Digital Network Twins are interactive virtual replicas of a physical utility-asset that empower system operators (transmission & distribution), asset managers and power generation companies at the strategic, tactical, and operational levels of the organization to embrace energy transition through tackling the present-day and prospective challenges faced by the industry.
Digital twins are dynamic virtual representations of complex systems which include exhaustive physical, operational, and functional data across the system’s lifecycle. The conceptualization of digital twins was first presented by Michael Grieves from the University of Michigan in 2002, however, not calling it the ‘Digital Twin’, but a model for Product Lifecycle Management1. In simpler words, a digital twin is a comprehensive digital replica of any physical asset. Having said that, digital twins can be described as a class of complex simulations but are yet different from conventional simulations. Digital twins combine Operational Technology (OT) and Information Technology (IT), by the means of a bi-directional dataflow between the twin and the physical asset itself. Furthermore, the ability to run ‘what-if’ scenarios over the lifecycle of a component combined with the elements of predictability and controllability, gives digital twins an unmatched edge relative to traditional simulation techniques.
Over the past years, the development of complementary technologies the likes of Internet-of-Things (IoT), Artificial Intelligence, Data Analytics and Machine Learning, has further advanced the potential of digital twins in the application areas of predictability and controllability. There is, at present, no common definition of a digital twin developed by academia or industry experts, given their utilization across diverse disciplines, contexts, and applications over disparate sectors. At BearingPoint, we acknowledge the impact that digital twins have already created for our clients in the manufacturing, aviation, and the supply-chain industries – however, at the same time, we foresee the promising benefits of digital twins when it comes to tackling the current (and future) challenges faced by grid operators and the energy sector at large.
The energy transition and the resulting shifts in supply-side and demand-side dynamics, requires new ways to actively manage electricity networks and re-think long-term investment strategies in order to tackle the present-day challenges faced by the energy sector and achieve both operational and capital efficiency.
The gradual shift to cleaner sources of energy has been a double-edged sword so far. Renewable energy sources are eco-friendly and emit lesser pollutants relative to fossil-fuel based power plants. On the other hand, there is no clear vision on the (future) production capacity of renewable sources given that some renewable energy sources are location-specific (e.g.: Windfarms), most require additional energy storage capabilities and not every form of renewable energy is commercially viable. To add to the uncertainty of supply-side capacity, the fact that only about 8,8% of the energy produced in 2019 came from renewable sources in the Netherlands (EU average: 19,7%)2, implicitly exerts geopolitical pressures from regulators, climate activists and policy makers on the Dutch power generation companies and grid operators. Moreover, the evolving use-case of prosumers i.e., the ability for consumers to produce electricity via renewable sources such as solar panels, adds to the mix of uncertain supply-side power generation variability. For system operators, this challenge implies that they are required to be capable to deal with the supply-side volatility in terms of varying peak-loads on the networks, which arise from the use of renewable energy sources.
Due to the COVID-19 pandemic, electricity demand slumped across the world in the past year. In the European Union alone, the annualized average growth rate of electricity demand in 2020 was -8,2% 3. Nonetheless, predicting electricity demand is increasingly complex and challenging. Demand forecasting techniques evaluate factors such as: time, weather conditions (humidity and temperature) and historical load data, however, they do not consider the influence of socioeconomic developments such as the pandemic which can dramatically impact social behavior. Furthermore, the electrification of large parts of the economy (e.g. E-mobility/EV infrastructure), increases demand volatility and makes it even more challenging to predict demand.
