In a simulation model solved using nonlinear mathematical programming, the role of underground storage facilities in California is analyzed in the context of a natural gas network where seasonal and weekly supply and demand are balanced with the combination of transportation, injections or withdrawals from underground storage, and short-term “linepack” in pipelines. Taking into account the decreasing efficiency of compressors as pipelines and underground storage facilities become full, the model suggests the cost-minimizing seasonal and weekly procurement of natural gas for California. Equivalently, the model represents decisions by utilities, customers, pipelines, and storage operators in the northern and southern systems that bring prices within California towards an equilibrium where spatial and intertemporal arbitrage opportunities have been exhausted, to the extent that regulatory and capacity constraints permit. The model is calibrated to consumption and production data from recent years, with the consumption data separated by customer type. The model’s temporal dimension encompasses 52 weekdays/weekend pairs. The model’s spatial focus is the two main consumption regions within California. The model distinguishes four underground storage facilities in California by location and compressor capacity. The model accounts for the links with other parts of the North American network with four “residual” supply relationships. The relative differences in weekly and seasonal cycles prove to be important for understanding the effective pipeline and injection capacities. The model is well suited to analyze the effects of substantial changes in infrastructure, such as the LNG terminal in Baja, Mexico, which will import the seasonal prices of northeast Asia as well as the natural gas. Another example of “scenario analysis” with the model is the proposed substitution of natural gas for coal in electricity generation in Ontario, which will reduce the advantage to California from importing from Alberta during the summer for injection into underground storage.
