points installed in our region alone – undoubtedly a great success story for the decarbonisa- tion of transport. Nevertheless, there is a risk that future local strategies are designed independently, built on individual data points and siloed from each other. The result? These chargers could be planned without con- sulting the network operator, meaning that when a con- nection request is made, deployment could be delayed. The value of stakeholders sharing plans, information and data early is that we, as an infrastructure provider, can plan ahead, and enable the sector to move at pace. Energy networks, local authorities, EV charging companies, home builders, and anyone else planning activity on the electricity network must work together to share activity, plans and data to actively break down these data siloes. Using the fundamentals of open, shared data to improve the accuracy of predictions will be central to e• ectively managing power networks as we rapidly adapt to and enable a net zero world. Further, by centralising this data and making it truly accessible to all via easy-to-use online tools and combining it with multiple scenarios of varying decarbonisation rates – for example National Grid's FES – we can con dently prepare our network for the approaching wave of multi-industry electri cation, maximising resources and protecting customers' energy bills. Demonstrating how this type of open-access platform could work, Northern Powergrid has led the way by mapping and making publicly available a swathe of raw data as part of its 2019 Distribution Future Energy Sce- narios (DFES) project. Cross-referenced against National Grid's FES and presented in easy-to-understand and visually engaging interactive time-lapse maps, the data is truly accessible to all, clearly displaying predictions on a number of parameters – including EVs, heat pumps and domestic solar PV – between today and 2050. We're now calling for all local authorities, academics and renewable energy project leaders to review this data and share any projects that will impact the underlying assumptions. By doing this, we will demonstrate how cross-industry collaboration can help ensure eŸ cient network and system planning, as the region's stakehold- ers mobilise to tackle the challenge of decarbonisation. Patrick Erwin, policy and markets director, Northern Powergrid that for energy to work in that world, we need to stop thinking in terms of power and move to a smarter, more exible world in which we manage energy across the system, "match-making" between supply and demand while optimising for cost and minimising carbon. Data and information have proved valuable assets for energy networks, helping us make more informed predictions. In fact, we already have a very good idea of what kind of network constraints to expect over the next 30 years. However, the stakes are high – if consensus with our stakeholders on the most likely future scenarios is built on an incomplete picture, our assump-

• Cover