The description of our methods are organized around the steps outlined in our framework diagram in Fig. 1.
Some modelling approaches are designed to provide a ‘prediction’ or ‘forecast’ of the future. That is not the case with our analysis; rather we are creating ‘simulations’ of potential futures based on a well-developed narrative, written by experts across a range of disciplines and fields. This narrative is used to inform a bottom-up analysis of energy service demands in each sector, which are subsequently used in an economy-wide model to construct ‘net-zero’ scenarios for the United Kingdom. Overall, our scenario approach is designed to give insights into the scale of change in energy demand and GHG emissions that is possible under certain circumstances.
Step 1 of our approach is the development of a scenario narrative that is underpinned by seven observable underlying trends that have impacted energy demand and/or are likely to continue to do so throughout the scenario timescale. Our list of underlying trends is not exhaustive as various other observable societal, political, economic and scientific or philosophical trends may impact energy demand, such as demographic changes. We also recognize that there is considerable crossover between the various trends and that some could have the potential to increase and further reduce energy demand32. Nevertheless, developing this narrative gives the sectoral scenarios a consistent underpinning. The underlying trends are listed below and described in further detail in Supplementary Note 4. Supplementary Table 1 in Supplementary Note 3 provides a mapping between the different energy services demands and the underlying trends provided below.
The integration of digital systems and information and communications technology into the energy system is already driving significant changes in the energy sector and is likely to accelerate change in the future from promoting new energy business models through to changing how consumers interact with energy services.
Sharing and circular economy
The ‘sharing economy’ is an approach that aims to decrease the number of under-utilized ‘owned’ assets in an economy by creating new business models that offer a service in its place2. The ‘circular economy’ is a concept that explores resource efficiency strategies to extend the time that resources are retained in the economy to reduce material throughput and environmental impacts33,34,35.
Energy efficiency has steadily improved by 1% annually over the past 30 years, with no indication that the potential for further energy efficiency improvements is saturated36.
The context of the COVID-19 pandemic has increased global focus on health, well-being and quality of life at the policy level and by individuals.
As climate and ecological breakdown accelerate and significant climate impacts of events, such as wildfires and flooding, gain global attention, public concern for the environment grows across the globe.
Increasing globalization in the form of international production networks and global value chains has a wide range of impacts on national final energy demand in different countries.
Work and automation
This trend considers the distinct impact that automation may have on working patterns in the United Kingdom and how this may change the demand for energy services.
The options for delivering energy demand reduction vary considerably across different sectors. The technologies and embedded social practices are entirely different for different energy service demands. We therefore employ specific sector models to determine the energy demand for each sector before combining this information into a central whole-system energy-modelling framework. A brief description of each model is included below, while more details of each model and the underlying sectoral modelling assumptions can be found in Supplementary Note 4.
For Mobility, the TEAM-UK (Transport Energy Air pollution Model for the UK)37 is used. TEAM-UK is a transport–energy–environment system-modelling framework that simulates passenger demand as a function of key travel indicators, built around detailed travel demand data from the UK National Travel Survey. The nutrition (including agriculture) and materials and products sector modelling utilized hybrid UK MRIO (multi-regional input output) models. The hybrid UK MRIO utilized for nutrition modelling is a Leontief physical input–output food system model, constructed from observed economic data of intermediate and final demand38. For the materials and products modelling, a hybrid UK MRIO model was used to determine the supply chain impacts and the ultimate impact on production requirements from UK industry. The construction sector energy service demands (included within the materials and products sector) are modelled separately using a bespoke model that considers 36 specific applications of key materials across 17 different built-asset categories. The UK National Housing Model (NHM)39 is a micro-simulation model that utilizes national housing survey data to explore the future of energy service demand in the domestic building stock. Finally, a bespoke sectoral model for non-domestic buildings was developed for this analysis and built around the UK Building Energy Efficiency Survey (BEES) dataset40 to explore different rates of energy efficiency uptake.
A critical part of the sectoral modelling process is to map the dependencies between sectors to ensure consistency so that key aspects of the narrative represented in one sector are also reflected in others. This is an important part of the approach to ensure that the scenarios are internally consistent.
Many linkages relate to the shelter analysis, highlighting dependencies where shifts in working patterns fundamentally change the types and patterns of mobility demand and the use of non-domestic buildings. In the case of non-domestic buildings, repurposing floor space in offices and retail spaces links back to impacts upon shelter demand. Other important cross-sectoral linkages include differences in house-building assumptions, and choices about necessary transport infrastructure are reflected in the demands for construction materials. Further, changes in vehicle sales impact manufacturing. Changes to diet impact the level of land take, which feeds into land availability for forestry and the level of output in specific food and drink subsectors, impacting energy demand in that sector.
