Document w84L7QXLJNgR9DG7GXOY79kd

r\ `^2* CHEMICALS n U) o INDUSTRIES INTEROFFICE / LAKE CHARLES TO FROM P/T Operating Personnel ArviUe Hoffpauir . DATE October 25, 1979 SUBJECT Startup of a Per-Tri reactor is critical from the standpoint of both safety and maximum tube life. Several things can occur if startups are rushed or for some other reasons done at other than a closely controlled condition. If organics Is introduced too fast, you not only greatly increase the velocity causing excess catalyst carryover, but also increase the chances of having liquid entrainment Into mixed feed headers and the heads on the tubes. If 03 is introduced too fast, you could have 02 break through into the dome, causing a fire. Also, the temperature increase should be where it is easily controlled with dowtherm pressure where drastic changes will not have to be made to the reactor. This new startup procedure should be followed closely and will work only with your cooperation. eg PER--TRI REACTOR STARTUP A. Organic vaporizer startup 1. Fill vaporizer with bleed valve open to insure all N2 is bled from system. Bring level to 10-20 percent. 2. Start force recirculation pump. Check seal flush to pump. 3- Line up steam trap to pad and isolate from condensate header. 4. Start 500-1000 Ibs/hr steam to vaporizer allowing it to heat up slowly. 5- When organics pressure reaches 25 ,psig, raise steam flow to 1260 lbs/hr. 6. When organics pressure reaches 35"^0 psig, introduce organics into reactor using below procedure. B. Reactor startup 1. Organic feed a. Slowly open organics valve to mixed feed header allowing vaporizer pressure to slowly drop to prevent shocking tubes and entraining liquid carryover from vaporizer. b. Slowly reduce Ns to mixed feed header. (Do not cut N3 completely out.) - c. Monitor bed temperatures closely. As soon as an Increase in tempera ture is noted, 03 should be Introduced into the reactor. d. Slowly increase steam flow until organics pressure lines out at 20-22 psig. At this time, N3 can be removed from mixed feed header. 2. Oxygen feed a. After introduction of organics feed and an increase in bed tempera ture is noted, 02 flow control valve should be open to introduce a minimum flow to the reactor. This is the critical stage of startup and should be monitored closely. b. Watch bed temperature and when the reaction has started, an increase in temperature will occur. At this time, raise 03 flow to the reactor to 7500 SCFH stepwise and cut back N3 proportionally. c. Watch bed temperature and when it reaches 700F, 03 flow and reactor pressure should be slowly increased to 15 lb. rates. This should be done slowly so that you are able to adjust dowtherm pressure to control bed temperature. d. When all W3 is out of the reactor, flush water should be taken off primary, brine put on secondary, degasser lined up to DH feed tank, and vents put forward. SL REACTOR PRE-STARTUP CHECK LIST Reactor Date 1. Reactor system leak check - Before charging catalyst. Put 35 psig nitrogen pressure on mixed feed headers, all organic feed lines, all O2 feed lines, reactor vapor lines, primary and secondary condensers and drain lines, and degasser by isolating system and using fluidization nitrogen. Tape flanges and use LEC-TEC. Depressurize when finished. Check/Done i tii 1. CI2 mix jet. ^^ ' 2. CI2 orifice plate flanges. 3. Mixed feed headers at bayonet flanges. 4. Organic feed lines spool flanges underneath reactor. 5. O2 feed lines spool flanges underneath reactor. Remove tape when finished. 6. Reactor head flanges (check by feeling for leaks with hands). _________ _ 7. Reactor head thermowell flanges. ____________ 8. Catalyst addition ram valve flanges. Remove tape when finished. 9. Catalyst blowdown ram valve flanges. Remove tape when finished. _____________ 10. Reactor vapor line flanges. Remove tape when finished. _ 11. Tant- tee by-pass flanges. 12. Tant. tee bottom spool flanges. 13. Degasser bottom spool flanges. 1A. Tant. tee inlet spools flanges. Remove tape when finished. _ 15. Others: 2. Vaporizer system leak check - Before or after charging catalyst. Leak check vaporizer by closing organic block valve to reactor, completely isolate vaporizer, and pressurize it to 35 psig with nitrogen. Tape all flanges and check with Lec-Tec. Depressurize when finished. 1. Float chamber flanges. 2. Sight glass flanges. 3. Uninsulated flanges on demistor block valves. '. Vaporizer manway. 5. Reboiler flanges. 6. F.C. 'pump suction ell and spool flanges. 1. Organic vapor line block valve flanges. 8. Organic flow orifice plate flanges. 9. Others: 3. Before catalyst is ---------------C---------- -- W--------------------- -J ----------------- steam on Centrifix, ram valves, bayonets, superheater, steam trac- ing, steam heaters. and steam supply up to vaporizer steam FCV. CONFIDENTIAL: Subject to Protective Order Of 14th Judrciai District Couj No. 91-1145 SL 019340 Check/Done 2- - A. This list is not to be taken as all inclusive. Therefore, make a walk-thru inspection of all systems and check that all piping flanges are made up, thermowells are in place, pressure gauges all in, and bleed valves closed. 5. Anytime before startup: Stroke all automatic valves from their controllers. 1- Both 1^2 fluidization MLS valves. 2. Cl2 FCV. 3. 02 FCV. A. Steam and organics to vaporizer. 5. Reactor PCV and vent MLS. 6. Watch all Dowtherm valves run thru -their motions. 6. After leak check is complete, fill vaporizer with 300-A00 gallons of organic feed. Then pad this out to the bottoms storage tank. After vaporizer is empty, refill it to operating level. 7. Be sure an empty drum is under the catalyst blowdown catch pot. C CONFIDENTIAL: Subject to Protective Order of 14th Judicial District Court No. 91-1145 SL 019341 RX MAJOR OVERHAUL INSULATION LIST 1 - BTM OF RX (FEED PIPING AND HEADS). 2- ORGANIC -CL2 MIXING TEE. 3- MIXED FEED HEADER BAYONETS. 4- BOTH REACTOR RUPTURE DISK. 5 - REACTOR OUTLET VAPOR LINE FLANGE (TOP OF RX). 6- REACTOR DOME THERMOWELL NOZZLES. 7- TOP OF CENTRIFIX. 8- CENTRIFIX. . 9- OUTLET OF CENTRIFIX. 