Document zdv58EEkKV8Dq6wpYpw4mRZQ6

UK*. 1-1.1* V to: subject: INTEROFFICE COMMUNICATION Westlake, Lou i sana October 14, 1969 Mr. R. H. Gerlach Heavy Ends Reboiler Fouling As a result of an engineering study on the heavy ends reboiler fouling problem, the following conclusions and recommendations can be made. Due to pressure drop considerations, use of one reboiler is not giving the high flow rates required to keep deposits from collecting on the tubes. This is true whether the coking is forming due to a large film between tube and liquid or if the residue is formed strictly as an evaporative process. There are several alternatives which can be chosen which may remedy this problem. Ths most inexpensive method is to operate with two reboilers in service. This has been tried once before but the unit was not given a fair chance, as one reboiler had operated for seven days before the other reboiler was placed In service. A test could be run with two reboilers with a clean bundle standing by. This spare unit could be placed in service within 12 hours after one bundle is taken out of service. While operating with.two reboilers, the coefficient will approach 1/2 the value of one reboiler and this should be pointed out to the operators in advance so they are not unduly alarmed. The 14" liquid line feeding each reboiler is too small, particularly when one unit is in service. This was noticeable in July when the 14" gate was 3/4 open and the coefficient was low. By opening the valve completely, the flow increased and the level in the heavy ends column dropped rapidly and hence raised the coefficient. By Increasing the line size to 18" the pressure drop will decrease. The baffling in the reboilers is very intricate. There are eleven baffles per reboiler (sketch attached). By changing and eliminating many of the baffles, the net flow would increase and have a higher liquid to gas ratio. CCR 000032330 Mr. R. H. Gerlach Page 2 The net head available for flow through the reboilers is 19.4 feet which corresponds to 9 PSID. The design ratio for the reboilers is 20% vaporiza tion at a design flow 4rate of 4553 GPM, The design line loss for one reboiler in service is 22.5 feet at this design flow rate. Since the differential pressure is greater than design, the percent vaporization is increased. As this design data are based on clean service and assumes high column level, high flow rates are highly optimistic. The flow rate is less than design. The use of a filtration system in the feed line to the heavy ends column was ruled out since the heavy tare were not filtrated by 40 micron equivalent paper. Another alternate system would be a recirculation pump system to give forced convection to each reboiler. The pump system, while offering the greatest chance of success, is the most expensive. Also enclosed is a graph which could be used by Operations and/or Maintenance to anticipate reboiler cleaning. Additional uses may be found in furnace decoking, EDC vaporizers, etc. In addition to bringing a reboiler on "cold" as was done by Operations during the last startup, it is also recommended that the deposits of a "dirty" reboiler be sent to the offsites tanks rather than to the EDC tar stills where there is a possibility that the dirty material could be transferred to the clean reboiler. After review of this letter, a meeting should be held with all interested parties to discuss any future action. PMW-MC Attachment cc: LNV . JDM^ DKP RDG RLH P. M. Winegar OCR 000032331 from J D. MINOTT To. 9$t> ^ .19 tyd /(f*<# r/6,^b Mnul *- Q, S / "Sta/ u^M- ("*&*'* t.*i> '/A#* JL^U j+Mito1 = /.a^ n,1 6.^0 1 &*J- 3.3 0 (o.S*) (). - %ss/fb 2 ('l`IS?j K 6263 * tfXA&i&r* 5f f 4Jf-X /fit ^*^^1,, tt>W. ^ $*A*vhe>1 /. <5-*-C **0' V* 5f C 5* <W. d%* %** 0*k+tLz m/'0+* <S> CCft 000032332