Optimal selection of intermediate storage tank capacity in a periodic batch/semicontinuous process

Abstract

Batch/semicontinuous chemical plants are usually designed either by assuming infinite intermediate storage or by assuming that the units themselves act as storage vessels, while the storage vessel,^ are sized by rules of thumb or experience. In this paper, the case of an intermediate storage vessel which links one upstream batch/semicontinuous unit to one downstream batch/semicontinuous unit is analyzed. The units are assumed to operate with fixed cycle times and capacities. Expressions for determining the minimum storage tank capacity necessary to decouple the two units are derived from a mathematical model of the periodic process. Effects of the relative starting limes of the two units on the required storage capacity are determined, thus suggesting the optimum process timings to minimize the same. Application of the results is illustrated by an example.

Optimal design of batch‐storage network using periodic square wave model

“…age analysis. Karimi and Reklaitis 1983 developed analytical results for the case of limiting storage volume in serial systems composed of arbitrary configurations of batch, semicontinuous, or continuous operations. They extended their results to the multiple inputrmultiple output intermediate Ž .…”

Deterministic variability analysis for intermediate storage in noncontinuous processes. Part I: Allowability conditions

“…The inductive proof clearly shows that the elementary variations are superimposable and this property will prove quite useful in the subsequent analysis of composite variations. If the variations are nonoverlapping then theorem I holds true for a general L-M system also (Karimi, 1984).…”

“…The quantitative analysis of this role of intermediate storage has been termed variability analysis (Karimi and Reklaitis, 1984). As reviewed there, such analyses have taken two basic directions: stochastic variability analysis and deterministic variability analysis.…”

“…Following the terminology introduced in Karimi and Reklaitis (1984), we define an L-M or parallel process to be one in which there are L batch/ semicontinuous units in the upstream stage, M batch/semicontinuous units in the downstream stage, and both stages are separated by a storage vessel of size V*. In this notation, a serial process is denoted as a 1-1 process For purposes of this paper theL-M process will be characterized by the nominal or average values of its parameters.…”

Optimizing inventory policies in process networks under uncertainty

“…Karimi and Reklaitis (1983) recognized the importance of intermediate storage for batch and semicontinuous processes, and derived expressions to find optimal storage capacities according to the periodicity of the production processes. Other models for multiproduct batch plants have included uncertainty in the design problem (Wellons and Reklaitis, 1989;Shah and Pantelides, 1992;Ierapetritou and Pistikopoulos, 1996), but they have not considered inventory management in their formulations.…”