Project update report: Development of energy system models to improve treatment of energy system architecture
This part of the project was both a response to the research agenda set out in phase one, and to the research questions set out in the original funding proposal. The work programme for this part of the project involved two whole energy system models, ESTIMO and UKTM, and two teams of modellers, led by John Barrett and Steve Pye respectively.
The ESTIMO model
The ESTIMO model has been used to develop predominantly renewable energy system architectures and associated Concepts of Operation (ConOps) that meet heat and other energy demands with zero emissions. Three heat technology shares have been explored: 70% consumer heat pumps, 70% district heating using a mix of heat sources including heat pumps, and 70% electrolytic hydrogen heating, with the remaining 30% assumed to be consumer heat pumps. Eight variants of these architectures have been designed to support an extended exploration of interactions between the heat sector and the rest of the energy system. All scenarios include a significantly increased cooling demand in the housing and non-domestic sectors, consistent with a mid-range estimate of warming by 2050, and two additional scenarios begin to explore the range of potential impacts of climate change.
Extensive work has been undertaken to develop operating algorithms for these systems. The intention is to build an optimisation capability into ESTIMO. In the meantime, limited optimisation studies have been undertaken using a simplified version of ESTIMO. The main goal of the work has been to explore the landscape of infrastructure options and associated ConOps that would be capable of providing the whole of UK energy demand with minimal constraints, across many decades. And in turn, to shed light on the trade-offs, for example between investments in different categories of infrastructure that are implicit in choices of heat decarbonisation strategy, and which need to be considered by policy-makers and energy system stakeholders in the development of energy policy and investment strategy.
The different systems have been simulated with fixed demand profiles and stocks of infrastructure, at hourly resolution using 35 years of historic meteorology data. The model therefore illuminates questions related to flexibility and resilience, but not to evolvability. Designs that are consistent with the reliability goal include 300-500 GW of generators, 10-20 TWh of storage of different kinds, and 20-50 GW of interconnectors linking the UK electricity system to various parts of Europe. To avoid trivial comparisons, both energy demand and renewable generation in the equivalent European energy system are also modelled explicitly. Among the findings is that even with a significant electrolyser capacity, an average of 40-50% of renewable electricity generation is spilled. The hydrogen-dominated architecture results in the largest requirement for renewable generation capacity, and highest overall cost, because of the thermodynamic inefficiency of the final stage of conversion from H2 to heat. A paper describing the ESTIMO model is near completion. The final months of the project will be used, among other things, to provide guidance on factors such as capacity margins and storage requirements to the UKTM team.
UKTM, a model primarily used for scenario analyses of the future of the UK energy system, but recently used for heat decarbonisation analysis (see Broad et al. 2020), was considered in view of two ESA features – evolvability and robustness.
Evolvability concerns the ability of the system to change over the medium to longer term, and the implications of changing course, even after specific choices and decisions have been made. This is an important capability for adjusting with minimum disruption to unforeseen future changes to the heat sector, such as rapid cost reductions in alternative technologies, large shifts in the amount and timing of demand, and additional requirements such as increased cooling provision or improved air quality, equity or energy security. Evolvability captures ideas of how existing infrastructure can be reconfigured to move to a different system, e.g., repurposing the gas grid to hydrogen, and how a system can start to transition while leaving options open and allow for dynamic adjustment (Scamman et al. 2020). For this research, we are using a myopic foresight version of UKTM (as described in Fuso Nerini et al. 2017), to reduce planning foresight to align more closely with real world decision making. Dominant heat decarbonisation pathways are being explored, to assess costs, deployment rates, decision points, and wider system implications of pathways.
Robustness concerns decisions that reflect future uncertainty and are therefore more robust to a range of outcomes. In this part of the research, we use a probabilistic version of UKTM using XLRM scenario framing (Lempert et al. 2003) to assess what uncertainties heat decarbonisation pathways are most sensitive to. This has meant adding the function to UKTM to run large ensembles of scenarios. For decision makers, it is important to understand these uncertainties, the impact that they might have, and therefore what might be a robust course of action.
As part of both of these research efforts, on evolvability and robustness, UKTM has been further developed to explore net-zero systems. This has been an important development of the model, in order to align it with current UK mid-century ambition.
The final year of the project will involve two stakeholder workshops to communicate the results of the Heat Challenge, and to explore the use of modelling results to support a participatory exploration of the tradespace for options to decarbonise heat in the context of the net-zero target and other key energy system objectives. The final months of work from both modelling teams will also include an inter-comparison of modelling results between the two models.
Banner photo credit: Ricardo Gomez Angel on Unsplash