Systematic Procedures for the Design of MultiProduct Plants,Multipurpose Plants, and Plant Complexes

这项研究的目标是为多产品和多用途工厂的概念设计开发系统的综合程序和快捷的评估技术，包括将问题分解为抽象空间的层次结构，其中层次结构中的每个层次包含的细节比上面的层次更多，但比下面的层次更少。首先考虑解决方案的最重要的功能，然后通过将不太重要的细节层添加到在先前级别开发的部分解决方案来开发解决方案。PI使用了这种技术，并已经表明，在层次结构中较低级别所做的决定永远不会反映出消除冲突解决问题的早期决定。在同时生产多个产品，但市场需求与通过反应器的一次生产的产品分布不相对应的情况下，合成石化过程的系统程序将是一个非常有价值的过程设计工具。同样，在某些情况下，在同一设备中通过不同的化学路线生产一种以上的产品可能是有利的，因为每种产品只在一年的部分时间内生产。这种操作通常需要多用途工厂以批处理模式运行，但并不总是如此。本项目要解决的第三类问题是工厂综合体的设计，即设计生产反应中间体的工艺，然后使用中间体生产最终项目。所有这些算法的开发将是有价值的，不仅因为将有更有效的设计程序可用，而且还将作为指导本科生设计的教学工具。

The goals of this research are to develop systematic synthesis procedures and short-cut evaluation techniques for the conceptual design of multiproduct and multipurpose plants which involves breaking the problem down into a hierarchy of abstraction spaces, where each level in the hierarch contains more detail than the level above it but less detail than the level below it. The most important features of the solution are considered first, then the solution is developed by adding layers of less important details to the partial solution that was developed at previous levels. The PI uses this techniques and has already shown that decisions made at lower levels in the hierarchy never reflect back on earlier decisions eliminating the problem of conflict resolution. A systematic procedure for synthesizing petrochemical processes where multiple products are produced simultaneously but where the market demand does not correspond to the product distribution produced in a single pass through the reactor (e.g. glycols and amines) would be a very valuable process design tool. Similarly, in some cases it might be advantageous to produce more than one product by different chemical routes in the same equipment, where each product is made for only part of the year. Multipurpose plants often, but not always, run in a batch mode are needed for such operation. The third category of problems to be addressed in this project is the design of plant complexes, i.e. the design of processes to produce reaction intermediates, and then use the intermediates to produce a final project. The development of all of these algorithms would be valuable not only because a more efficient design procedure would be available but also as teaching tools for the instruction of design to undergraduates.