(500d) A Novel Method to Compare Chemical Process Flowsheets

A chemical flowsheet is a diagram that contains a wealth of information to describe a chemical process. By recognizing similarities in process flowsheets, chemical engineers can understand ways in which to improve process design and operations. A systematic methodology has been proposed in the literature [1] to identify process patterns in chemical flowsheets. In their proposed method, SFILES strings [2-3] are generated and sequence comparison algorithms [4] are used for identifying flowsheet patterns. The purpose of the present work is to provide an algorithm that is complementary to that proposed in [1] and, as a result, is capable of identifying novel types of patterns in flowsheets.

In this work, we develop a novel comparison method to compare chemical process flowsheets. In the first step, we identify key process units in a chemical flowsheet. Subsequently, the flowsheet is converted into a tree with identified key units. Each node in the resulting tree corresponds to a key unit. We perform a depth-first graph traversal to store the process information in each key node. In this process, the traversal history is stored as an ordered sequence. Similar structures of two process flowsheets are then identified by comparing sequences in key nodes. We use the F1 score to measure the similarity between two compared sequences. Several case studies are conducted, including a comparison of 14 flowsheets covering 7 diverse chemical products. The results show that the proposed method successfully identifies multiple similar structures in two compared flowsheets.

Keywords: flowsheet comparison, sequence alignment, pattern recognition

References

[1] Zhang T, Sahinidis NV, Siirola JJ. Pattern recognition in chemical process flowsheets. AIChE J. 2019;65(2):592-603.

[2] d’Anterroches L, Gani R. Group contribution based process flowsheet synthesis. Fluid Phase Equilib. 2005;228:141-146.

[3] Tula AK, Eden MR, Gani R. Process synthesis, design and analysis using a process-group contribution method. Comput Chem Eng. 2015;81:245-259.

[4] Smith TF, Waterman MS. Identification of common molecular subsequences. J Mol Biol. 1981;147:195-197.