Continuous-time look-ahead flexible ramp scheduling in real-time operation

•This paper proposes a novel continuous-time look-ahead optimization model in Fig. 1 to co-optimize energy and flexible ramp products in real-time operation.•This paper models up and down flexible ramp capacity provided by generating units as explicit continuous-time decision variables and ensures the deliverability of ramping capacity by reserving the respective energy requirement in generation capacity constraints of units.•The regulation up and down capacity of generating units are modeled by continuous-time trajectories and their respective ramping requirement is accounted for in the ramping constraints of units.•In order to ensure the availability of adequate energy and ramping capacity for real-time operation, an enhanced continuous-time day-ahead unit commitment (UC) model is developed to co-optimize the energy as well as balancing and regulation reserve over day-ahead scheduling horizon.•A function space solution method is developed that converts the continuous-time real-time and day-ahead operation problems respectively into linear programming (LP) and mixed-integer linear programming (MILP) problems, which can be solved using commercial solvers. This paper develops a novel continuous-time look-ahead optimization model for co-optimizing balancing energy and flexible ramp products in real-time power systems operation. In addition, a continuous-time day-ahead operation model is developed to commit and schedule an adequate subset of generating units for providing balancing energy and ramping requirements in real-time operation. The up and down flexible ramp products are defined as continuous-time decision trajectories that continuously supply the real-time up and down ramping requirements of the system. The deliverability of flexible ramp products is secured through reserving the respective power requirements in generation capacity constraints of generating units. A function space solution method is proposed to reduce the dimensionality of day-ahead and real-time models through projecting them into finite-dimensional function spaces spanned by Bernstein polynomials, where the real-time operation decisions are modeled with a higher accuracy level than day-ahead decisions. Simulations are conducted on the IEEE-RTS system using CAISO load data, and the numerical results showcase the effectiveness of the proposed model in supplying the energy and flexible ramp requirements and avoiding ramping scarcity in the system, while reducing the operation costs.