10- INLET TO PRIMARY TEES. 11 - OTHER 1 - ___________________________________________________ 2- _____________________________________________________________________________________________________________________________ 3- ___________________________________________________ 4- ___________________________________________________ PERNELL 2-7-80 SL 019342 THE ABOVE SHOULD BE DONE PRIOR TO REACTOR START-UP. CONFIDENTIAL: LCP -161C Subject to Protective Order of 14th Judicial District Court No. 91-1145 Flanges to be Tightened after Reactor Startup 1. Both reactor rupture discs 2. Reactor outlet vapor line flange 3. Spool top of centrifix 4. Spool outlet of centrifix 5. Spools inlet to primary condenser tees. 6. Spools Btm of primary condenser tees 7. Other a. b. c. d. The above should be done prior to increasing reactor rates. % SL 019343 No- 91-1145 NORMAL PER-TRI REACTOR SHUTDOWN 1. The reactor rates must be lowered to 15 pslg using rate guide in SOP manual, 2. Stroke flush water MLS and solenoid valves. 3. Reactor vents lined up through startup header and degasser lined up to acid pit. 4. Brine to secondary condenser blocked. 5. Flush water put on primary condenser. 6. Nitrogen is slowly introduced into the "0" ring and oxygen feed slowly removed. When the nitrogen flow is at 9000 SCFH, oxygen should be blocked. 7. Nitrogen started into windbox. The flow should be brought up to 25,000 SCFH and the organics to the reactor blocked. 8. Both the above steps should be done by insuring extra nitrogen flow at all times to prevent defluidization. 9. After all feeds are blocked, call to have the pancake removed. 10. Pad out organic vaporizer to bottom storage tank. 11. After pancake is removed, open fluidization valve, close vent valves, and shut off flush water. 12. Block, tag, and drain reactor for necessary work to be done. ABNORMAL REACTOR SHUTDOWNS These shutdowns occur for various reasons and at these times the reactor will have to be put on nitrogen as quickly as possible. The same steps that are taken in a normal shutdown will have to be taken only much faster. Therefore, the possibility of defluidization of the reactor is greater. Each operator involved in reactor shutdowns should know exactly what must be done to safely secure the reactor during any upset condition. Study and practice all emergency procedures. _ CONFIDENTIAL: Subject to Proactive ,, Ot 14 tn Judicial District Court No. 91-1145 SL 019344 SL 019345 'X n 77'J 7 Tt'l\M ^>T tff'A. Taw Farm SfW fkt)6 ,-V "Uo'3 "3Wf| Cf/JFf).7'o P, , .Gaff(thrift n?y Preheater Catalyst D 0//.t ino o' ft ui* Xf ft o S o- C pj tQ ft c Q o z r`l H* ^ vc to *3 >- o Si (*' Cf >-3 U1 ttf h frt'* < >r* ** H- !5 nrt ort o/*. oili* W C)f. Cllt.Z' JI ( "A " ' iTZ mt(r. [> fa1tf till V~.~ r-* 'it 5 1\%- J._ A[fU #'3 R* ' u:s m\ I 5T1 Art UW-C 'IWftHjAM ^ */'"(ft .&fKliIAlt'V irsTKir .fMFtjWMrK Ih. PfCjU-A TQl "iiij 7$#" SrCVl'CnACtAsi S'MRWCmfi I Tfir flw/iX 4 [77^ | lllt'A^ i k <- t Per ,' i Xu* useju ~E7 tt-1 lai_ rh Toppiho S'tui. ,_vL X*X ___ t HrAViE i h-t i.u #ii1ij/i/jRT * Simmrrt V--. A Oi/C 30iF $6W| To one t Tf-t/m] fitfi'TdMS ItAVT. i tloTE : I SO^SrlSAM Ail StEAM Utility DftoPS XriTAAt To )d w r h cr'M 'ThrX SttsaM Txacinc For CIZ Line: Stbaa\ 7TiA&M&JkC/iTiApp. $Ce/irftTO MONITORING OF PER-TRI FEED QUALITY Since the quality of organic feed to Per-Tri is controlled by other operating units, it is critical that operators and lead operators closely monitor the sample analysis. The composition of the feed is the single most important factor in controlling ratios and conversions. The higher percentage of TCE in the feed the lighter the total feeds will be. The lighter the feeds, the lower the ratio. Since PPG does purchase feed material from outside sources, there is a percentage of the material that we don't know the composition of. But there are some bad actors that we do know about and guidelines have been set up on acceptable quantities. EDC Less than 3% Ferric Chloride (FeCl3) Less than 15 ppm Water Less than 50 ppm Hexachloroethane Less than 2% A high EDC concentration will cause increased cracking and burning in the reactor. Since burning (pyrolysis) causes a direct loss of carbon, it is very undesirable. Also, when the EDC increases, other light materials also increase. These react the same as EDC so that the problem compounds. Ferric chloride also increases cracking and burning at the higher pressures and temperatures. It also causes rapid fouling of the vaporizer reboiler. The 15 ppm limitation is because it concentrates in the vaporizer and causes fouling and plugging. , Water acts much the same as FeCl3 as far as cracking. It also causes severe corrosion. CONFIDENTIAL: Subject to Protective Order of 14th Judicial District Court SL 019346 No. 91-U45 Hexachloroethane causes plugging in the feed piping and head. When hexachlor condenses from the vapor state, it goes to a solid form. The only way to remove it is to dissolve it in liquid organics. CONFIDENTIAL: Subject, to Protective Order of 14th Judicial District Court No. 91-1145 SL 019347 9/17/76 Clary PER-TRI REACTOR CONTROL There are several different types of chemical reactions taking place simultaneously in the Per-Tri .reactors. In order to obtain the optimum yields and production rates, these reactions must be controlled. The intent of this scenario is to point out methods of desired reactor control. One way to view reactions in the Per-Tri reactors is in zones. Shown below is a graph of the many reaction zones. These zones deal mostly with O2 feed. l. Zc ne 7 - Denoted by reactor dome reactions or fires and possibly free CI2 in vent. l Zc ne 6 - Denoted by high HC1-TPD in reactor vent, high 7. burning, and dark crudes. i Zc ne 5 - Denoted by high HC1-TFD in reactor vent and high % burning. . <, Zone 4 - Optimum zone denoted by lowest HC1-TPD in reactor . vent possible with the 7. burning as low as possi| ble. 1 Zone 3 - Deacon reaction going but not optimum. Denoted by high HC1-TPD in reactor vent and low % : burning. i Zone 2 - pyrolysis nappenir.g. Denotea by high HC1-TPD in reactor vent, low % burning, and dark crudes. ; Zone 1 - Some deacon reaction going. Denoted by higr. j HC1-TPD in reactor vent and low 7. burning. !