A flexible ramping product: Can it help real-time dispatch markets approach the stochastic dispatch ideal?

• Compares a deterministic dispatch market with flexiramp with stochastic dispatch.
• The results illustrate how up- and down-flexiramp enhance market efficiency.
• But flexiramp constraint is insufficient to produce the least (expected) cost.
• The exact parameters of the constraint affect the quality of the solutions.
• A penalty approach is proposed to decrease the possible inefficiency of flexiramp.

U.S. Independent System Operators (ISOs) are creating short-run markets for so-called “flexiramp”. The aim of these markets is to ensure that enough flexible generation capacity is on-line to manage the increasingly volatile net loads resulting from growth in renewable energy. In particular, we assume that the purpose of flexiramp is to improve the expected performance, in terms of costs, prices, and reliability, of the ISOs’ deterministic market models. Therefore, we compare the solutions of (1) a deterministic dispatch model with a flexiramp constraint that simulates ISO operations with (2) a stochastic dispatch model that, by definition, obtains schedules that minimize expected cost. Dispatch, prices, settlements, and market efficiency are contrasted in a simplified case study to explore the fundamental reasons for successes (and failures) of flexiramp markets. The results illustrate how flexiramp can enhance market efficiency. However, they also show that procuring flexiramp is insufficient to minimize expected costs, and that market parameters affect the quality of the solutions. The simulations furthermore show that deterministic markets with flexiramp can yield either higher or lower prices than the stochastic optimum. We propose a penalty-based approach to mitigate possible biases towards choosing capacity with high energy costs to provide flexiramp, and conclude that market operators will need to monitor market performance and adjust flexiramp parameters in order to maximize market efficiency.