In the Netherlands a dense gas distribution network delivers gas to the majority of households. Traditionally only a few big suppliers deliver gas to the network, but currently there is a growing number of new small green gas suppliers in an increasingly flexible and competitive gas transport market. The gas transport providers, responsible for delivering gas to their customers, are confronted with a much more challenging task in operating the network, especially when confronted with fast changes in supply and demand and during maintenance and incidents. This complexity asks for operator support operating the network. In this paper we show, based on physical models of a typical network, involving hundreds of kilometers of pipelines and thousands of nodes and branches, how a fast and versatile simulation model, can predict the pressures and flows in the network. This gas transport simulation: predicting pressures, flows and mixtures at every location in the network, is the basis of controlling a less robust gas network, through periods of high and low demand. We will simulate a real network with the physical relations between flow, pressure, pressure drop, time-delay and mixing. Due to the nonlinear relations between pressure and flow, the standard techniques of dynamical (electrical) network simulation cannot be applied. They lead to instabilities, numerical inaccuracies and slow convergence. We will show how we solved these problems. We further show how this prediction model fits in an operator support environment that helps network operators in real-time management of operations.??
展开▼