\ 0- increases Subject to Protective Order or 14th Judicial District Court No. 91-1145 SL 019348 2- - A reactor can run in any reaction zone except Zone 7. The reactor will not give optimum yields and produce maximum production rates in any zone other than Zone 4. Listed below are conditions, results, etc*, occurring in each reaction zone: Zone 1 This is the first reaction zone during and after starting up a reactor. The deacon reaction will be going some but not very much. There will be high HCl-TPD in the reactor vent and the 7. burning will be low. This would indicate a need to increase O2 feed flow. One word of caution... you should not make large 62 feed flow increases because you could put enough O2 feed into the reactor to go all the way to Zone 7 and not know it until there is a fire in the reactor. Zone 2 Here is where pyrolysis will occur. Pyrolysis (chemical decomposition caused by heat) will occur before the desired control points (the lowest " burning possible that gives the lowest HC1-TPD in the reactor vent possible) are reached. This means the deacon reaction is not going all the way. High HC1-TPD, low % burning, and dark crudes will show up in this zone. You have to add enough O2 feed to get past pyrolysis but do so in small increments while calculating 7. burning and HCl-TPD in the reactor vent. All 0? feed adjustments should be based on % burning and HCl-TPD in the reactor vent. Tee plugging is experienced while in Zone 2. L. one 3 At this point in the zoned reaction chain, you have the deacon reaction going but is not being driven to optimum. One nature of the deacon reaction is that the Oj fed will go to the deacon reaction first until the deacon reaction has been driven as far as possible without increasing the 7 burning. Hign HCl-TPD in the reactor vent and low 7. burning will be present in Zone 3. Dark crudes will disappear wner. you reach this zone. Again, more 0^ feed is needed. You need to add Or in very small increments (say 200-30C SCPH) at this point. .Again, all 0- feed changes must be based on calculated % burning anc calculated HCl-TPD m the reactor vent. Remember that you are very near optimum at this stage of the game. Zone + We have reached the top of the mountain. Zone 4 is the optimum zone. Ideal control is when there is enough CA feed to drive the deacon reaction to the desired point while maintaining the burning reaction at the lowest possible. This means the hold point on Og feed is when the HCl-TPD in the reactor CONFIDENTIAL: Subject to Protective Order 14th Judicial District Cour No. 91-1145 -3- vent is as low as possible with the % burning as low as possible. Past this point, the 7. burning and the HC1-TPD in the reactor vent will both increase. The actual control range, which is Zone 4, is not very wide. In comparison to most of the other reaction zones, it is very small. This means this is the hardest zone to stay in. This is why you need to make small O2 feed adjustments--200 SCFH of O2 addition or reduction can put you out of Zone 4. Zone 5 You just went past the optimum. Too much O2 feed caused you to go above __ the_ limits ..(opt imum_7o burn in g_ and_. HC lzTFD__in_th e_rea c 1o'r._ven t).___Calculate. 7. burning and HC1-TPD in the reactor vent. Study this data along with past data and decide on amount of O2 feed reduction--keep it small. Zone 6 You are headed toward an unsafe and uneconomical situation. 3urning has increased way too much. Any additional O2 feed will increase burning rapidly to the point where you have entered or could enter Zone 7. High 7= burning and high HC1-TPE in the reactor vent will show here. Back off Ch feed and calculate 7. burning and HC1-TPD in the reactor vent at the lower C>2 feed rate. Always keep O2 in reactor vent below 67.. Dark crudes will be present in Zone 6. Also tee plugging. Zone 7 The reactor will probably be on fire at this point. There is the possibility that free Cl will show in the reactor vent. Either lower O2 feed and get the fire out or shut the reactor down--you may not have a choice because the fire may have destroyed the TFE expansion joints on the primary con denser piping or you may blow a reactor rupture disc. All seven reaction zones have been defined but there are a few ether things to keen in mine:i. i. Study the composition 01 the organic teed. As the organic feeds heaw up. less Or teed is required. As the organic feeds lighter, uo, more 0- feed is recuired. Ir. order to control a reactor, you must look at the HC1 m the reactor vent in lbs/hour or tonsday--not in 7= 'Cl ir. tne reactor ver.t gas sample. 3. You must calculate T burning. CONFIDENTIAL: ofSl4thCT Frtective Order or 14th Judicial District Cou No. 91-1145 SL 019350 -44. Each reactor will have to be compared to itself--not to the other reactors. This is because of different conditions on each reactor such as catalyst level, catalyst condition, heat transfer, etc. 5. Catalyst level and condition of catalyst must be maintained properly in order to run a reactor properly. Attached are four graphs that show the results of O2 feed decreases and increases. Graph #1 points out what happens to the HC1-TPD in the reactor vent when you do not have enough O2 feed--Graph #2 shows how the % burning moves without enough C>2 feed. Both graphs point out an optimum range of 7, buming vs. HC1-TPD in the reactor vent. These two graphs are based on actual O2 feed decreases. Graph #3 points out what happens to the HC1-TPD in the reactor vent when you have too much O2 feed--Graph tl4 shows how the 7,, burning moves when too much O2 is fed. Again, both of these graphs show an optimum range of 7. burning vs. RC1-TPD in the reactor vent. These graphs are based on actual O9 feed increases. Both cases--not enough O7 feed or too much O2 feed--shcw there is an optimum point to run a Per-Tri reactor. In the past, we have not really tried to control the chemical reactions in the Per-Tri reactors. If we are to get the best from the reactors, we have to control them. Control means to calculate % burning and HCl-TPD in the reactor vent, compare these two controls with previous calculated data, and make the changes on feeds, etc. to hold the reactions at outimum levels. SL 019351 CONFIDENTIAL: Sublet to Protective Order of 14th Judicial District Court No. 91-1145 5 u RbJl AJ G- C4LC U CATfOAjS [SHORT COT -HeTHVP ) yfint/t&l .,.JJ e eoEV : ________ /o CO... Taj_ % cox iv vsvt utfJr VfipoxteEft Prep *A7e ($ p ~m BlOUjpoujU FtDuJ KATe fvpA ) A OF V^/Jt rcouj CcJ^aT tU^iA<r *) d O/ /o & U PJ*VJ Or - CONFIDENTIAL: Subject to Protective Order of 14th Judicial District Court No. 91-1145 '-o + /0coI)(% Ve>J ^r LOU J[0,0 V R FTtt>^)Cl 1*7 } (b LQuj DoujJO p 6CU/)| v lo 0 tXAyiPLE: Vc JJJ h^Oixj ^ 55/p %^ , Itg ya /p CO> - H% Peep B LbUJDOlV*J 3 -r J/*-- buRuiu^ - ~ ( 5 > s' ' C,Qt.