The next step of the approach is to combine the sector analyses into an integrating framework, primarily to explore the system-wide implications of lower energy demand requirements on energy supply and the role of CDR. For this, we use UKTM, a technology-explicit, whole-system, partial-equilibrium model that relies on the TIMES model-generator framework41. The model optimizes future energy system evolution using linear programming, optimizing future investment choices to meet energy service demands at least cost (based on minimizing the discounted net present value for the whole system). The model has been used across a wide range of energy scenario studies42,43,44,45,46. In recent years, it has been co-developed with the United Kingdom’s energy ministry (BEIS), which has used it extensively to inform energy strategies47,48,49.
The model represents the existing energy system in 2010, including the existing infrastructure assets (power generation plants, vehicle stock and so on) and flows of energy both between and across sectors. In UKTM, the whole system is represented, from resource extraction, through to primary and secondary fuel production (electricity, hydrogen, biofuels) and finally consumption in the residential, industrial, service, transport and agricultural sectors. This final consumption is used to meet the wide range of final energy service demands needed across the economy, such as mobility, heating and industrial production. In addition to energy service demands, GDP and population growth are exogenous to the model. These two variables have not been adjusted in the scenarios, and further research is required to explore the potential effects on GDP in particular with considerably lower energy demand.
For scenario exercises, projected energy service demands are exogenous inputs into the model. Their changes over time are the main drivers for investment and system development in the model which solves by exploring least-cost supply-side solutions to meet these future service demands. ‘Supply-side’ refers to any part of the system used to supply energy to meet energy service demands. This includes transformation/conversion processes, for example, electricity generation, and all of the technologies used in end-use sectors, for example, gas boilers, cars, cement kilns and so on. It also includes some explicit energy-saving measures in the buildings sector, such as fabric retrofit. The whole-system representation allows for explicit trade-offs between sectors in respect of resource allocation. Demands for energy vectors, such as electricity and hydrogen, are endogenous to the model and sensitive to changing prices driven by the dynamics of balancing demand and supply. The other benefit of the whole-system representation is that it allows for comprehensive and internally consistent accounting of both energy- and key non-energy related greenhouse gases, including, for example, agriculture and land use. This means the model can be used for exploring energy systems that meet climate and energy policy goals.
Under each scenario, the sectoral modelling (steps 2 and 3) provides estimates of energy reduction through ‘improve, shift and avoid’ measures. Two types of information relevant to energy demand are passed for use in UKTM:
Energy service-demand projections
These projections inform how energy services will change over time, based on ‘avoid’ measures and some ‘shift’ measures in transport. They are exogenous inputs to the whole-system model and are based directly on projected energy service demands from the sector models (Supplementary Tables 3 and 4 in Supplementary Data 1 provide projections).
Technology efficiency measures
The sectoral analyses take account of opportunities for ‘improving’ the efficiency of energy use and shifting to cleaner energy use. Such measures include improved efficiency of technologies, switching to electricity-using appliances and building retrofits. Such measures are typically considered endogenously by UKTM; therefore, we have not hardwired efficiency-related energy demand reductions in to UKTM from the sectoral modelling. Rather we have tried to align input assumptions on technology efficiency and deployment rates, followed by an iterative process of checking model outputs with sector teams. The approach to endogenize sectoral assumptions in UKTM means that there are some differences between the sectoral and UKTM outputs. Differences have been tolerated where these are not significant, particularly as the key insights from UKTM relate to implications for energy supply to end-use sectors. Appliances were included within domestic and non-domestic sectors.
The linkages between sector models and UKTM are described in turn below. Further information on specific UKTM assumptions across sectors can be found in Supplementary Note 4.
The sectoral modelling approach for mobility includes the development of energy service demands based on in-depth assessment of a range of behavioural levers that are then fed into the UK Transport Energy and Air Pollution Model (TEAM)37 to explore journey distance by purpose, mode choices, vehicle choices and rates of deployment. International aviation was restricted to an assessment of energy service demands influenced mainly by socio-economic, demographic and policy (for example, changes in the cost of air travel via pricing such as frequent flier levy) drivers. Shipping was not included in the TEAM analysis and was considered separately by the UKTM team.
