Search

The inherent coupling of the electric and natural gas systems due to the operation of gas generating units and power-to-gas facilities, along with the uncertainties faced in both systems due to the variability in electricity and gas demand and the vastly increasing volatile renewable injections, create an imperative need to schedule and operate the two systems in a coordinated manner. In this paper a new model for the fully integrated stochastic day-ahead scheduling of electric and gas systems is presented, coping with the uncertainties of both systems. The stochastic parameters comprise the electricity demand and the renewable injections, which collectively create several net electricity load scenarios, and the gas residential/industrial demand. The integrated scheduling problem concerns a unit commitment for the electricity problem, amended with additional constraints imposed by the underlying natural gas transmission system considering steady-state flow. A two-stage stochastic programming model is devised, having as second stage the possible realizations of net electricity load and gas demand in real-time. The model is tested in medium-size real-world test systems - the Greek electricity and gas systems - deriving useful insights on the advantages of the integrated stochastic scheduling versus the deterministic scheduling of the electricity and gas systems.

A coupled optimization of the electricity and gas systems is presented in this paper. The electricity problem involves a unit commitment with co-optimization of energy and reserves under a power pool, considering all system operational and unit technical constraints. The gas problem involves a medium-scale highly non-convex and non-linear problem structure, which is modeled as a Mixed Integer Non-Linear Programming model. The decomposition of the overall problem is based on the Augmented Lagrangian method. An iterative process is implemented, coordinating the two interdependent systems using an alternating minimization method, in which the Lagrange multipliers are updated using the subgradient method. The gas problem is solved in two phases in order to avoid numerical instabilities; first, the direction of flow is defined, and then the gas flow is derived in the second phase. The solution algorithm is evaluated using the Greek power and gas system, comprising thirteen gas-fired units and fifty-three gas network nodes. The test results indicate the strong interdependence of the two systems, and demonstrate the efficiency of the presented algorithm in coordinating them.