^ //7 ) J - 3 t S' \4q1 *t - e, ?& % Hci IkJ\y exlt SL 019352 fr [/ HCI )(/ ve/Jr flolo) (o.oc z j - tv MC/ C. Vent and Crude Analysis The sample analysis is very important in reactor operations. By watching the different components in the crude and vent, the reactor can be operated at the optimum. This will vary depending on the com position of the reactor feeds. A typical range for the different com ponents are listed below. Crude Per + Tri (conversion) 89-92% TCE . 1% Lights 5% Heavies 6% Vent HC1 50% 0a 1-3% Na Depends on dead tubes CO 4-7% COa 30-40% D. Vaporizer Blowdowns A blowdown stream is taken from the bottom of the organic vaporizer that should be maintained at approximately 10% of the vaporizer feed. This stream is for two reasons, first to control the ferric chloride concentration and secondly to prevent heavy materials from concentrating in the vaporizer. Either of the above can cause plugging and fouling in the vaporizer and reboiler. CONFIDENTIAL: Subject to Protective Order of 14th Judicial District Court No. 91-1143 SL 019353 C'OWF' J, ATj : Subject to Protective Order of 14th Ju.dici<i 1 District Court Ro. 91-1X45 F. Reactor Catalyst Level The reactor was designed to maintain a certain velocity (.55 ft/ second) through the catalyst bed. If either the velocity or the amount of catalyst changes one or both of two things happen. First the reaction is not completed and there will be oxygen breakthrough in the reactor dome causing excess burning, and secondly the heat transfer area will be affected causing the vapor spaces to run hot. This can be monitored by watching the vapor space temperatures and by checking for dark crude. Reactor Dowtherm Levels A low Dowtherm level acts like a low catalyst level in that the vapor space temperatures will run hot and the crude will be dark. This is caused by loss of heat transfer area, and when the reactor vapors are not cooled properly, then burning increases with an increase in tempera ture. Feed Piping and Checking Individual Tubes on Reactor Total After the reactor has been dumped and the mixed feed header cleaned, the bayonets are reinstalled and a wash flange put on where the mixing tee normally goes. All organic valves to individual tubes should be closed except for tubes 1 and 3. A water hose is then attached to the wash flange and the mixed feed header filled. Then with water pressure on the feed piping to each tube is flushed. Nitrogen is then put on the "0" ring and each rotometer is checked to be sure it is clear and working and that the 02 piping to each tube is clear. To complete work on each individual gauge, valve and tubing on the individual nitrogen system is checked and any bad ones tagged. Sl- 019354 CHEMICALS INDUSTRIES INTEROFFICE / LAKE CHARLES TO Per-Tri Operators FROM Arviile Hoffpauir DATE SUBJECT January 19, 1981 Reactor SOP TUBE KILLING This writeup is to present a few ground rules on when to take a tube out of service. It also discusses our philosophy on tube deactivation. At all times we want the maximum production out of the reactors, but at the same time we want the equipment protected from damage. (The same holds for people, but this will only discuss equipment.) If at any time you feel that some piece of equipment is being damaged, immediately remove it from service. You are the people who will be most familiar with a certain problem and your judgment decision will be honored if you kill a tube. You will never be criticized for removing a tube, but if it is felt that the tube should have been left operating, we will discuss it with you so there will be no misunderstanding. A dead tube representa 6.3 TPD of lost production. A reactor shut down for overhaul represents 240 to 250 tons of lost production. If killing a tube prevents a reactor overhaul, that dead tube would have to be unproductive for 40 days to make up for the overhaul. Therefore, you can see that dead tubes, on a production basis, are better than reactor overhauls. There are many things that can cause a tube to act abnormally. I will present a few cases where a tube must be killed. There are other ways the tube can act, and in these situations you must use your own judgment on whether or not to let the tube operate. I will repeat again, you are operating the plant and you must make these decisions. The three most common things that can happen to a tube are: 1. Bad thermocouple, 2. Plugged mixed feed orifice, 3. Plugged windbcx. In case (1;, the temperature of the tube will gc anywhere or just follow the tube in front of it. The ring and windbox pressure will be normal and show good bed fluidization. Here, get the thermocouple replaced. Case (2) plugged nixed feed orifice - The temperature of the tube will fall, the oxygen ring pressure will be normal but the windbox pressure will be low. COSFIDE * 3i order CaC SL 019355 2- - The tube temperature drops because not enough organic is being fed to the tube and the oxygen is acting as a coolant. If you let this condition exist, a fire will develop in the dome of the reactor. The oxygen passes unreacted through the tube and builds up in the dome and vent system. High oxygen con centration in contact with reactor crude and catalyst fines is ideal for rapid oxidization. There is also a chance for an explosion in the dome under these conditions. The action you would take would be: 1. Replace 03 with N3 to the ring of that tube. 2. Replace mixed feed with Ns and close mixed feed valve to that tube. Add extra N3 and build up pressure between windbox and closed valve. Then rapidly open the mixed feed valve and allow N3 to flow backwards through valve to flush out the valve port. Repeat this procedure about three times. Then replace N3 with mixed feed. 3. Put one-half the normal oxygen back in the tube, leave one-half N3. 