Planning Optimization Of Polyvinyl Chloride Production By Calcium Carbide Method

Polyvinyl chloride(PVC)is a kind of important commercial polymer which has been widely used as raw materials of various kinds of petrochemical products.PVC production by calcium carbide method accounted for 76.2% of market share in China because of abundance in coal and shortage of gas and oil.However,as a “heavy energy consumption,heavy pollution and resource-related” process,planning optimization used to direct production is imminent under the unprecedented pressure of energy conservation and emissions reduction.In this paper,a real PVC plant by calcium carbide method is studied considering process operation and energy consumption.An integrated idea is putforward considering both material production system and combined heat and power(CHP)utility system.For energy intensive equipment,such as calcium carbide furnaces,electrolytic cells and CHP units,the nonlinear characteristics of energy consumption are studied respectively.Then,a mix integer nonlinear programming(MINLP)model is proposed for a multi-period planning of a PVC plant by calcium carbide method in whole process.The proposed MINLP model is in large scale so that it's difficult to solve.In order to accelerate the solving speed,a decomposition strategy based on hierarchy is studied.However,there are many uncertain factors,such as demand uncertainty,price uncertainty of raw materials,price uncertainty of sales,production capacity of equipment,utility system and so on.These uncertain factors have important effects on the production plan.Therefore,planning optimization for PVC production under uncertainty is also studied.The main researches in this paper are as follows:(1)Integrated planning optimization of a typical PVC production process is studied in order to solve a problem that material production system and utility system represended by CHP are isolated from each other.For material production system,nonlinear characteristics between energy consumption of energy-intensive equipment and product rate are studied.For utility system,the nonlinear characteristics between energy consumption and power of CHP units are studied and the decision-making of electricity purchase from the state grid is also considered.The integrated planning problem is described as a MINLP model based on the discrete time representation.(2)Operating state in binary variables,operating load in continuous variable and nonlinear characteristics are optimized simultaneously in a MINLP problem.The resulted model scale is too large and the solving difficulty hinders the industrial application.In order to accerate the sloving speed,a separate optimization strategy of operating states and operating loads is proposed.First of all,load hierarchy and energy consumption hierarchy of calcium carbide furnaces,electrolytic cells and CHP units are establied according to the principle of “priority to choose the lowest energy consumption equipment”.Then,load hierarchy and energy consumption hierarchy can take place of nonlinear characteristics.The MINLP model is converted to a mixed integer linear programming(MILP)model.Thus,operating states in binary variables are obtained quickly by optimizing MILP problem.What's more,operating loads in continuous variables are determined through a nonlinear programming(NLP)model which is got by fixing the binary variable.(3)To improve the robustness of proposed model,modeling and solving in PVC production are studied considering a variety of uncertain factors.Planning model of a PVC plant based on the discrete time representation under uncertainty is establied using the chance constrained programming theory to describe uncertain factors,such as demand uncertatinty and price uncertatinty of raw materials.Uncertain constraints can be converted to deterministic constraints using deterministic model transformation method and then optimization results are obtained using proposed decomposition strategy based on hierarchy.