Managing demand uncertainty in natural gas transmission networks

Abstract Natural gas transmission networks (GTNs) are widely used to transmit natural gas over long distances while the efficient management of GTNs requires as accurate as possible information about the demand of the customers. However, the demand of the users can only be predicted with a limited accuracy. Accordingly, the uncertainties resulting from the demand prediction errors should be addressed to increase the robustness of the daily planning of GTNs. An optimization algorithm based on chance-constrained programming technique is developed to consider the uncertainty of the forecasted future demands, whose solution is the set point of the compressor stations discharge pressure. An important prerequisite of chance-constrained programming formulation is the mean and variance of the objective function and constraints. Accordingly, unscented transform is used as the uncertainty propagation tool. Then, the methodology is applied to a long distance transmission pipeline with 8 compressor stations to illustrate the effectiveness of the proposed approach where the end user load is characterized by a Gaussian random variable with known mean and variance. The results of the optimization based on the proposed algorithm are compared to the deterministic ones where an optimum pressure buffer can be obtained based on the mean and the variance value of the customer demand which is required for the safe operation of the network. Pressure buffer is also affected by the probability assigned for the constraint satisfaction. The results of the algorithm are compared to deterministic approach in terms of power consumption and it is concluded that to gain robust solutions to overcome demand uncertainties, extra power usage is essential. Also, it is shown that demand uncertainty handling requires a specific strategy rather than operating the system based on the expected condition.