4. Check tube temperature. If it lines out below pack and mixed feed pressure is back to normal, remove all Na and bring 03 back to normal. Keep a close watch on this tube. If it goes back into the pack and both pressures are normal, the tube is all right to operate. 5. If both pressures are normal, but the tube temperature remains below the pack, kill that tube. Case (3) plugged windbox - A plugged windbox will make a tube run cold. A plugged windbox can be noted by high windbox pressure and low tube temperature. The usual procedure for attempting to unplug the windbox is to blast the windbox with nitrogen. If the windbox cannot be unplugged, the tube must be killed. If allowed to operate, a fire will develop in the reactor dome. The possibility of severe clinker formation is good as the C/02 ratio will drop and excessive burning will take place. This also holds fcr plugged mixed feed orifice. Unusual tube operation where a tube must be killed is as follows: 1. A tube will start going high and low temperature and oxygen adjustments do not affect the tube's temperature. 2. A tube cools off and oxygen increases do not heat up the tube. 3. A tube goes hot and oxygen cuts do not cool off the tube. 4. A tube cools off a few degrees and the oxygen ring pressure drops. 5. Any other condition where you think a tube needs to be removed from service. cosncEOTWw order set to Ptotecttvcourt SL 019356 -3- Normally, 1, 2, and 3 can be attributed to clinker formation. No. 4 would indicate the riser has severed, and the primary reaction zone has been displaced a few inches from the thermocouple. To sum it all - When in doubt about a tube, kill it. CONFIDENTIAL: Subiecfc to Protective order o 14th Judicial District No. 9I-U45 Iv SL 019357 TUBE OPERATION AND KILLING This writeup is to present a few ground rules on when to take a tube out of service, saving tube, and recovery of dead tubes. At all tines we want the maximum production out of the reactors, but at the same time we want the equipment protected from damage. The same holds for people. If at any time you feel that some piece of equipment is being damaged or people are in danger, immediately remove it from service. You are the people who will be most familiar with-a certain problem and your judgment decision will be honored if you kill a tube. You will never be criticized for removing a tube, especially if every attempt has been made to keep operat ing. If it is felt that the .tube should have been left in service, we will discuss it so that there will be no misunderstanding. An attempt to recover a tube is no indication that it should have been left on line. Conditions when tube is killed and when,trying to revive can be completely different. At maximum rates, a dead tube represents 6.6 TPD of lost production. You can see the importance of trying to keep a tube operating. The other side of the coin is a reactor outage costs 5.2 T/Hr of production. A complete overhaul takes at least 36 hours or a loss of 190 tons of production. A "mere rupture disc blowing will cause a minimum s'notdown of eight hours or 42 tons. A tube can be unproductive for a month before it makes up for a total reactor outage. If you consider that a few tubes can be replaced faster than all 19, you can see that a dead tube is a lot better than reactor outage due to damage. The following is an outline the rest of writeup: 1. Items to check which should indicate it tube must be killed immediate or attempt to save. 2. Possible things wrong and procedure for trying to keep operating. 3. Recovery procedure on dead tubes. CONFIDENT!/*?..: Subject to I'roteotivi- ..r Of 14th Judicial District Court Bo. 91-1145 gL 019358 We cannot possibly list every symptom problem or situation that may occur. You must use your own judgment, and, we repeat, you are operating the plant and you must make these decisions. To sum it all up, out goal is to try every method of keeping a tube operating if the situation allows. When in doubt about a tube--kill it. I- Checking Procedures on Errant Tubes The majority of the time the first indication of problems on a tube will be a change in temperature. This can be heating up or cooling off. You should form the habit of watching every tube temperature punch once. The possibility of a bad thermocouple exists and one tube fallowing another. This can cause serious problems which could be corrected before hand. Finding a tube heating up or cooling off early enhances the saving procedure. Other indications of possible problems may be change of windbox and O-ring pressure, and rotameter positions. When causes the above will be discussed in Section II. The checks below apply when a tube is gradually heating up and not when it jumps to 900 F in two punches, which calls for immediate deactiva tion, unless you know it is thermocouple. Usually, a tube does not cool off drastically enough to necessiate immediate killing, but caution should be exercised. A tube at 600 F or below can cause fire in dome if adding oxygen or just working with. A tube this cold should he killed. Checks: A. Check reactor vapor space temperature on L & W and vapor tempera ture to primary on IBM. Any sharp increase, kill bad tube. B. Check reactor pressure--any increase, kill bad tube- C. Check color of reactor crude--if black and shocking or cutting back on oxygen does not help, kill bad tube. 0. Check percent of oxyger. flow or. rotameter. CONf'IDRNTIM.: Subject to protective Order of 14th judicial District Court No. 91-U45 SL 019359 E. Replace thermocouple if checks do not show extreme conditions. F. Shock windbox and watch the rotameter bob. If it is driven down, we are losing fluidization. G. Shock O-ring. H. Check motion of rotameter bob and pressure gauges; if they are moving, the tube is still capable of being recovered. The reason for killing a tube should be stated in log book so it can be determined if tube can be tried with recovery procedure and how much effort should be