2023 Theses Doctoral

Decarbonizing the electricity sector in Qatar

Al-Aali, Ibraheam Ali

Limiting global warming to 1.5℃ requires transitioning to low-carbon electricity grids. In Qatar, high and predictable insolation synergetic with demand makes exploiting solar energy particularly attractive to decarbonize the electricity sector. With a hot desert climate, space-cooling drives demand, accounting for nearly half of annual electricity use. This dissertation analyzes a decarbonization pathway by exploiting solar PV generation combined with ice storage for cooling load shifting and battery storage for electric load shifting in a top-down approach by (i) assessing the potential for large-scale deployment, (ii) examining the subsequent problem of distributed energy resources capacity sizing, and (iii) proposing a solution to the arising demand side management problem. A carbon tax is examined to oppose cheap and plentiful natural gas.

The analysis outcomes using a linear program show a strong potential for decarbonizing using PV-enabled solutions. While they cannot displace gas generations, their role is reduced to aid in meeting summer demands. Although buildings are well suited for distributed PV, Qatar is a better fit for utility-scale implementation because of reduced costs and higher output from solar tracking technology, and accessibility for cleaning as soiling on PV is a concern.

Under the current gas price of $3.3/MMBtu, PV with ice storage could reduce emissions by 43% while cutting annual costs by 20%. Carbon pricing at $60/ton of CO₂ reduces emissions by 60%. Further reduction is difficult due to the misalignment of the summer electricity demand peak with the solar insolation peak, and ice storage cannot outcompete existing gas generation for a seasonal cooling load. Ice storage is fit to utilize the large idle chiller capacity in the shoulder season, particularly in less efficient systems, because an equal tank volume corresponds to a greater electric load shifting. Battery storage becomes economical with a carbon tax above $100/ton of CO₂ to manage non-cooling loads and is unsuitable for seasonal loads. Without a feed-in tariff, battery storage is better suited for utility-scale applications due to a reliable aggregate non-cooling load. Supported by battery storage, emissions could be reduced by 92% at $140/ton of CO₂ carbon tax. However, peak gas generation demand was only lowered by 66%.

Linear models are useful to describe large systems, but they cannot be applied to an individual system. Instead, hybrid models combining models from first principles with data-driven parameters are developed. The distributed-scale capacity sizing problem is formulated in a bi-level optimization. The upper-level decided equipment capacities using particle swarm are passed down to solve the scheduling problem to estimate electricity charges in a mixed-integer linear program with piecewise linearization. The distributed-scale analysis affirmed the suitability of the decarbonization pathway. Buildings with dominant day-time demand, such as commercial buildings, are well positioned to benefit from exploiting distributed PV generation.

Demand-side management for cooling systems becomes essential in transitioning to low-carbon power grids since intermittent renewable generations cannot be dispatched or perfectly predicted. An optimization strategy is developed to schedule and dispatch chiller systems with ice storage. The strategy decomposes the problem into a bi-level formulation solved using the genetic algorithm. The upper level decides the storage dispatch amount, and the lower level solves the scheduling problem at each time step. The penalty function method handles the scheduling problem's constraints, and with penalty factor tuning, premature convergence is eliminated. Compared to commonly used heuristic strategies, optimal control reduced cost by 11-33%. The gains are augmented with a more complex tariff structure like demand charge.