(661b) Lessons from an Integrated Optimisation Model of the UK Power, Heating, and Hydrogen Sectors | AIChE

(661b) Lessons from an Integrated Optimisation Model of the UK Power, Heating, and Hydrogen Sectors

Authors 

Mersch, M. - Presenter, Imperial College London
Markides, C. N., Imperial College London
Mac Dowell, N., Imperial College London
Future energy systems are likely to be highly integrated. Increased electrification of heating and transport, and fuel switching from natural gas and petroleum products to blue or green hydrogen will likely be key enablers of the transition towards carbon neutral energy systems. Together with a higher share of volatile renewables in the power sector, with large-scale storage, firm generation capacity and demand-side management as balance mechanisms, this will break up the traditional structure of separated energy systems and markets.

Conventional energy system models tend to only consider single sectors, most commonly looking only at power generation. For future projections, they rely heavily on assumption regarding e.g., added demand from electrification of other sectors. Likewise, many studies on low-carbon heating technologies depend on electricity and/or hydrogen price assumptions. The enforcement of emission targets is also associated with implicit assumption. For example, it could be more economical to allow residual emissions from heating, which are then balanced by negative emissions from the power sector. Therefore, integrated energy system models of multiple sectors are of high interest when investigating energy system transitions.

We present here an integrated capacity-expansion and unit-commitment optimisation model of the electricity, heating, and hydrogen sectors in the UK. The development of technology portfolios is developed starting from today’s installed capacities until 2050, allowing investment, retrofit and decommissioning decisions every 5 years. Additionally, the dispatch schedules are optimised on an hourly basis for each of the decision years, using either full-hourly time series or a typical-day clustering approach. The clustering method preserves periods of peak demand and low renewables availability, such as not to underestimate the required technology capacities. A central planner perspective is taken, meaning that all decisions are made by one entity with the target to minimise total system costs, rather than individual stakeholders minimising their costs or maximising their profits. Perfect foresight is assumed, which means that technology portfolios and operation are optimised over the entire time horizon simultaneously rather than applying a rolling horizon. The model is implemented as an LP model in GAMS, solved by Gurobi, with data pre- and post-processing done in Python.

The electricity sector is modelled using adapted equations from the original electricity sector model developed by Heuberger et al. [1][2]. A variety of electricity generation and storage technologies are considered in the model:

  • Conventional thermal power plants (coal, open-cycle gas turbines (OCGT), combined-cycle gas turbines (CCGT), nuclear)
  • Coal plants and CCGTs with carbon-capture and storage (CCS)
  • Hydrogen-fired CCGTs and OCGTs
  • Biomass plants with or without CCS (the former of which can provide negative emissions)
  • Renewables (on- and offshore wind, solar, hydropower, geothermal plants)
  • Interconnectors
  • Large-scale storage (pumped hydro, Li-ion batteries)

Constraints include hourly demand matching, reserve and inertia provision, build-rate constraints to limit capacity expansion to reasonable levels and operational constraints of technologies. Additionally, capital, fuel and various other operational costs are calculated, as well as emissions.

The heating sector is split into three parts: domestic heating, commercial heating and industrial heating. Each come with their own demand, own set of technologies, and constraints. One important distinction to the power sector is that for the heating sector the demand has to be satisfied per end-user, since end-users cannot share heat with each other. For that reason, the domestic building stock is clustered into a number of representative buildings, and technology portfolios and operation are optimised for these representative buildings. The overall technology capacities, fuel use, costs and emissions are then aggregated based on the number of buildings that fall into each category. A similar approach is taken for commercial heating, where properties are split into office and retail buildings, restaurants and cafes, and warehouses/storage facilities/industrial buildings. In each category, a high-demand and a low-demand property is considered. Industrial heating is split into high-temperature and low-temperature demand. Technologies considered for heating are natural gas and hydrogen boilers, air-source and ground-source heat pumps and electric resistive heating, as well as water tanks for storage.

Hydrogen is an option as fuel for both, low-carbon heating and electricity. It can also function as storage medium. Following a similar structure as the electricity sector, hydrogen generation and storage capacities are optimised, and ideal dispatch schedules are determined. Considered generation technologies are methane reformers with CCS, electrolysers and biomass gasification with CCS, which can provide negative emissions, similar to biomass power generation with CCS. Salt caverns and line pack are considered as storage options.

The different sectors are explicitly integrated in the modelling. Electrification of heating for example will result in a higher electricity demand, which in turn requires more electricity generation assets. The same is true for the hydrogen sector, where all demand is due to heating or electricity generation, no exogenous hydrogen demand is assumed. This integration and simultaneous optimisation eliminates many of the aforementioned key uncertainties related to single-sector energy system models.

We present results from different optimised energy system transition pathways, focussing on trade-offs and interactions between the different sectors. The integrated model allows us to explicitly quantify the impact of decisions in one sector on the others. Additionally, we are able to determine optimal emission trajectories for the different sectors that lead net-zero overall emissions. The prioritisation of resource use, e.g., is it better to use biomass for power generation or for hydrogen generation via gasification, can also be evaluated using the integrated model.

References

[1] C. F. Heuberger, I. Staffell, N. Shah, and N. Mac Dowell, “A systems approach to quantifying the value of power generation and energy storage technologies in future electricity networks,” Comput. Chem. Eng., vol. 107, pp. 247–256, Dec. 2017, doi: 10.1016/j.compchemeng.2017.05.012.

[2] C. F. Heuberger, E. S. Rubin, I. Staffell, N. Shah, and N. Mac Dowell, “Power capacity expansion planning considering endogenous technology cost learning,” Appl. Energy, vol. 204, no. August, pp. 831–845, 2017, doi: 10.1016/j.apenergy.2017.07.075.