made. Results of checks should be noted, i.e., rota meter driven down when shock windbox, changed thermocouple, etc. II. Causes of Bad Tubes, Indications, and Correction Tubes may have to be killed for mechanical reasons. (1) Thermowell leaks--this will be found when have to replace thermocouple or noticing organics dripping on pad under reactor. (2) Flange leaks on piping or heads. If we can not tighten up, the guilty tube must be killed. (3) Thermocouple failure--heat will cause failure of wire and you will get erratic temperatures or follow tube in front of it- The checks will lead you to bad thermocouple. Change and the tube should be operable without killing. Errant temperatures are basically due to improper C/02 ratio which may be due to the following. A. Plugged mixed feed orifice in the mixed, feed pipe. B. Plugged windbox. C. Improper oxygen flow. D. Loss of fluidization. (A) Plugged mixed feed orifice - This is a rare problem but can be caused by cracking and/or buildup on orifice plate. The temperature of the tune will fall, the oxygen ring pressure will be normal, but the windbox Sub'jet:T; ^a'w'ia. .-.*.\- uis*"1 I4tb Of 91145 SL 019360 pressure will be low. To safe the tube, you must shock with nitrogen through the individual feed line. If this does not improve situation immediately, kill the tube. (B) Plugged windbox - A plugged windbox will make a tube run cold and can be noted by a high windbox pressure. The tube temperature drops because not enough organics is being fed to the tube and the oxygen is acting as a coolant. If you let this condition exist, a fire will develop in the dome of the reactor. The oxygen passes through the tube and builds up in the dome. High oxygen concentration in contact with reactor crude and catalyst fines is ideal for burning. There is also a chance for an ex plosion in the dome. Notice that adding oxygen to a tube with plugged wind box can contribute to an already serious situation. The same thing is true with plugged orifice. Therefore, before adding oxygen be sure you do not have a plugged windbox and/or orifice. The usual procedure for attempting to unplug the windbox is to blast the windbox with nitrogen. If the windbox cannot be unplugged, the tube must be killed. The tube may be attempted to be revived later by recovery procedure. If allowed to operate, a fire may develop in the dome and possibility of severe clinker formation is good- (C) Improper oxygen flow -- The temperature may go up or down depen ing on whether it is too much or too little. With people ir. the area, it will happen that block valves be moved bv accident. A simple check of flow will correct this. Plugging of oxygen risers will be indicated by higher pressure low flow, and low temperatures. Shocking of oxygen risers may unplug. The s\- tube will cool off a few degrees and oxygen ring pressure falls off. The opposite on temperature can occur since there would be less pressure drop across ring and excess oxygen going to tube may cause burning and high temperature. Upon determining this as a cause of erratic operations, it is a must kill. Be sure to record that this is reason for killing as we do not want to try and recover. Plugging of O-ring will be indicated by higher pressures, low flow, and low temperatures. Shcoking of O-ring may unplug. CD) Loss of fluidization - Caused by windbox plugging, vaporizer boilup drop off, improper velocities, bad catalyst distribution. Tube will get warn depending on degree of fluidization. Check pressure gauges. What happens when shock windbox? Immediate shocking is proper corrective pro cedure for one tube. Lower pressure on reactor and increase boilup if this situation is affecting entire reactor. III. Recovery of Dead Tubes The attempts to recover a tube will probably be done on or after day shift after supervision has reviewed the characteristics of the tube, production demands, and maintenance required done. Instructions will be given when to do it. The following will be the procedure: A. Have thermocouples changed. 3. Have pressure gauges changed. C. Bring reactor to atmospheric rates. D. Shock dead rubes checking pressures to make sure you have fluidization. E. Add chlorine to reactor - at least 3000 SCFH total. Sub'jec 14 th DBUTIAh: Protective Order ,al District Court 91-1145 SL 019362 F. Replace nitrogen to windbox with mixed feed. The temperature should come up to 600F. A low temperature is a bad omen and should be con sidered when tube does not perform normally after oxygen is added and there is doubt as whether to kill or run. G- Add one-half normal oxygen, leave one-half Na H. As tube comes up to pack consistently, add oxygen and take out nitrogen. I. Observe closely. <***$?&* O* SL 019363 EMERGENCY PROCEDURES FOR POWER FAILURE Reactor Operator A. Switch control room to N2, open doors (if necessary). B. Immediately put all reactors on nitrogen. C. Secure reactors. Lead Operator (Subject to L.O. judgment and circumstances) A. Check well water booster pumps. If not on, block discharge valves and open up fire water. -- B. Block reactor feeds. - C. Check vent header rupture disc, If blown, isolate blown disc and insure the spare disc is unblocked. D. Secure plant. C Operator A. Block discharge of condensate feed pumps. B. Open bypass valves as necessary at metering stations. C. Block reactor feeds. D. Secure plant. Auxiliary Operator A. Heir lead operator get fire warer on to flush water system. B. Eel: as lead ooerator sees fit. ___________ I. Have pancakes removed or. all reactors. ___________ 2. Block 0- PCY. 3. Steam to 02 and creanic nreneaters. CONFIDENTIAL: Block recycle tank LCY. Subject to Court of 14th Judicial Uisct . 