Short- and long-term variability in future electricity systems – Ensuring the flexibility to manage the grid frequency and inter-annual variations

Abstract: To achieve climate mitigation targets, the net climate impact from electricity generation must be drastically reduced in the near future, and eventually eliminated. The level of generation must also be increased to accommodate the increasing demand as the transport and industry sectors undergo electrification. However, the ongoing transition to weather-dependent electricity generation poses challenges in terms of providing ancillary services and ensuring security-of-supply in the electricity system. This thesis applies techno-economic optimisation modelling to investigate two challenges related to non-dispatchable generation from wind power and solar photovoltaics (PV): (i) the cost-effective provision of sufficient short-term flexibility to control the grid frequency; and (ii) the cost-effective provision of long-term flexibility to manage inter-annual variability. This investigation is focused on identifying the key technologies that may provide the two forms of flexibility and on whether the cost-optimal mix of technologies to supply electricity is affected. This is achieved using a linear electricity system model that optimises the installed capacity levels and the operation of generation and storage technologies. Specifically, the model encompasses nuclear and bio-fuelled thermal power, wind power and solar PV, Li-ion batteries, hydrogen storage systems, and heat-generation technologies for inter-connected district heating systems. Electricity transmission between neighbouring regions is also modelled in Papers II–IV. It is found that batteries play a key role in limiting the cost of providing flexibility for grid frequency control, although other technologies, such as hydro- and thermal power, curtailed wind and solar power and flexible power-to-heat for district heating, also contribute. The results indicate that, with grid-scale batteries as an investment option, the demand for inertia and reserve power availability would increase the system cost by 0.5 €/MWh of produced electricity. However, the results also show decreasing costs as the shares of wind power and solar PV increase, and that a weaker impact on system cost can be achieved if reserve power is provided by flexible household loads. In terms of inter-annual variability, the results support previous studies that have demonstrated that the levels of cost-optimal investments generated by individually modelled weather-years vary widely (e.g., total thermal capacity is in the range of 74–141 GW in Northern Europe). Furthermore, the results show that accounting for inter-annual variability affects only slightly the total capacity levels of wind, solar and nuclear power. However, inter-annual variability increases the capacities of biogas turbines and may increase the need for long-term biogas storage.