On October 13, 2017, within the CAPE-OPEN 2017 Annual Meeting, Professor Ross TAYLOR, together with his co-workers Harry KOOIJMAN and Jingsong ZHOU at ChemSep and Clarkson University, delivered a recorded presentation (also available as a PDF, 1275 Kbytes).
Ross TAYLOR is the Kodak Distinguished Professor of Chemical Engineering at Clarkson University in Potsdam, New York where he has been since 1980. The ChemSep project was started in February 1988 at The University of Delft and is now a consortium supported by several major companies. Harry KOOIJMAN is the co-architect of ChemSep while Jingsong ZHOU is a PhD candidate at Clarkson University. Some of Jingsong’s work is presented here.
Ross starts by thanking the CAPE-OPEN community for its contributions in making ChemSep more widely known thanks to its CAPE-OPEN interfaces.Then Ross demonstrates how a summary report including cost estimation of a given column may show up in a flowsheet simulator when hovering over the icon representing a ChemSep Unit Operation. Ross would like this functionality to be commonly available in all process simulators (PMEs) and part of the CAPE-OPEN standard. Within the specific Graphical User Interface provided with ChemSep, there is now a rating panel. Rating provides also cost estimation and permits comparison between different designs based on economics.
Ross supports the point that the costing should be done unit operation per unit operation, within each unit operation while the PME should be made aware of the costing information supplied by each unit operation (this looks like an application of Flowsheet Monitoring – editor’s note)
The rating panel in ChemSep gives the possibility to save information from the column calculation in a file which adheres to the specific formats of a number of vendors of column internals.The KGTower software from Koch Glitsch and the SulCol software from Sulzer ChemTech are two examples of such vendor tools.Results from these rating tools need to be entered manually in ChemSep in order for these results to be taken into account.
Next Ross addresses the issue of the slowness of rate-based separation models. While such a slowness in inevitable considering the kind of model used, Chemsep was recently capable to speed up significantly calculation and this is significant especially for non well-mixed flow patterns.
Ross goes on with discussing appearance issues. In a large number of PMEs, CAPE-OPEN Unit Operations come with all the same icon, making them not recogizable one from another. This proves even more difficult for the end-user when the same flowsheet carries more than one CAPE-OPEN Unit Operation. The issue is also that having the same icon to represent a CAPE-OPEN Unit Operation whatever its configuration, reduces considerably the transmission of detailed information to the end-user. The capability to let evolve the icon representing a CAPE-OPEN Unit Operation according to its configuration has been introduced in COFE and proves very useful. Ross would like this functionaliy to be somehow picked up by the CAPE-OPEN standard. An icon could also be used to represent a sub-flowsheet such as the one described for the representation of divided-wall columns. However the true response to the challenge of modeling divided wall columns relies in an equation-oriented approach that ChemSep is experimenting as the Parallel Column Model (PCM). PCM can also be applied for dealing with maldistribution models. PCM will be made available in ChemSep 8.according to the current development plans.
Finally Ross adresses a number of issues with respect to CAPE-OPEN. First of all he warns against the deprecation of CAPE-OPEN Thermodynamic and Physical Properties interface specification in its version 1.0 since all ChemSep users have not access to implementations of version 1.1 even though ChemSep is capable to use CAPE-OPEN Thermodynamic Servers adhering to version 1.1. Ross also asks for specific gravity at 60 F to be introduced in the list of properties recognized by CAPE-OPEN.
In terms of conclusion, Ross summarizes the new features of ChemSep 7.2, then gives a wish list on CAPE-OPEN re-stating the points mentioned above, and then ends up with the request for a CAPE-OPEN Cost Library of callable functions.
Feedback and comments to this presentation are welcome by Ross and Harry. Contact information is available on the ChemSep member page.