5. Block discharge valves or, CL i' tints. NO- "*- 6. Block steam PCY's if r.ecessarv SL 019364 CONDITIONS ANDJPROCEDURES IN EMERGENCIES INVOLVING UTILITY AND FEED LOSSES Unusual or unexpected operating problems in the Per-Tri Plant have developed in the past, and, of course, will occur in the future. Each emergency situation such as a loss of one of the utilities or reactor feeds allows us a better knowledge of operations through experience. To summarize some of the previous problems, and consider other possible emergencies. this report is being compiled. Prior thought to loss of feeds or utilities, or other adverse conditions may avoids major damages or a disaster within the plant. Additions and corrections to the following discussions by any of the Per--Tri personnel are encouraged. This information should be kept up to date and should be especially beneficial for personnel newly assigned to the Per-Tri Plant. Abstract Conditions and operating procedures are discussed in this report for the following utility and feed losses: 1. Loss of 02 2. Loss of N2 ' 3. Loss of both02 and N2 4. Loss of instrumentair 5. Loss of Cl2 6. Power loss to the Per-Tri Plant 7. Power loss to the organics area (includes air compressors) 8. Power loss to include Liquefaction and the alternate emergency lighting power source S. Loss of norm. and alternate emergency lighting power dicial District Court Ho. 91-1145 SL V " 10- Loss of 400 lb steam _ 11. Loss of well water Discussion 1. O2 Loss A loss of O2 has occurred on one occasion in the past and organic vapors backing into the 02 feed lines resulted in explosions which broke a number of rotameters on 3 of the 4 reactors. Plant B has 2 sources of 02 feed. Liquid Air is the primary source with approximately 125 lbs of pressure on their pipeline. Big Three Ind., the other source, has approximately 550 lbs of pressure on their pipeline. A set amount is flow controlled from Liquid Air and the remaining pressure or flow is made up by Big Three. In the event we lose Liquid Air oxygen flow we would have no trouble making up the difference with Big Three, but if Big Three were to lose flow, Liquid Air could not maintain enough flow to operate all the reactors requiring 02 in Plant B. Rates would have to be cut back to prevent shutdowns. When an 03 loss is indicated downstream of the Plant B metering station, open both N2 block valves and put U2 to the O-ring. Then begin shutting down the reactors. If the 0- reaches zero before N~ can be admitted, the 02 feed flow control valves should be closed to prevent organic vapors--03 mixtures from reacting in the feed lines. Particular attention should be given to ensure that proper fluidization (N2 flow) is attained during shutdown. - bV>ct to *,'rti'LiA^I'*r* tcOt`r'dceort SubJfCT;,,jViriaI lUSl U4b SL 019366 Emergency Procedures _ .. .Loss Upstream of Metering Station `a,-" If liquid Air 02" is lost, loss must be made up with Big Three pressure. b. If Big Three O2 is lost, loss must be compensated for by cutting back the rates on all area reactors using O2. \ Loss Dowtherm of Metering Station a. Admit N2 into the O2 feed header as soon as possible. b. Close the 02 automatic valve when the flow reaches zero. c. Shut down and fluidize reactors.- The 0 Contract First 120 tons from Liquid Air Next 57 tons from Big Three Next 60 tons from Liquid Air All the rest from Big Three 2. N2 Loss N2 entering Plant B is provided by two sources. Big Three is the primary cource with a pipeline pressure of 375 lbs. The second or supple mentary source is Liquid Air with a pipeline pressure of 125 lbs. Both pipelines are connected at the metering station where there is the ability to both regulate and monitor N2 flow. The N? contract First 70 tons from Big Three Nex.t Mj.2r0. .tons tromt L icu-zj c *Arr All the rest -from Big Three CONFIDENTIAL. Subiect to Protective Order n01f r1s4ctuh Judic..i.a..*l-* D istrict Court NO. 91-LL45 The N2 is then reduced and regulated at 96 lbs inside the plant. As it enters-Per--Tri, it devides into two streams at the rack, north of the control room. The smallest of the two lines is for field instrumentation SL 019367 only, while the large line provides Nj for reactor fluidization when needed. pads for tanks and condensors, utility drops, and for the Dowtherm pressure system. If N2 loss occurred upstream of the metering station by one of the suppliers, then the plant pressure would have to be maintained solely by the other. Both N2 suppliers have the pipeline capacity to supply N2 to Plant B under normal circumstances. In the event of a loss of N2~down stream of the metering station. nothing could be done except shut down and dump the reactors. All field instrumentation is on N2 and at present time not capable of being switched to instrument air. It is important that plant pressure should be maintained at 96 lbs at all times. Emergency Procedures Upstream of metering station (loss) a. Immediate action should be taken to place the surviving supplier on line totally. b. Isolation of defective pipeline. c. Notify both suppliers. Downstream of Metering Station (Less) a. Immediate shutdown b. Block both N2 sources c. Notify suppliers 3. Loss of N2 and 02 Total loss of N2 and 02 would cause a total shutdown of all reactors. This would cause ail reactors to be dumped also. Subject Of 14th J n2 red light PLANT AIR SWITCHING CONTROL ROOM INSTRUMENTS TO N2 4.XO PI -tST 5^ O 1. CLOSE VALVE 1. 2. OPEN VALVE 2. (YOU WILL GET RED LIGHT) 3. OPEN N2 VALVE ; CLOSE AIR VALVE 3. 4. CLOSE VALVE 4. 5. OPEN VALVE 5. I 6. OPEN CONTROL ROOM DOORS. t* l* Emergency Procedure Control Room - a. Close the 02 automatic valves. b. Close steam automatic valves c. Release reactor pressure Field ^ .. ` ' a. Block off organic vapor lines * from all vaporizers. b. Block steam valves to the organic vaporizers. 4. Loss of Instrument Air This will affect control room instrumentation and pressurization* The control room is pressurized with air to keep air that is contaminated in the atmosphere out pf the control room air. The control room is equipped with a switching station just outside the east door. At this station, the control room air can be switched to Na. The control room doors must be opened to protect from over exposure to N2. Emergency Procedure a. Switch from air to N: (outside east door) b. Block open control room doors. 