Research on Sensitivity Analysis Application in Optimal Design of Gas Pipeline

In traditional pipeline optimal design, the coefficients of design variable in optimal models are certain. However, these coefficients' change result in the change of optimal solutions due to the external reasons. Thus there exist a problem that how to gain the changed optimal solutions when these coefficients change. The sensitivity analysis of optimal design just can solve this problem quickly and effectively. According to the characteristic of the natural gas transmission pipeline optimal model, and to solve it, the model is transformed to the standard form of Generalized Geometric Programming. After having gotten the optimal solutions of the model, the sensitivity analysis algorithm of Generalized Geometric Programming is used to analyze influences on optimal solutions when its parameters such as capital construction investment, natural gas price, pipeline materials characteristic and so on have been perturbed. It is concluded that the optimal results got by the sensitivity analysis algorithm of Generalized Geometric Programming meets the needs of natural gas pipeline engineering design within the ±20% ranges of parameters perturbation by comparing the results of sensitivity analysis with the ones of the optimal model newly-optimization. The example indicates that the optimal solutions are the most sensitive to the perturbation of such the three parameters as the cost being direct ratio with the diameter of the pipe, the cost being direct ratio with the unit weight of the pipe and the intensity limit of the pipe, are rather sensitive to the perturbation of such the parameters as the initial compressor cost unconcerned with power, the initial compressor cost being direct ratio with power, the annual management cost of the compressor stations unconcerned with power and the internal charge for gas and are little sensitive to the perturbation of such the parameters as the unit power management cost at each compressor station and the cost for maintenance and utilities.