Market design innovations for decarbonized grid operations
Date
2023-05
Authors
Li, Wanning
Major Professor
Advisor
Tesfatsion, Leigh
McCalley, James
Dobson, Ian
Ajjarapu, Venkataramana
Ryan, Sarah
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The U.S. electric power industry is undergoing a remarkable evolution to achieve its decarbonization goal with 100% reliance on low-carbon or zero-carbon power resources. However, the most cost-effective carbon-free power resources, wind and solar, are intermittent. The increasing reliance on these Intermittent Power Resources (IPRs) posts the greatest challenge to electric power system operations. Additional flexibility will be required to mitigate the impacts of the variability and uncertainty of IPRs' generation. However, the current U.S. centrally-managed wholesale power market designs are not designed to handle the high penetration of IPRs and not able to fully recognize and properly compensate this flexibility. In addition, since the marginal operating costs of the system will quickly approach zero, today's marginal cost-based market mechanisms are increasingly facing challenges in the provision of price signals for independent power producers to make decisions about the construction, operation and retirement of plants. Therefore, it creates the needs to re-think how electricity market design could provide incentive to flexible resources.
This work focuses on developing a new design for grid-supported RTO/ISO-managed wholesale power markets able to operate efficiently and reliably with high penetrations of IPRs and energy storage resources. The novel features of this proposed new design, called the Linked Swing-Contract Market Design, involve changes in product definitions, contract forms, and settlement rules, not in real-time operations; hence, these changes can be gradually implemented within current U.S. RTO/ISO-managed wholesale power markets.
The key novel feature of this proposed new design is its dynamic focus on reserve as the basic transacted product, where reserve consists of the contractually guaranteed availability of power-production capabilities for possible RTO/ISO dispatch during future operating periods. Reserve offers take the form of two-part pricing ``swing'' (flexibility) contracts in firm or option form. These contracts permit dispatchable resources to ensure they receive appropriate full compensation for all of their rendered services, including flexible reserve availability in advance of operating periods for the reduction of volumetric grid risk as well as RTO/ISO-dispatched power deliveries during operating periods to meet real-time customer power demands and grid reliability requirements.
My key conceptual contribution to the original development of this proposed new design was the complete analytical formulation of an RTO/ISO-managed contract-clearing optimization for a swing-contract market. A novel feature of this formulation -- the representation of a binary unit-availability indicator as the product of a binary contract-clearing indicator (selected by the RTO/ISO) and an availability-offer indicator (automatically extractable from a swing contract) -- permits the contract-clearing optimization to be represented as a Mixed-Integer Linear Programming (MILP) problem with a relatively small number of RTO/ISO binary choice variables to which any standard MILP solver can be applied.
My two key conceptual contributions to the subsequent development of this proposed new design are as follows. First, I have carefully formulated linkages among day-ahead, hour-ahead, and real-time swing-contract markets to ensure efficient reliable dynamic performance. Second, I have developed a new stochastic MILP formulation for improved handling of uncertainty and variability of net load arising from high penetration of IPRs with partial firming by energy storage.
My key practical contribution to the development of this proposed new design has been the development and implementation of computational studies to test the feasibility and performance capabilities of my conceptual contributions. For example, performance studies conducted by means of a modified IEEE 118-bus test system demonstrate that better reliability performance is achieved using my stochastic MILP formulation than my original deterministic MILP formulation.
The next two core chapters of this dissertation address related challenges that arise when grid operations are approaching high or even complete 100% decarbonization. A stochastic production cost simulation method is proposed for the evaluation of generating capacity reliability in power systems with penetration of renewable power. And an open-source 8-zone test system is developed for teaching, training, and research purposes that is based on ISO New England structural attributes and data. The 8-zone test system is then further extended to include wind turbine agents, which enables agent-based modeling of a power system with high renewable penetration.
Series Number
Journal Issue
Is Version Of
Versions
Series
Academic or Administrative Unit
Type
dissertation