Interdependency of security-constrained electricity and natural gas infrastructures

The electric power generation relies increasingly on the natural gas supply system as additional natural gas-fired power plants are installed in restructured power systems. In this context, the economics and the reliability of electric power and natural gas systems will impact one another. This dissertation addresses the interdependency of electricity and natural gas systems and proposes integrated approaches for the operation and the scheduling of the coupled energy systems.;This dissertation considers combined-cycle gas turbine units (CCGTs) as key elements for linking electric power and natural gas systems and proposes mode and component models for representing CCGTs. The two models are used in scheduling of CCGTs by mixed-integer programming (MIP).;This dissertation proposes two integrated short-term scheduling models. The first one is from the viewpoint of the ISO (Independent System Operator) which proposes a security-based methodology for the unit commitment solution when considering the natural gas transmission system and contracts. The proposed solution applies a Benders decomposition method to incorporate the natural gas transmission feasibility check subproblem in the security-constrained unit commitment (SCUC) solution. The second integrated model considers a joint-operator for the coordinated scheduling of the interdependent power and natural gas systems. The integrated operator utilizes an augmented Lagrangian relaxation (LR) based model for the coordinated least-cost allocation of natural gas resources to individual gas loads and power plants.;The natural gas flow exhibits remarkable differences from the electric power flow because of the slow response of the former system and storage nature of pipelines. This dissertation also proposes an integrated short-term scheduling model with the transient state natural gas flow formulations which is represented by a group of partial differential equations and nonlinear algebraic equations. The implicit finite difference method is adopted to approximate partial differential equations into algebraic difference equations. The scheduling coordination problem is described as a bilevel programming formulation. The objective of the upper-level problem is to minimize the operating cost of the electric power system while the lower-level optimal natural gas scheduling problem is nested as a constraint. Based on Benders decomposition methodology, a coordination scheme is proposed and corresponding optimization algorithm is developed.