Chemical Processing - June 2008 - (Page 21)

Key considerATions When moving models from design into operations, pay particular attention to the following points: • Number of chemical components. Design models may contain more than are needed in a process model. Fewer components will speed up simulations. • Is the model topology up to date? Has the plant changed since the design model was developed? • What are the valid ranges for the process model? At what throughput? • Does the model need to handle different product grades? If so, you may need a range of alternative models. • Does the model need to account for different ambient conditions such as different heat losses and different utility temperatures in winter/summer or night/day? • Does changing catalyst activity have to be considered? • Which model inputs can be fixed, which will be manually entered and which will come from real-time data systems? • What are the lower and upper limits for all model inputs? • Some equipment models may require changing from design to “rating.” For example, in design the heat exchanger is specified by outlet conditions with no utility stream included. In rating, both sides of the heat exchanger are included and simulated with heat-transfer coefficient and area. • Distillation column efficiencies may have to be matched to plant data or equilibrium-stage models converted to mass-transfer-based ones. • Can the plant still operate with some equipment switched off? If so, the model will need to account for this. • What are the keys results to be calculated by the model? • Which equipment items can be deleted from the design model? Which of these are not required for the particular on-line calculation? • Must any additional equipment items be included? For example, long pipes, valves and pumps sometimes are left out of design models. • How robust is the design model? Can it cope with the different input values in the plant model? A model path Process engineer on plant Troubleshooting Simulation engineer Performance monitoring Traditional approach Open-loop Real-time model Tune model Design model Plant engineer, shift supervisors, operators Decision support Off-line process model Simulation engineer Conﬁgure to plant Process design engineer Model development simpler custom interface, such as one based on Excel instead of their normal “engineering” user interface. The sidebar provides other practical pointers for moving models from design to operations. The nexT sTeps Figure 2. New uses of design models in operations require changed workflow. If an off-line process model gets regular use in operations, it may be appropriate to convert it to a realtime open-loop model. The model execution then can be automated to occur, say, once per shift, every N minutes or when triggered by a process event. Such open-loop models also may write results back to the plant’s real-time data systems. However, the results of the model are always evaluated by a person, who ultimately accepts or rejects any advice or data. Additional effort is needed to make these automated models even more robust, read in more real-time plant data and reconcile plant data (e.g., measured mass flows in and out of a unit that don’t balance). The final level of modeling in operations employs real-time closed-loop models, with their results implemented in an automated way to optimize processes. engineer knows how to overcome such problems. These systems require additional effort to make them For use in operations, the model must be tuned fully robust and safe. However, they promise even greater to match plant conditions and the particular calculabenefits, particularly where processes need to respond to tions being executed. For instance, the plant setup may change from day to day — with different product predictable variability (e.g., in feedstock characteristics). Chemical companies already are realizing significant grades being produced and individual units or conbenefits in plant operations from each of these approachtrollers switched on/off. The off-line process models es [6]. The experience reported by INEOS is instructive. must account for these specifics. It used a modeling approach to optimize heat exchanger In addition, because the simulations only are valid within a limited range of operating conditions this range monitoring and cleaning in its vacuum distillation units, saving more than $3 million dollars per unit per year [7]. must be strictly understood and enforced. The models will need to be made robust, so they always converge within the valid operating ranges. Model inputs (both FuTure direcTions those entered manually and those coming from realThe biggest challenges in integrated engineering are time data systems) must be kept within these ranges; along two fronts: this often is done by running the models through a • for collaboration between disciplines; and 21 chemicalprocessing.com JUne 2008 http://chemicalprocessing.com

Table of Contents Feed for the Digital Edition of Chemical Processing - June 2008

Chemical Processing - June 2008 Contents From the Editor ChemicalProcessing.com Field Notes In Process Energy Saver Compliance Advisor Achieve Model Operations Bolster Your Condition Monitoring Toolbox Particle Analysis Makes Solid Progress Improve Batch Reactor Temperature Control Improve Your Job Security Plant InSites Process Puzzler Equipment & Services Product Spotlight/Classifieds Ad Index End Point

Chemical Processing - June 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.