Managing wind power variations through dispatchable generation in carbon-constrained energy systems

Doctoral thesis, 2024

If the European Union is to achieve climate neutrality by Year 2050, as envisioned in the European Green Deal, electricity generation from variable renewable energy sources will have to increase, both in share and in absolute volume, to replace fossil-based electricity generation. This will be necessary to meet the rising demand linked to the anticipated widespread electrification of the transport and industry sectors. Given the inherent variability of renewable electricity generation, the aim of this work is to advance understanding of the interplay between electricity generation, energy storage options, and flexible demands in future energy systems that match the European Union's ambitions for Year 2050. To study this interplay, techno-economic optimization models are applied. These models determine the cost-optimal capacity mix and system dispatch by minimizing the total system cost and ensuring that the supply consistently meets the demand.

A key focus of this work is the potential system value of shifting electricity generation in time via hydrogen, including its production and storage, as well as the conversion of hydrogen back to electricity. This hydrogen pathway, primarily involving hydrogen-fueled gas turbines for reconversion, presents an opportunity to enhance the value of electricity by shifting it from periods of abundant generation to periods of scarce generation. This work demonstrates that incorporating hydrogen-fueled gas turbines – or any other technology with similar cost characteristics and fuel flexibility – can lower the total system cost and reduce the amount of curtailed electricity. This hydrogen pathway is particularly competitive in wind-dominated regions, even though hydrogen production costs can be lower in solar-dominated regions. The critical factor is the residual load profile, which in wind-dominated regions often fluctuates over timescales that range from hours to several days. Notably, this hydrogen pathway remains competitive despite the low round-trip efficiency, an aspect that is often highlighted as a significant drawback for this hydrogen application.

This work also delves deeper into hydropower in energy systems modeling, aiming to enhance the representation of this technology and avoid overestimating its operational flexibility, in addition to evaluating the future role of hydropower. Concerning the role of hydropower in a future Swedish electricity system, the findings suggest that the trend of a weaker correlation between hydropower generation and intra-day load variations will persist and will even grow stronger in the future. This implies that hydropower will mainly serve as a complement to wind power rather than acting directly to balance the demand. Consequently, the value of operating hydropower at varying levels for periods ranging from several days to a couple of weeks will be higher in future electricity systems, underscoring the importance of accurately accounting for internal hydropower limitations in energy systems modeling. Moreover, the development of the Swedish electricity system appears to have a limited effect on the dispatch of Swedish hydropower. This is largely due to the fact that variations in generation that occur outside of Sweden are dominated by wind power, and due to the interconnecting transmission capacity, Swedish hydropower is exposed to these variations regardless of the system that is built in Sweden.

In addition to the intra-year fluctuations of weather-dependent electricity generation, this work also examines inter-annual variations, which are primarily driven by variations in wind power generation. Concerning the inter-annual variations, it is vital to distinguish between annual capacity factors and hourly generation profiles. While variations in the annual capacity factors mainly influence investments in the volume of wind power, variations within the generation profile mainly affect investments in different storage technologies and peak power capacity.  From the work conducted for this thesis, it can be concluded that during years or extended periods of low-level wind generation, fuels such as biomethane, methanol, biodiesel, or even fossil equivalents are likely to be cost-competitive options for balancing inter-annual variations that exhibit low recurrence. These fuels are not only storable at reasonable cost but can also be used in gas turbines, which have low investment costs and, thus, do not significantly affect the total system cost. In fact, gas turbines offer a unique complement to address low-occurrence variations, including both shorter fluctuations within years and extended periods of very low generation, such as consecutive weeks of scarce generation, which occur once a decade or even less frequently.