This article is part of FORESIGHT Climate & Energy’s upcoming special print issue on City Electrification, available soon More and more urban areas are creating virtual representations of their cities
MONEY SAVING
Using simulations based on real data rather than assumptions can save grid operators money WINNING SUPPORT
Conservative building owners take some convincing to invest in smart building systems KEY QUOTE
We’re not gathering new data, we’re just bringing it together to create powerful new information
Cities are home to more than 50% of the global population, account for two-thirds of global energy consumption and more than 70% of annual global carbon emissions, according to the International Energy Agency (IEA). The tendency for humans to cluster in huge metropolitan areas is projected to continue. On current trends, more than 70% of the world’s population will live in cities by 2050, triggering massive growth in demand for urban energy infrastructure. Meanwhile, distributed energy generation is increasingly being integrated into electricity grids, just as electric technologies for transport, space and water heating are adding to demand. Without more effective management of building technologies, coupled with energy efficiency to cut demand, expensive network reinforcements and new electricity generation will be needed to cope with these changes. Digital technology holds many of the answers to an affordable energy transition, with plenty of unused potential. A digital twin is a virtual representation of an asset, such as a building, electrical grid or even an entire city, which allows a particular component, process or system to be studied. Progressive city authorities and grid operators have begun utilising the option to bring together diverse data sets, such as from sensors, smart devices and mapping systems, combining them with historical data to make a complete virtual picture. 
COMPLEX MODELS Data pictures enable the streamlined modelling, monitoring and management of real-world built environments. The simulations produced by digital twins allow the past, the present and possible futures to be studied, with Artificial Intelligence (AI) and machine learning models harnessed to make sense of the data. Such complex models are backed by faster computer processing and advancements in AI, allowing for analytical work to be performed at a highly granular level. Data from real-world events can be added to ensure the accuracy of forecasts. “We’re right at the tip of where we think this is going,” says Michael Jansen from digital twin software start-up CityZenith. “The key was rallying around data science, gaming technology and radical new forms of 3D visualisation. These are all different areas of technology, so merging those together has been a decades’ work. We’re at a place now where we’ve achieved the first ever true operating system for a city.” LONDON SNAPSHOT In the UK, a team at University College London’s (UCL) Energy Institute has been working on 3D simulations of the UK capital’s building stock to create a database of all the energy and carbon data collected through the city’s energy programmes and policies. This provides a snapshot of all London’s buildings and their energy performance to help identify the worst-performing buildings and the homes most likely to be in fuel poverty. Digital twins are a cost-effective way of identifying solutions to decarbonisation, professor Paul Ruyssevelt at UCL believes. “This is largely because we’re not gathering new data, we’re just bringing it together to create powerful new information. In the past, obtaining the same data would have involved extensive physical surveys of buildings and would have been extremely expensive,” he says. The same argument applies to a project to twin Finland’s entire grid system, which has enabled transmission system operator Fingrid to cut the proportion of time spent collecting and verifying data compared with analysing it from 80% to 20%, allowing it to plan 25 years in advance. Non-smart methods of gathering data are more expensive from both a Capex and Opex perspective, according to Sabine Erlinghagen, chief executive officer of digital grid at the German engineering multinational Siemens. In addition, the complexity of data gathering will ultimately increase to levels that are hard to handle without smart technology such as digital twins, she says. 
Twin Cities Using already available data with new machine learning technology can help maximise carbon reduction potential. ILLUSTRATION Masha Krasnova-Shabaeva
SENSITIVE ANALYSIS Simulation enables the user of the digital twin to identify where to place sensors to make informed decisions, such as where to best collect data, where to install infrastructure such as transformer stations, and where EV charging can take place. “Grid operators see huge Capex savings using simulation based on real data rather than on assumptions. If you get it wrong, you’ve spent millions on copper and transformers that might be the wrong size or in the wrong place. Digital twins give you the true picture and much more sensitive analysis of scenarios,” Erlinghagen says. Several cities have already been putting digital twin technology through its paces. In 2016, Los Angeles set a goal of transforming its electricity supply to 100% renewables. The city council directed the Los Angeles Department of Water and Power (LADWP) to determine the technical feasibility and investments needed to achieve this, as well as the implications for jobs, electricity rates, the environment and environmental justice. LADWP partnered with energy modelling experts at the National Renewable Energy Laboratory (NREL), a federal agency, on a research project known as LA100. NREL modelled and analysed different projections for LADWP’s customer electricity demand, local solar adoption, power system generation, and transmission and distribution networks, and worked with local institutions to examine changes to air quality and the potential for jobs and economic development. Using NREL’s high-performance computers, the team ran millions of simulations of thousands of buildings to examine how the adoption of new design elements, equipment, or appliances could change how much and when people use electricity; explore opportunities to electrify different transportation modes and assess when and where people might charge electric vehicles; investigate how much rooftop solar could be installed; examine costs and benefits of a wide range of technologies, including solar photovoltaics, wind, concentrating solar power, geothermal, biofuels, nuclear, batteries, hydrogen storage, and encouraging flexible demand for electricity. NETWORK OVERLOAD It also performed a detailed analysis of site-specific data for both the distribution and transmission network to ensure new resources would not overload networks and simulate how different technologies, including energy storage, could be used to ensure demand is met every hour of every day of the year. The study integrated several methods of modelling, including electricity demand modelling, economic impact analysis, distribution grid modelling, and lifecycle greenhouse gas analysis. The study proved that 100% clean electricity was not only feasible in