Utility Week

UTILITY WEEK | 4TH - 10TH OCTOBER 2019 | 11 modelling (BIM), a system applied to construction projects that puts all data in the same format and the same language. We want to apply this to existing infrastructure, most of which is old and we need to be able to get more out of it." A lack of information about what data already exists is another challenge that creates ineff iciency under the current energy system. "It sounds very 'meta', but it means that people search for data that doesn't exist, for a very long time in some cases, or they build a business model based on a false notion that they will be able to get something that doesn't exist," says Dobson. "Companies also re-collect data that they could have gathered from an existing source at a much lower cost, potentially with a higher degree of accuracy." e taskforce therefore recommends that a data catalogue is set up to capture all available energy system datasets across government, Ofgem and industry as standardised metadata, with industry participation in the catalogue mandated via regulatory and policy frameworks. Missing data is a particular concern on older networks, but is there a way to avoid the major investments required to upgrade them with sensors? Emerging technologies such as data analytics and machine learning could provide a solution. Electralink, an organisation that runs a data transfer service for the energy sector, has been working with the DNO Western Power Distribution to explore how machine learning can interrogate customer energy consumption patterns to reveal network constraints. Legal restrictions currently prevent access to granular domestic smart meter data, so the project explored how machine learning can be applied to non-smart consumption data already available in Electralink's archive. Dan Hopkinson, director of data and transformation at Electralink tells Flex: "Using machine learning on the data we were able to identify some very strong correlations, and the potential for the existence of a lot of low carbon technologies that DNOs are not currently aware of. e data is a strong indicator if a technology like solar panels, or an EV recharger, has been applied to a property." Hopkinson admits that the model is not perfect, but the intention in future research is to incorporate additional data sets, including contextual data, switching behaviour, and demographic data such as property type, to create a more accurate picture, is could allow DNOs to model the low voltage network without the need for protected consumer data. " is form of virtual network monitoring could improve our understanding and therefore reduce costs by ensuring that real-time monitoring is only put where it is needed," says Hopkinson. B e n e f i t s m u l t i p l i e d Data may be the new oil, but if, like the black stuff , too much of it is allowed to accumulate in the hands of private companies there is a danger it will not empower and benefi t the rest of society. Reducing barriers to public and private sector data can attract innovators that create operational eff iciencies or develop new business models; it can increase the speed that new markets develop and improve decision- making across the energy system. Data sharing improves 7 Planning data Local Authority planning data is important to enable the energy system to grow and adapt to the changing needs of the nation. Having visibility of developments further in advance could enable more innovative and eff icient energy system solutions. AMBER Generally available but with inconsistent formats 8 Asset monitoring data Asset monitoring data includes technical asset monitoring (network, generation, storage, etc.), smart meter metrics and the associated models. e blend of this monitoring data enables actors to operate the system eff ectively, identify areas for optimisation / innovation and utilise proactive network maintenance models. RED Data is collected for some assets but it is not ubiquitous and the data is not widely shared 9 Market data Data relating to energy markets and the fi nancial operation of the energy system. is information is critical as it lets innovators identify and target areas where costs are disproportionately high, enables carbon pricing to have maximum impact and enables price signals to drive benefi cial behaviours. AMBER Existing market data is often available but via paid services, emerging markets are less transparent. 10 Cross infrastructure data Sharing data across infrastructure sectors (water, communications, transport, waste, etc.) could help to minimise public disruption from groundworks, reduce costs for consumers and identify innovation opportunities. RED There are good projects such as the Geospatial Commission Underground Asset Register but data which is currently openly available is limited 11 Weather data Granular weather data is fundamental for modelling output of renewable generation, asset condition, asset eff iciency, etc. GREEN Data accurate and widely available (open and paid) 12 Demographic data is is critical for modelling, business model innovation and low carbon transition modelling. GREEN Data is of a good quality and widely available (open and paid) Dr Richard Dobson Practice manager – data systems, at the Energy Systems Catapult Dr Richard Dobson Practice manager – data systems, at the Energy Systems Catapult Data overview

• Cover