สำนักราชบัณฑิตยสภา

«“√ “√ √“™∫— ≥±‘µ¬ ∂“π ªï ∑’Ë Ú˜ ©∫— ∫∑’Ë Û °.§.-°.¬. ÚıÙı Pramote Chaiyavech ˆˆı planned for the production of 315,000 tons per annum (tpa) of ethylene and 105,000 tpa of pro- pylene. Although an E/P cracker could have been chosen, inad- equate availability of propane ruled out this alternative. The pro- pane dehydrogenation route giv- ing a very high propylene yield was the only way to use the valuable propane for producing propylene. Two propane dehydrogenation processes were available for licens- ing at that time, both with no com- mercial plant experience. UOP Oleflex process was selected by NPC, and plant construction was completed at the end of 1989. The start-up of the NPC’s oleflex in early 1990 marked the first opera- tion of a commercial scale propane dehydrogenation unit in the world. The Operating Experience It is quite normal that, in op- erating a new process with a new design, one has to go through a learning curve to get to the desired target, solving any teething prob- lems along the way. Corrections and modifications are usually un- avoidable. This was particularly the case for NPC, which had no previous experience from other plants on which to rely. The chemical equation for propane dehydrogenation to pro- pylene looks quite simple. How- ever, to put it to work is no simple matter. The severe themodynamic equilibrium limitation of the reac- tion pushes the required operat- ing temperature to a high level coupled with the need for an ac- tive and selective catalyst. The high severity of the operation gives rise to several undersirable side re- actions, both thermal and catalytic, leading to several forms of coke de- posited on catalyst and equipment. Understanding these side reac- tions is of paramount importance for the success of the process. To take care of the coked catalyst, con- tinuous regeneration is required. The Oleflex unit thus consists of four radial flow moving bed reac- tors together with the UOP CCR system (Fig. 1). The process set- up is very similar to the world fa- mous UOP CCR Platforming pro- cess. However, the high-severity environment (reactor inlet tem- peratures 650 ˚ C and 1 kg/cm 2 g reactor outlet pressure) required gave rise to several new unique problems of their own, encompass- ingmechanical, metallurgical, pro- cess and operational aspects. Fortunately UOP is well equipped with a pool of highly knowledgeable experts well versed in all aspects of the technology, supported by modern research fa- cilities. The UOP support team for the Oleflex operation was of first class quality, particularly in the problem solving area. They worked hand in hand with NPC Technical and Engineering staff and jointly came up with the re- quired solutions in a speedy man- ner, thanks to the support of both UOP and NPC management as well as the construction contractor, without which the correction work could not be as effective. Table 1 lists the major opera- tional highlights of NPC’s Oleflex. Several incidents occurred during the first year of operation, requir- ing immediate correction. Later modifications are for process im- provements carried out at properly planned periods. Several lessons have been learned through these experiences. The more important ones will be discussed below. 1. Void blowing January 16, 1990 was the date of the initial start-up of the Oleflex unit and on- grade propylene was produced on January 19. However on Feb. 16, 1990 while the feed rate was 96 per cent of design, the temperature drop across Reactor 2 suddenly de- creased and it was not possible to transfer catalyst from Reactor 1 to Reactor 2. Pressure drop across Reactor 3 increased sharply, neces- sitating unit shut-down. Inspec- tion revealed a reactor screen de- fect allowing passage of catalyst into the feed inlet area during shutdown. At each re-start the feed pushed back the catalyst chips against the screen partially blind- ing the screen cumulatively. With less open area for the feed flow, the feed velocity through the catalyst area increased to a point that a void was created in the catalyst bed, blocking the flow of the catalyst. The high velocity attrited and turned catalyst into chips and fines and transported them to the next reactor. Figure 2 shows the cata- lyst flow through the internal parts