System operators are required to always maintain a balance between electricity supply from powerplants and demand from consumers. Our analysis in the plot above depicts a constant linear trendline in load forecasting deviation for 15-minute intervals over the last 5 years (2015 – 2020) in the Netherlands. In other words, the average deviation is the root mean square error between the day-ahead load forecast and the actual load and is significantly variable between 1,1 GW to 2,25 GW. Hence, if system operators forecast a load of 10 GW between 12:00 and 12:15 for the next day, the actual load can lie anywhere between 8,9 GW and 11,1 GW given a deviation of 1,1 GW. The median deviation on the plot shows the extent to which load forecasts can be overestimated in some years and underestimated in others. To put these numbers into perspective, 1 GW can power around 1 to 1,5 million homes assuming an energy usage of 3500 kWh/year per home and normalizing for peak demands. Given the increasing complexity of networks, partly due to the non-linear nature of EV and related infrastructure adoption, we are of the view that forecasting deviations may tend to fluctuate even further going forward. For energy market participants, this translates into a higher level of uncertainty when it comes to applications such as: load balancing and demand forecasting.
There is growing evidence of rising energy losses on transmission and distribution electricity networks as a result of increasing complexity of these networks, higher voltage fluctuations and longer transmission distances from the power generation plant to the end consumer. In the Netherlands alone, the transmission grid losses increased by approximately 150% from 2010 to 20194. Transmission losses are not only economically unreasonable for system operators but can also lead to premature asset breakdowns and instrumentally increase their carbon footprint. The fact that decentralized energy storage is still in its infancy and requires substantial investments in infrastructure at present, grid losses remain a challenge for the energy sector. For grid operators, this challenge is indicative of more bottlenecks in the context of utility-asset management, predictive maintenance and repair planning among others.
A Digital Network Twin is a multidimensional digital representation of a physical network that offers a helping hand to energy market participants in order to combat alike issues of today and the potential challenges of tomorrow. The Digital Network Twin offers the architecture where raw data, data models and analytics techniques come together. The digital twin has various application possibilities at different levels in the organization, such as improving the implementation of a particular process at the operational level to performing simulations at the strategic level.
At the strategic level, Digital Network Twins offer limitless capabilities such as Emissions Calculation, Demand Forecasting and Asset Portfolio Planning, among others. In the specific application of Emissions Calculation, Digital Network Twins can estimate the levels of emissions, a grid discharges as a byproduct of electricity transmission and distribution, in turn, enabling organizations to stay on track with their Sustainable Development Goals and ESG considerations. At the tactical level, in the application of Asset Portfolio Management, Digital Network Twins can anticipate when component(s) on the network may breakdown over their lifecycle by the means of predictive maintenance, allowing a reduction in network downtime and preventing untimely breakdown costs. In addition, Digital Network Twins enable organizations to effectively manage their network components and optimize their service portfolio to forecast asset performance and eventually, maximize their asset lifetime value. At the operational level, Digital Network Twins ensure that every crucial decision being made on the grid (e.g., in the case of, load balancing or congestion) is tried, tested, and simulated prior to the deployment on the field. Digital Network Twins are a class of comprehensive solutions that empower utility-asset managers, system operators and power generation companies to achieve both operational and capital excellence.
Eager to know more about how we empower our clients to build core capabilities when it comes to their digital twin investments? Stay Tuned for our next article!
Want to get started on your digital twin journey or just interested to know more about Digital Network Twins? Talk to our experts and learn how digital twins can optimize your inter-and-intra organizational processes and create lasting value for your business.
Manager People & Strategy
Service Line Leader Data & Analytics
1. Grieves MW. Virtually Intelligent Product Systems: Digital and Physical Twins. In: Flumerfelt S, Schwartz KG, Mavris D, Simon Briceno, eds. Complex Systems Engineering: Theory and Practice. American Institute of Aeronautics and Astronautics; 2002:175-200.
2. European Commission. Share of renewable energy in the EU up to 19.7% in 2019. Published 2020. https://ec.europa.eu/info/news/share-renewable-energy-eu-2020-dec-18_en
3. International Energy Agency (IEA). Global Energy Review 2020. Published 2020. https://www.iea.org/reports/global-energy-review-2020
4. TenneT. Transmission Grid Losses. Published 2020. https://www.tennet.eu/e-insights/energy-transition/grid-losses/