Modelling the impact of the electrification of heating and transport on the mix of wind, solar, and storage.

Abstract

The mix of technologies in a future energy system is an important area of research. As the proportions of wind power and solar power generation capacities in the power system increase, energy storage becomes necessary to complement their variability due to weather. The subject of this thesis is how the wind/solar/storage mix will be impacted due to changes in electricity demand caused by the electrification of heat and transport using the UK as a case study.

To account for long-term weather effects, whilst keeping socioeconomic and technological factors constant, a novel method has been developed for incorporating heating electricity into the historic electricity demand. This enables the impact of heat pumps alone to be studied. The research reveals that for predicted 2050 heat pump penetration levels the monthly demand for electricity doubles in winter. This leads to an increase of approximately 30 TWh for each winter month and a 37% increase in year-to-year variability of electricity demand due to weather. Temperature dependent electric vehicle time series incorporated into the electricity demand show a seasonal variation far lower than for heat pumps, with annual demand varying over a range 14 TWh between years with EVs compared to only 10 TWh without.

These electricity demand time series were used in a novel high-level energy model to calculate the minimum energy storage required for a generation mix including wind, solar, base load and dispatchable power sources. To provide sufficient granularity and variation in conditions, hourly demand and generation time series based on 40 years of historic weather were used. The optimum wind/solar energy mix from perspectives of storage, excess energy and cost were investigated. It was found that heating electrification increases the proportion of wind energy required by between 3% and 5%. In contrast, a change to providing most transport by electric vehicles does not significantly change the optimum proportion of wind energy required. Providing all heating with boilers fuelled by hydrogen generated via electrolysis would require 7 times as much renewable energy compared to all heating provided by electric heat pumps. Energy lost due to curtailment exceeds energy lost to storage efficiency by an order of magnitude.

A sensitivity analysis on alternative model inputs showed that different wind generation time series have the most impact on predicted energy storage capacity. Simulated offshore and onshore wind and actual national grid generation all have different patterns. This can have a significant impact on storage requirements, a result that has not been noted before. Although previous research has shown the impact that wind turbine locations have on the amount of energy they generate, when they generate this energy has had little attention.

A novel comparison of four heat demand methods found that the method used by the when2heat dataset most accurately predicted the measured data which is an important
result considering it has already been used by several other studies. The methods were validated against national gas time series, and measured data not previously used for
this purpose. It was also found that peak electricity demand is very sensitive to the method of generating heat demand and hourly heat pump operating profiles, suggesting
inaccuracies of 25% in previous estimates of future peak demand.