466581 Troubleshooting Start-up of a Newly Built Intensified Heat Exchanger Reactor - Using Reaction Engineering and Dynamic Modeling to Solve Real-Life Problems

Tuesday, November 15, 2016: 4:43 PM
Franciscan A (Hilton San Francisco Union Square)
Eric Grolman, DSM Ahead B.V., DSM, Geleen, Netherlands

A large existing batch process was retrofitted with an intensified heat-exchanger loop reactor, in which most of the highly exothermic reaction is performed in a small reactor volume, at well controlled conditions. The change in design was initiated to alleviated serious safety and corrosion concerns and bring the production installation up to modern day safety standards. However, after the new heat exchanger loop reactor was constructed and initial hick-ups in flow control systems were solved, startups still stranded in runaway temperatures and emergency shutdowns. After 7 months of attempted start-ups and repairs, not even 5 minutes of steady-state production was achieved and plant operations were far removed from a solution to the problem.

A new team was formed to analyze the problems and deliver a final solution and investment scenario. In a day-long meeting under significant (time) pressure, available (transient) plant data were collected, as well as reaction kinetics relations obtained from earlier pilot-plant experiments. These data were used to construct a chemical engineering, dynamic model of the loop reactor. This model was used to analyze the problem and run through various scenarios that could explain the anomalous reactor behaviour. After running about a dozen scenarios, the apparent rate of reaction after scale-up appeared to be only 1/4 of the rate obtained from the pilot plant runs, while the heat exchanger appeared to deliver only 1/2 of the expected heat duty. The model predicted that plant start-up should be possible, but at only 1/3 of the design capacity. This was demonstrated in a successful trial run a few weeks later, in which the reactor produced on-spec product for the first time. To reach full design capacity, a change in hardware was still required. After 6 months the project closed, with the reactor up and running at full capacity.

The presentation shows the hardware changes implemented, the problem solving and modeling procedures used, and compares the modeled temperature and plant temperature profiles. A plausible explanation for the strange differences in apparent reaction kinetics before and after scale-up is proposed. Finally, the presentation demonstrates that scale-up can present really nasty surprises in actual practice, but also shows how a Chemical Engineering modeling approach can be used to solve issues and find minimum investment solutions, even after an installation is built.

Extended Abstract: File Not Uploaded
See more of this Session: Reaction Chemistry and Engineering II
See more of this Group/Topical: Catalysis and Reaction Engineering Division