Abstract
We discuss how an optimization model can be used together with a scenario generation procedure to provide valuable analysis for companies operating in a natural gas value chain. The solution time of the optimization model can be considerable for some model specifications, so a large scale sampling from the distribution of the uncertain parameters would lead to intractable solution times. By using a scenario generation procedure we can, however, drastically reduce the required amount of analyses necessary to run. We discuss two different procedures in this paper: moment-matching and copulas. We also demonstrate the application on a gas transportation network similar to the one on the Norwegian Continental Shelf. The data used in the analysis are synthetic, but with realistic values.