the sprawling Californian city, but it is also beneficial to the economy and job numbers as well as the environment. It also stressed the need to create robust energy efficiency programmes to reduce demand by prioritising low-income customers. As a result, in early September 2021, the Los Angeles City Council passed a motion requiring that 100% of the city’s electricity come from clean, zero-carbon energy by 2035. LADWP is in the process of preparing a long-term resource plan for the investments needed, which should feature substantial funding for a new transmission, storage and energy efficiency infrastructure, and a ramp-up of distributed energy generation programmes, particularly rooftop solar. VILA OLÍMPIA In Brazil, multinational power company Enel has developed South America’s first digital twin of an electricity network in Vila Olímpia, one of Brazil’s main financial centres just outside São Paulo. The twin replicates the local electricity infrastructure from individual physical elements to the most complex dynamics of its operations. The project—known as Urban Futurability—was made possible by a network of sensors, which monitor voltage, current and temperature of transformers, installed on the grid, each communicating information on the grid’s status in real time to the distributor and other stakeholders such as local government authorities and national regulator, Agência Nacional de Energia Elétrica. The digital twin enables information and data from the electricity infrastructure to be made available in real-time. It can facilitate grid inspections, which are carried out with the help of augmented reality, which the team trains for using virtual reality headsets, and enable cost-cutting preventative maintenance, where risks are spotted ahead of time and acted on in advance. The system has been an “essential tool” for transforming the electricity sector and distribution networks, with the data generated improving operational efficiency, according to Enel. The $23.6 million project began in August 2020, and so far has identified more than 500 anomalies, such as rusty transformers, broken crossheads or poles, loose cables and overheating, which were quickly resolved, avoiding interruption of the power supply, and identified overheating in equipment which was then resolved, avoiding network faults. “The expectation is that the financial model will become increasingly competitive, allowing for adoption on a larger scale and an effective contribution in planning for electrification and decarbonisation of networks and cities,” according to Enel. FREE ACCESS In the US, CityZenith is so confident in digital twin technology that it has offered to freely donate it to 100 cities globally over three years through its Clean Cities, Clean Future programme. The first participant, announced in September 2021, was for a digital twin system covering a group of buildings, systems and infrastructure at Brooklyn Navy Yard in New York City. It will monitor building performance and plan energy sustainability and climate resilience projects at the yard. The digital twin will automate many manual services that are currently provided by green building retrofit consulting services, which are labour intensive and costly, explains Jansen. It will effectively create a “green building calculator” that any property can easily use. Jansen says the project will demonstrate how an investment of $0.10 per square foot will unlock $3-$5 per square foot in energy savings, with a payback of three to five years on retrofit investments. Further projects have since been announced in Las Vegas, Nevada, and Pittsburgh, Pennsylvania, while CityZenith is also talking to other US cities on the east coast and in the midwest, and in Europe, according to Jansen. 
PEOPLE PROBLEM Creating the digital twin is straightforward, according to CityZenith. The hard part is convincing building owners to invest in smart building systems identified by the digital twin as viable solutions to decarbonise the building. Payback periods need to be within five years, not ten, Jansen says. “It’s expensive to invest in a smart building system and the payback periods can be long, so many building owners will say forget it. That’s why only 1.7% of buildings in the US are green. Part of the aim of this project is to stimulate how to combine various elements to achieve the financial payback that’s required to make these investments,” he says. “The digital twin will show the city authority and investors where they should be placing their [resources] and how they should finance green building retrofits over time. The biggest challenge we’ll have is inertia—building owners are conservative. The key is to get early adopters and make sure that they have good things to say about the technology,” he adds. Erlinghagen at Siemens believes the technology itself is not a challenge to integrate with grid networks, especially since the roll-out of smart meters in many locations, which is providing readily available data that can be taken advantage of. “But to really change the way that grid operators design and carry out their processes, you need to get people, technology and processes all in sync,” she says. The IEA stresses the need to train workers as a potential challenge to making digital twin a mainstream technology in cities. Those working on projects will need advanced analytical capabilities to handle AI systems and machine learning models in community-scale projects, it says. DATA WOES There can also be issues with the data itself due to the low granularity of the measurements available, diversity in data sources, and data unavailability, the IEA says. Ruyssevelt agrees that there are often problems with data quality and his team had to undertake a lot of work on some of the datasets they used to make them useful for a digital twin. Privacy issues could also come into play in the future use of digital twins, he believes. At the moment, all the data is in the public domain. Other datasets that could be used include those used however to define fuel or income poverty, which is much more sensitive information, he points out. Advocates of digital twin technology see a strong future for the technology in decarbonising cities. “In five to ten years’ time, we envisage that we will essentially have a 3D planning system, whereby each local authority will have a model of their existing buildings that will allow them to test the applications for new build and retrofit and look at the potential implications for planning and for energy infrastructure requirements,” says Ruyssevelt. Once there is a simulation of the whole of the UK, large property holders with disparate estates across the UK would be able to use it to build a portfolio of 3D representations of their buildings, in order to review and manage their performance, he adds. Jansen believes that the real estate sector could get to net-zero much more rapidly if the use of digital twins becomes mainstream. “I can see no reason why real estate can’t get to that within five to ten years—why do we need 30? You can simulate the building, and invest in building management systems and renewable energy within that time, so why not?” •
TEXT
Catherine Early
ILLUSTRATION
Masha Krasnova-Shabaeva