Using improved control schemes to utilise the buffering capacity in gas headers to absorb upstream disturbances

A spectral element method (SEM) model was developed and validated using measurements from a real process for simulation and control applications. The figures show that model is complex enough to represent a real header process accurately, but simple enough to be implemented practically.
The importance and optimal placement of pressure measurement locations on the gas headers for buffering control is determined through the combined use of the SEM model collocation points, an extended kalman filter (EKF), and a staggered pressure control scheme in a self-optimising control (SOC) framework. The figures show that the optimal pressure measurement location was found to be dependant on the distance from the flow disturbances and final control elements as well as the individual importance of consumer flow stability, expensive supplier use and gas emissions.
Suitable control strategies to reduce gas emissions and improve flow stability for consumers were investigated and compared using the SEM model and a Monte Carlo simulation.The figures show the effect of variance in the gas properties on the total cost using the Sobol method. A chosen buffering control scheme was implemented on a real gas header and compared with standard PI control. The figures show that the chosen scheme (CLMPC) reduced emissions and improved consumer stability by intelligently utilising the available pressure buffering capacity in the industrial gas headers.
All these figures therefore show that: it is possible and beneficial to mitigate flaring and improve consumer stability in industrial gas headers through the use of suitable pressure buffering control strategies, that the pressure measurement location impacts the performance of the control strategy and can be provided through a SOC framework using the combination of suitable state estimators and the SEM model developed, and that the SEM model accurately represents real industrial gas header dynamics and is suitable for use in practical control and simulation applications.

基于真实工艺测量数据，构建并验证了一种光谱元素法（SEM）模型，以应用于模拟与控制领域。图示表明，该模型在精确表征实际进料过程方面复杂度适中，同时足够简洁以供实际应用。通过结合使用SEM模型的配置点、扩展卡尔曼滤波器（EKF）以及交错压力控制方案，在自优化控制（SOC）框架内确定了气体主管道压力测量点的位置及其对缓冲控制的至关重要性。图示显示，最优压力测量点的确定依赖于从流动扰动和最终控制元件的距离，以及消费者流量稳定性、昂贵供应商的使用和气体排放的个体重要性。采用SEM模型和蒙特卡洛模拟，研究了旨在降低气体排放并改善消费者流稳定性的适宜控制策略，并进行了比较。图示展示了利用Sobol方法对气体性质变化对总成本影响的效应。选取的缓冲控制方案在真实气体主管道上得到实施，并与标准PI控制进行了比较。图示表明，所选方案（CLMPC）通过智能利用工业气体主管道中可用的压力缓冲容量，降低了排放并提升了消费者稳定性。因此，所有这些图示均表明：通过适宜的压力缓冲控制策略，在工业气体主管道中缓解燃料气排放并提升消费者稳定性是可行且有益的；压力测量点的位置影响控制策略的性能，并且可以通过SOC框架提供，该框架结合了适宜的状态估计器和所开发的SEM模型；SEM模型准确表征了真实工业气体主管道动力学，适用于实际控制和模拟应用。