Simulation And Optimization Of The Polypropylene Process Based On Pesponse Sruface Method

The development status of ethylene industry is an important yardstick of the degree ofthe industrial modernization, and polypropylene (PP) industry is an important part of theethylene industry. With the development of economic, PP is becoming more and moreimportant in the national economy. For a long time, the PP equipment in China had beenrelying on imports and lacking in the indigenous innovation, recent years, this situation haschanged, but the technical level is still low. To achieve the indigenous innovation, one of themain content is the study of propylene polymerization mechanism, establish a mechanismmodel by high-precision chemical process simulation, solve the existing problems andbreakthrough the technology bottlenecks, and finally design and develop new processesbased on the mechanism model. In view of the above background, this paper use a responsesurface method that based on Box-Behnken design (BBD) to simulate and optimize theSpheripol PP process, research the process parameter’s impact on the target, such as PPquality, production, and so on.1. The GPC of PP was deconvoluted by Schulz-Flory most probable distribution, theMgCl2-Supported Ziegler-Natta Catalyst DQ-III had five active site, the deconvolution resultsis very good (R2=0.9999793) the Polymeric Uniform Ball (PUB) Model is applied tosimulate the catalyst particle, the polymerization occurring in each active site of the PMLMis described by the Monte-Carlo technique, is called the Uniform Ball Monte-Carlo Model(MLMCM), finally, the propylene polymerization kinetic constant, Mw and MWD of PPwas calculated by the UBMCM, the results showed that the UBMCM was appropriate forsimulate the polymerization kinetic constant, Mw and MWD.2. The simulation scheme were designed by response surface method that based on BBD,and simulated the propylene polymerization process by Aspen Plus which based on PC-SAFTequation of state respectively in accordance with the predetermined scheme, then analyzedthe result and got a multivariate quadratic response Regression Model, checked the adequacyand statistical significant of the model by a variety of mathematical methods, the importanceorder of the process conditions were described by Pareto chart. Within a certain range, theMw of PP increased with the increase of the polymerization temperature (T) and pressure (P),catalyst ([Cat]) and propylene ([Pr]) flow, decreased with the increase of H2concentration([H2]); the activity of the catalyst increased with the increase of T, P,[Pr], and [H2], decreased with the increase of [Cat]; PP yield increased with the increase of T, P and [Pr], decreasedwith the increase of [H2], with the increase of [Cat], PP yield increased to the maximumvalue and then decreased; pressure of the reactor increased with the increase of T and [Pr];liquid hydrogen concentration increased with the increase of P,[Pr] and [H2], decreasedwith the increase of T.3. The quadratic response Regression Model were solved by Newton iterative method,and got the optimized polymerization process conditions: T(201)=70.58°C, P(201)=3.38MPa, T(202)=69.98°C, P(202)=4.44MPa,[Pr,201]=20.39t/hr,[Pr,202]=4.35t/hr,[Cat]=0.69kg/hr,[H2]=3.86×10-4vol%, comparing the results with the plant data shown that, Mwof PP did not change, and PP yield increased by4.2%, activity of the catalyst increased by1.5%, and conversion of propylene increased by5.16%, the optimize effect was significant.4. The grade transition from grade G to grade T and the reverse transition were use theOvershoot-Quick Venting Strategy (OQV), the stabilization time of Mw and yield were lagsbehind the transforming operation, in the grade transition from grade G to T, the stabilizationtime of Mw and PP yield in R201and R202appear in3.06hr and2.5hr respectively, gradetransition time is2.5hr, in the grade transition from grade T to G, the stabilization time of Mwand PP yield in R201and R202appear in5.84hr and4.35hr respectively, grade transitiontime is4.35hr, the grade transition time of transition from low Mw PP to high Mw PP isshort its reverse transition. The transition material of grade G to grade T is29.74t, and thereverse transition is52.87t, the former is smaller than the latter.The quadratic response Regression Model can be used to: when PP quality unstable,Mw, MWD and MFI deviates from the set value, determine the problems and adjust basedon the3D response surface chart and regression equation, or set process conditions in thedevelopment of new grade PP.