(46h) High-Pressure Polymerization Process Technology: Modeling and Control of Polymeric Micro-Structure and Safety Considerations

The high-pressure process technology is extremely demanding with respect to invest, operation and safety considerations. Operation pressures up to 3000 bar and temperatures up to 300°C+ as well as production scales up to 450000 t per year and plant put special demands on material, operations and process safety. It is easily understood that at such operation conditions und scales explorations by trial and error are no longer an option. Even though one would accept the extraordinary costs, the safety aspect will tell to reject such ideas. Therefore, introducing new chemistry into processes such as the production of low-density polyethylene (LDPE) still the classical route is chosen with trials on mini-plants and on small and intermediate production scale. Examples for this will be presented at this meeting by Duchateau and Stimeier. However, understanding and improving the world-scale operation plants is also an issue. At this point modeling is an extremely helpful tool. It is this motivation why efforts have been taken in this field for quite some time and early on. Quite a remarkable level of quality in description has been reached up to now applying process modeling on the level of ODEs or Monte Carlo techniques to gather insight into the detail of polymeric micro-structure. In recent years a combination of modeling on the level of ODEs and Monte Carlo techniques together, so-called hybrid technology, has made it feasible to investigate complete production processes on a level of detail such as polymeric micro-structure, branching on the level of individual molecules together with their topology and incorporation of co-monomers on a molecular level. Reliable data for the kinetics and thermodynamics are indispensable for this as it will be discussed by Pflug at this meeting. All this helps building the bridge from control of micro-structure by process conditions towards application properties being characterized by rheology. Besides our group specifically the group of Kiparissides, Tobita and Reed generated remarkable progress in this field.

In the course of these developments elementary kinetics have been discussed with increasing detail. Subject of such discussions have been the propagation of terminal double bounds generating an excess of long-chain branches up to networks, the effect of molecular topology onto the resulting structures after β-scission or degradation by shear forces inside the reactor. The quality of description was excellent in any of the demonstrated case studies. It is for the nature of such projects that the application is often restricted to specific cases of processes or selection of cases. A question that remains open for some instances is whether the found solution is unambiguous and no other mechanisms could generate similar results. Objects of inspection have been in such cases the high molecular weight shoulder of molecular weight distributions produced in autoclave reactors featuring attractive properties for coating material when having the right shape. A wider view onto the application of models such as tubular or autoclave processes, varying process licenses and/or production facilities should help to better understand the quality of a model with respect to universal application. It is this criterion that brings a model closer to its final target to be useful as design tool for processes and products. For such a purpose our group applied systematically simulation models onto various process applications. Results and conclusions about the relevance of the individual elementary kinetic processes for the different production strategies will be presented. It is interesting to see how these findings can be transferred to processes close to the reactor wall such as the laminar layer of flow.

Beyond product design by modeling, it would be favorable for the application of models to process optimization being able to describe the upper decomposition limit of ethene high-pressure mixtures within such models. The decomposition is an ultra-fast process imposing temperature trajectories of more than 1000 K/s and pressure trajectories of way beyond 1000 bar/s. Besides the loss of production running into such decompositions means significant stress for the operation equipment that should be prevented. Pioneers for the incorporation of decomposition reactions into models have been Zhang and Ray. For the sake of simplicity, they separated polymerization and decomposition kinetics to have a clear non-coupled action of parameters within their model. However, inspecting the two independent kinetic schemes one finds similar species and it is easily understood that the kinetics must be coupled also burning the produced polymer into ashes as it is observed for decompositions in practice. This contribution will inspect how far progress in coupling the kinetic mechanisms can be achieved although access to experimental sensors is limited. The experimental counterpart for this study are decomposition test under well-defined laboratory conditions. Decomposition temperatures and pressures are matched acceptable so that a potential can be seen that in near future simulation models may inherently incorporate limiting process conditions. At the moment the state of the art is that model based process suggestions have to be checked manually against experimental data.

The setup for experimental decomposition test on laboratory scale is equipped with pressure and temperature sensors for data acquisition on μs base. This is sufficiently fast for pressure release experiments under well-defined laboratory conditions. Models may support analysis of pressure release experiments for a better understanding. First results will be shown for this.