5. Loss of Cls This is not a critical feed loss since the reactors are presently being operated without CI3. The potential hazard would be a leak or break to the atmosphere. In the event of a break or loss of line pressure, the CI2 auto feed flow control valves must be closed immediately to prevent organics backing up in the line. Then appropriate action would be to isolate the leak or break. Make appropriate feed cuts on reactors. COWF1 DENT1 AX* * Subject to niatric:t Court of 14th JuOiow _ Emergency Procedures ' - _ a. " Close CI2 automatic valves ` " b. Isolate leak or break c. Notify supplier (Liquefaction) d. Notify other on line users (EDC, Tetra, and Tri-Ethane II) . .......... 6* Loss of Power to the Per-Tri Plant (air supply not affected) The emergency lighting and instrument power system should switch automatically to the alternate power source from the cylinder loading building bus: control room lighting, plant instrumentation, and much of the process area lighting would remain in service. Critical flow losses would be cooling water, condensate feed for the steam drums, and all still flows. High pressure well H20 to the r \ reactor primary condensers is on emergency power, but an effort should be made to ensure that the power has been switched and the pump is on. Steam flows to the stills should be shut off at the panel board since pressures will quickly develop without condenser water. Except for the DH still, gradually decrease the steam flows to zero during an approximate two minute time interval. This will allow time for the safety relief valve to relieve the sudden pressure in the line which blew a steam line gasket curing a previous still shutdown. The steam drums will not boil dry for at least ;ive minutes at full rates. Therefore, aaecuate time should be available tor putting the reactors on Nj. If a steam drum would happen to boil or drain empty, safety valves an each reactor dowtherm vapor line should relieve the sudden pressure buildup. A. large C2 cut initially ------ CONFIDENTIAL: Subject to Protective Order Of 14th Judicial District Court No. 91-1X45 SL 019371 would lower reactor temperatures and allow a dowtherm pressure increase to slow boiling in the steam drum. The discharge valves on the condensate feed pumps should be closed to prevent the steam drums from draining back into the condensate storage tank. In the event of a temporary power failure, quick and immediateL action by the personnel turning on all pumps should ensure continuous operation of the plant. If performed quickly enough, all product should stay in spec. Emergency Procedures Control Room Field a. Shut off steam to the DH stills a. Close valves at the condensate pump discharge. b. Gradually shut off steam to the other stills during a time period of about 2 minutes. c. Immediately reduce rates to approximate atmospheric, or until N2 can be substituted for the flows. Make a large initial 02 cut and maintain a small flow with a high dowtherm pressure. b. Drain the DH still to the sewer to minimize reboiler corrosion. ftnr Ft e~ - 7. Power Loss to the Organics Area (includes air compressors) Emercencv Procedures Control Room Field a. Switch control room from air to N2 outside east door. a. Follow the above power loss procedure. b. Block open control room doors. c. Follow the above power loss procedures. CONFIDENTIAL: Subject to Protective Order of 14th Judicial District Court No. 91**1145 8. Power Loss (includes Liquefaction and the alternate emergency lighting power source) ... - _ .- In addition to the previously mentioned critical flow losses of cooling water, condensate to the steam drums, and all still flows, there would be no lighting, temperature readings, or still panel board controls, and the Cl3 flow would drop to zero* However, we would have use of all pneumatic instruments since all field instruments are on N3 and control room instruments can be switched toas discussed previously. Shut down. Emergency Procedures Control Room a. Switch control room to K2, open door. Field a. Close valves at the condensate pump discharge. b. Shut off steam to DH stills. \ . --"p ric b. Block off liquid Cl3 line to vaporizer. c. Dump reactor as soon as possible due to no flush water. d. Close Cl3 automatic valve. c. Drain the DH stills to the sewer. a _ - =rer* ^ 9. Loss of Normal and Alternate Emergency Lighting Power This is an unlikely problem but has happened in the past. The Per-Tri and heavies still temperature control valves would fail closed on a power loss to the board so it would necessitate a shut down. Both Tri stills and both Per still would have to be put on total reflux. The topping still would have to be shut down. In the case of the reactors if you would lose the well water booster pumps, you would have to shut down as soon as possible and dump the reactors. Without well water booster pumps you couldn't keep the primary condensers cool enough to fluidize the reactors. CONFIDENTIAL eet to Protective Order of I4th judic No. *91 lMtriCt Court SL 019373 10. . Loss of 400 lb Steam 400 lb steam. Powerhouse C sends P/T 175 lb steam. The 400 lb steam header is sent directly to the reactors as 400 lb steam, and the 400 lb header is also reduced to 250 lbs. The 175 lb header is reduced to 150 lbs. All headers are tied into each other. The 175 lb header can be equalized with the 250 lb and the reactors at reduced rates can run on 250 lb steam if time permits. If not, reactors will have to "be shut down and restarted. 11. Well Water Loss The main concern with a loss of well water is the cooling tower. Well water is used to makeup water to maintain a constant level. Fire water can be used for snort periods, if necessary to maintain a water level above the pump suction. In case of a well water loss put the stills on recycle because of the neutralization systems. Loss of well water would also mean that the flush water to the primary condensers would also be lost. In the event of a reactor shutdown, it would have to be dumped instead of fluidized. Conclusion and Recommendations Since an emergency shutdown is inevitable sometime in the future, we have at our disposal many means by which we can shut down without injury to personnel or to equipment. The quicker and safer the response to an emergency the more profits that are gained. It is recommended that all personnel be familiar and follow these procedures for safe and profitable operation of the Per-Tri unit. SL 019374 Subject to t of 14th Judici