UKTM received energy service-demand projections for all transport subsectors, except shipping. On energy technology assumptions, the key alignment was on vehicle efficiency factors to those in TEAM. The modelling teams also iterated on UKTM constraints, including rates of technology deployment.
This analysis uses the UK’s National Household Model (NHM), focusing on heating requirements under different scenarios, which factor in varying levels of new house building, retrofitting and other behavioural changes.
For this sector, UKTM received the energy service demands for space and water heating. Given that these energy service demands already include heat demand savings from energy efficiency measures, building-retrofit options were switched off in UKTM to avoid double counting. The NHM already underpins energy technology assumptions in the UKTM residential sector, meaning that further alignment was required only around elements adjusted for this analysis, for example, heat pump deployment rate and coefficient of performance as they differ between scenarios.
The sectoral modelling for non-domestic buildings was built around the UK Building Energy Efficiency Survey (BEES) dataset50 and is used to explore different rates of energy efficiency uptake across the three modelled scenarios. By reviewing the current and future expected building stock for each of the main subsectors, including main commercial, leisure and public service building uses, this approach estimates the full technical savings potential across the sector and translates different levels of ambition into varying growth rates for corresponding energy efficiency options.
The results from this model provided direct input into UKTM by informing the total energy efficiency gains from building-retrofit and management measures that are not related to technology replacement. These include fabric, building instrumentation and control and carbon and energy management systems. Their roll-out was then limited in UKTM according to the levels of ambition relevant to each scenario. Efficiency gains from technology switching (primarily through electrification) are estimated endogenously in UKTM.
In parallel, different growth trajectories for future building stock number—proxied through total floor space requirements—were developed for each scenario. These were built specifically for this sector but were developed in consultation with experts across the project to mirror changing pressures on, for example, storage space requirements in line with changes in retail shopping and home delivery. These floor space requirements then provided the main energy driver input into UKTM as their change over time was used to inform future growth in energy service demand in the model.
Materials and products
The input to the industry sector of UKTM was from the sector analysis of resource efficiency gains as estimated in UK MRIO. The approach was to first apply resource efficiency percentage gains to the UKTM growth drivers; these growth drivers are largely taken from the UK government econometric energy demand model, EDM51. In addition, further adjustment factors to account for changes in infrastructure construction in other sectors, for example, buildings and transport, were applied to key sectors producing construction materials, for example, iron and steel and cement.
The integration into UKTM of the sector analysis of nutrition covers the resulting on-farm agricultural changes in terms of emissions and land availability, due to changes in the overall national diet and scenarios for the reduction of food waste throughout the supply chain. Because of the low level of detail in the UKTM agriculture sector, the link was made through the emissions of CH4 and N2O relating to crops and livestock which were updated to follow the trends seen in the sectoral modelling for each scenario. The resulting land freed up by a shift to a more plant-based diet was used to define new limiting constraints on the planting of forests within UKTM (both for biodiversity and energy crops) such that the more ambitious the nutrition scenario, the more land that becomes available for forests out to 2050. Finally, the assessment of food waste generation from the sector analysis for each scenario were used to adjust the trends shaping the scale of food waste production in UKTM.
We have developed four scenarios, Ignore, Steer, Shift and Transform. These are described in more detail in Supplementary Table 2 (Supplementary Note 4).
Our Shift and Transform scenarios are our so-called low energy demand scenarios. They assume a national attempt to rapidly reduce energy demand in the United Kingdom to increase the likelihood of meeting ambitious climate outcomes in the short and long term. Our scenarios provide an analysis of the total final energy demand in the United Kingdom and are also broken down into the five high-level categories of mobility, residential buildings, non-domestic buildings, nutrition and materials and products.
In addition to the low energy demand scenarios, two additional scenarios were considered at the system level in the UKTM analysis for comparative purposes. These include a scenario called Ignore Demand, based on achieving reductions as estimated in the Climate Change Committee 2018 progress report, including medium risk policies52. The second scenario is called Steer demand, which aims for net-zero GHG emissions by 2050, based on all the United Kingdom’s legislated carbon budgets, including the sixth carbon budget (−78% GHG reduction in 2035 relative to 1990 levels). However, the scenario fails to achieve the 2050 target, falling short by 27 MtCO2 despite high levels of removal. It relies on improved energy efficiency and supply-side options only, with no consideration of measures for avoiding energy use or shifting to options that supply energy services with less energy, for example, private cars to public transport. This implies that without further demand-side efforts, the United Kingdom’s net-zero target will be very difficult to meet.