Long-duration energy storage plays a vital role in energy systems with large amounts of renewable power. Its value is highly dependent on the dynamics of the market in which it operates and on its ability to store energy efficiently over time. This thesis investigates how operational flexibility and market volatility influence the performance of an integrated storage system that combines Compressed Air Energy Storage, Thermal Energy Storage, and reversible Solid Oxide Electrolysis Cells. The analysis is based on synthetic day-ahead prices representative of the market behavior of 2020-2024. A numerical framework is developed that links detailed physical models of the three storage assets to a stochastic dynamic programming controller formulated as a Markov Decision Process. The generated price signals vary in one measure of volatility, the chance of a price spike. The operational flexibility of the system is expressed in three ways: the duration of a process, the limits of power transfer between technologies, and the ability to trade multiple types of energy. A factorial study of 32 scenarios explores the effects of process duration, internal power limits, market accessibility, and probability of a price spike.

The results show that annual revenue increases consistently with higher spike probability, demonstrat- ing that the value of long-term energy increases with wider intertemporal spreads. Shorter process durations allow quicker reactions to price changes, increasing both profit and variability. Expanding access to the thermal and hydrogen markets improves revenue across all scenarios by expanding fea- sible dispatch options. Electricity trading remains the dominant operation mode at the assumed thermal and hydrogen price levels. The framework ensures that all energy flows are physically consistent, but does not yet include degradation effects, hydrogen leakage, or partial-load performance. In conclusion, integrated storage of multiple types of energy gains the most from operational agility and broad market access. Furthermore, increasing the transfer of power from and to storage facilities, without increasing the corresponding storage capacities, can decrease the overall performance of such an integrated system. Future research should include component degradation and evaluate the model against both real market data and increase in seasonal volatility of prices.