Operating strategy for load service entities using flexible real-time pricing through stochastic dual dynamic programming

The increasing penetration of renewable energy sources has presented challenges in maintaining power system reliability through traditional supply-side controls. Consequently, utilities have been developing demand-side management strategies to ensure grid stability. Dynamic pricing (DP)-wherein electricity prices vary with market conditions to control demand-has been widely adopted. Among DP approaches, real-time pricing (RTP) shows the highest responsiveness but remains limited because it typically mirrors wholesale prices and the elasticity of electricity demand is extremely low. This paper proposes a flexible operating strategy that utilizes real-time pricing (RTP) and the elasticity of electricity demand for demand response (DR). The core design of this strategy is as follows. First, to enhance the accuracy of demand response, a demand model was designed that integrates both self-time and cross-time elasticity to capture the inter-temporal effects of demand. Second, the optimal price-adjustment mechanism is formulated as a multi-stage optimization problem, which is solved by applying Stochastic Dual Dynamic Programming (SDDP) for the first time in the field of demand response. Third, the uncertainty of consumer response is managed through probabilistic scenarios based on a novel modified Markov Chain. The primary objective of the proposed strategy is to flatten the demand curve by minimizing the variance between hourly demand and day-ahead forecasts. Case studies demonstrate that this strategy balances demand and stabilizes both grid operation and utility profits.