Networks and Spatial Economics, 3: (2003) 97–122 C 2003 Kluwer Academic Publishers, Manufactured in the Netherlands. Computational Experience with a Large-Scale, Multi-Period, Spatial Equilibrium Model of the North American Natural Gas System STEVEN A. GABRIEL Project Management Program, Department of Civil & Environmental Engineering, University of Maryland, College Park, Maryland 20742-3021, USA email: sgabriel@eng.umd.edu; http://www.eng.umd.edu/sgabriel JULIO MANIK SHREE VIKAS Energy Cluster, ICF Consulting, Fairfax, Virginia 22031-1207, USA email: jmanik@icfconsulting.com email: svikas@icfconsulting.com Abstract In this paper we describe numerical results for a market equilibrium model for the North American natural gas system. The model is based on the notion of maximizing total surplus less transportation costs (Takayama and Judge, 1971; Samuelson, 1952) resulting in a large-scale nonlinear program (Gabriel et al., 2000). The model relies on building up supply curves from the “bottom up” using a data base of some 17,000 natural gas reservoirs. This feature provides a good deal of realism in simulating the effects of technology, market forces, and policy considerations on the supply side of the market while making the computations challenging due to the lack of closed form supply curves. A successive linear programming strategy is employed to solve the overall nonlinear problem. We describe several mathematical algorithms that are employed in the successive LP approach to efficiently compute market equilibrium values. These algorithms are heuristic in nature with excellent convergence results. In the numerical results section of this paper, we describe several experiments regarding schemes to accelerate the overall convergence based on iterative smoothing (similar to a Gauss-Seidel strategy) as well as tests aimed at optimal spatial and temporal aggregation. Keywords: Spatial modeling, energy equilibrium, natural gas, energy sector optimization, nonlinear program- ming, successive linear programming, Gauss-Seidel 1. Introduction The North American natural gas system is an example of a spatially diverse market connected via a pipeline network structure. This market includes Canada, the United States, and Mexico and is well integrated with for example, suppliers in Canada competing against producers in the U.S. for selling natural gas. Natural gas is used in four main consumption sectors: residential, commercial, industrial, and electric power. Its role in the electric power sector is becoming increasingly more prominent in the greatly expanding North American electrical power market for two main