Premium
Establishing independence of continuous process product samples
Author(s) -
Christiansen Reed L.
Publication year - 2000
Publication title -
aiche journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.958
H-Index - 167
eISSN - 1547-5905
pISSN - 0001-1541
DOI - 10.1002/aic.690460716
Subject(s) - residence time distribution , process engineering , process (computing) , residence time (fluid dynamics) , laminar flow , product (mathematics) , continuous flow , computer science , continuous reactor , heuristic , flow (mathematics) , mathematics , engineering , mathematical optimization , biochemical engineering , chemistry , biochemistry , geotechnical engineering , catalysis , aerospace engineering , operating system , geometry
Abstract New product qualification frequently requires that a manufacturer provide customers with several independent samples for evaluation. For continuous manufacturing processes, equipment residence time distributions (RTDs) introduce serial dependence among the samples. Determining RTD by pulse or step modulation of a tracer or process variable can produce off‐spec material and add costs by limiting the opportunity to conduct other product or process development experiments. A novel method of determining a continuous process' RTD does not risk production of off‐spec material. Through analysis of serial component analytical data, the process' washout curve can be inferred. The theoretical basis for the method is presented and experimental requirements enumerated. Applicability to well‐mixed, laminar flow and convective‐dispersive processes is demonstrated. Experimental validation in a specialty polymer pilot plant resulted in an error of about 5% in mean residence time, when compared to estimates from pulse tests and calibrated level measurements.