ABSTRACT

For the technical planning of the transmission system Gasunie is using a model which is balancing the gas supply and demand over a period of a year. The balancing takes place on a daily basis. The demands are temperature dependent and can be calculated for each day. Each supply has got a annual contracted quantity, a daily minimum and maximum flow, and a given quality. There is a network (pipelines, valves, etc.) involved and minimum and maximum flows are taken into account. Looking at the load duration curves of the total supply and demand it happens that maximum demand is higher than maximum supply in winter and vice versa in summer. To overcome these differences Gasunie has got the possibility to use underground storage as a balance-tool. Gas supplies have different gas qualities. Consequently this means that gas has got different qualities all over the network. At some nodes of the network, the gas quality is constrained (blending stations and Nitrogen-plants). This whole gives rise to a non-linear optimisation problem. Using a linearisation process this problem is solved iteratively. Piecewise Linear programming is used in each iteration. Formerly Gasunie solved this problem sequentially day by day. For each supply an initial supply plan is generated. In winter the supply is at its maximum and in the summer the supply is at its minimum. The sum of the daily flows in this plan equals the annual contracted quantity. Sequentially each day of the year a non-linear optimisation problem has to be solved. Deviation from the plan will be corrected as soon as possible. Use of the model showed that during the sequential process supply occasionally balanced less gas into the system as was foreseen. This shortage could have been avoided if supply had the possibility to react in summer on the demand and its interference with other network-elements in the following winter. Balancing more gas in summer would have solved this shortage. Nevertheless this disadvantage the heuristic has been applied with success. Lately Gasunie had to deal with more complex balancing problems. Analyses showed that as a consequence of the above described heuristic, some supplies didn't balance their annual contracted quantity into the system, although this was expected, and the usage of underground storage was not optimal. Because the sequential model is not sufficient and not reliable any more, Gasunie is introducing a new balancing model. This model is solving the yearly balancing problem as a whole and not sequentially. This means that, instead of solving a set of daily balancing problems after each other, one optimisation problem has to be solved which is a lot bigger in magnitude and complexity then the daily balancing problem. Realising this new model some numerical problems had to be analysed and solved. The result was a much better usage of the underground storage as a balance-tool with some nice performance and surprising side-effect that more high-caloric gas (blended with less nitrogen) was balanced into the system.

This content is only available via PDF.