Document G1JKn65qRJ8RojrZGEOYZYwV

PROCESS FOR THE K&HUPACTURE OF DIPHEHYL & SAHTOVAX Compiled by E. Mather, November 1930 Distribution Llsti lb. E. A. O'Neal, Jr., London IT. V* H. Garrett, London Er. V. D. Scott, London Mr. J. S. Brough, London Mr. D. V. Thrift, London Rnabon Research Files (Hr.y.H.Ritchle) Mr. 0. V. Taylor, Bewport (2) H*. P. J. 0. Haywood, Newport (Analytical section only of the report) M*. 0. B. Durgln, Anniston -ft*. A. K. EUehburg, Anniston Mr. J. V. Pennington of K.O.L., * (at Anniston, Hoveafoer 1950) St. Louis Library St. Louis K.C.L. file (E.Mather) WATER PCB-SD0000050648 IHTRODOCTORy MOTE 1 Preliminary information wu sent from Anniston to Hr. E. A. O'Beal, JT., in London, Jcne 1946. Mr. D. S. Bavsrcroft and the writer paid a brief visit to Anniston In September 1947 to collect Information needed for valuation and design work In London, and they produced the report "Kotos on the Process for the Manufacture of Diphenyl & Santowax at Anniston", September 1947. A set of Anniston drawings, out of over 350 drawings on file, was sent to London shortly afterwards, the selection being such as would give full constructional details relating to Anniston's preferred designs. Since then there has been a great deal of correspondence be tween M.C.L. and Anniston; Hr. H. L. Barden visited Anniston In August 1950 to review the process from the angle of pro duction work rather than design work, (see the relevant chapter of his "Report on a Visit to U.S.A.", August 1950), and the present writer visited Anniston November 6-10, 1950, mainly to deal with certain questions raised hy the M.C.L. mechanical engineers. This present report Is Intended to condense the information, collected as above, into a form convenient for use within Monsanto Chemicals Limited, particularly In production work. It Is hoped that it will be of use In the final stages of the construction of the Newport plant, and that It will form the basis of an operating manual for use at Newport, to be supple mented and kept tg> to date as new Inforation cooes forward. r DSW 462056 WATER PCB-SD0000050649 EARLIER REPORTS In a letter to Up. D. S. Havercroft, Sept.il/47, Mr. M. Buis listed the following six reporta which were already In London files: Demon Chenloal Industry Report Series Demen Production of Diphenyl, Terpheayls and Chlorinated Deri ratIres, by Wayne 2. Alexander, K.C.C., St.Louis, Oct.4.1946. Extract fro* Demon Report, C.I.O.S. XXH-14 pp.5^-57. Diphenyl. - Design Data 4th Diphenyl unit, D.A,Roper, with Study of Diphenyl Costs, E.H,Buford, 8.9,39. Final Report on Plant Tests on Ho .6 Diphenyl Utalt - Part 1. Job Ho.171-41 File H0.I38, 17th Sept. 1941. R.A.Ewlng. Design Data for npbesyl Still, D.A.Roper, Sept. 1939. Report HO .1924 Final Report on Separation of Diphenyl High Boiler Into fractions or known mixtures Part IH * Developing Plant Processes for Production of Santowaxes R.O.M.P. * . Job-Vo. 171-61, PHe Ho.2l8-B. Prepared by A.K.EUenburg. "Equlpnent Design for Santowex R Plant, May 20/43", Job 1-352-Report #31 by O.R.Shockley, accospanled this report. In Septeaber 1947, D. S. Bavercroft & E, Mather wrote a report "Hotes on the Process for the Manufacture of Diphenyl & Santowax at Anniston." On Sept,25/47 copy #11 of Anniston report: 1640, Final Report on Study of the Diphenyl Process Part 11." A Review of the Process. Job 171-41, File 138, F.2.Hubbard, Anniston, Oct.15/41# was handed to Mr. Bavercroft for London, also a copy of Addendum to 1712# Final Report on Study of Diphenyl Process, Part 111, Continuous Let-down Procedure, Job 171-41, File 138. A. M. Ellehburg, Anniston, Jhn.31/45* DSW 462057 WATER PCB-SD0000050650 Earlier Report*, oontd On Oct.13/*7, microfilm copies of the following two reports wore sent to Mr. V. M. Cooper, London: 1642. Final Report on Diphenyl Operations, J.E.Moose, October"19*1. ----------- 1712, Final Report on Study of Diphenyl Process, Part ill, Hubbard, TerSTk Miles, Anniston, 0ct,19/*2. On Mar.ll/*8, a copy of report 1937t "Highly Chlorinated Quaterphenyls" was sent from St.Louis to Ru&bon. Mr. EUenburg, January 20/*8, roriwed the following reports: 21* "Benzol Losses In- Waste Itydrogen from Diphenyl Plant", J.L.Howerton, Sept,22/33. **5 "Reooassendations for the Operation of the Crude Diphenyl Still", J.H.Carothers, June 23/33, 9*8 "Investigation of Benzol Losses --- at the Diphenyl Plant", Horton Harris, Aug.21/35, Mr. Ellenburg, Apr.28/*8, cited the following reports on the German tube process: (1) I.G .Farbenlndustrle at Leverkusen Germany, Ju1^25^*5^^A Visit by Turnock & Lowderailk, (2) B.I.O.S. Final Report 893* Item #1, (19*7) pay & Richards A copy of Anniston report 2211 "Interim Report on Alloy Test ing , 171-*29, V.W.Wade, Apr.22/48, discussing the corrosion erosion of the lead pots by the vapour blast, was sent to London (Hr. Barercroft), July l/*o. A microfilm copy of the Organic Division report "Diphenyl fro* Benzene", 117-881, Oct*5/35, on the effects of impurities In benzene, was sent to Hr. Thrift of M.G.L., Bay 9/*9. The process description In this report is said to have been based largely on that of the report "Design Data far *th Diphenyl Chit", by D.A.Roper. DSW 462055 WATER PCB-SD0000050651 ( Earlier Report*, contd 4 On Jan,lK/h% the Phosphate Division Engineering Department Report 117. "Pinal Design Data for 12-Uhlt Diphenyl Plant" was sent to London. The Phosphate Division Monthly Research Reports go to M.C.L. The London Central Engineering Design Data Books "Diphenyl Phase HID, Jan.49" and "Santowax R, Phase HI, Feb/49" have been read at Anniston, and coaanents offered. Anniston staff Intend to write up the Operating Procedures for their own use, but the process Is so well established at Anniston that this task has no great urgency. Anniston drawings which have been sent to M.C.L. at different tines are listed under "Pshawing List" on pages 77-Si later in this present report. DSW 462059 WATER PCB-SD0000050652 OPTLIHE OP PROCESS AS WORKED AT AHNI3T0K, 1950 5 Benzene vapour Is passed continuously through molten lead at a high teaperature, causing sons 16$& to 20$ of the benzene to be converted. In a single pass through 3 lead baths In series, into a mixture of diphenyl, terphenyl, etc, with the evolu tion of hydrogen. The resulting mixture of hydrogen with the reaction products and the unchanged benzene Is passed up a fractionating coluiav which has a condenser and reflux split box, so that the sensible heat of the vapours Is used to give a rough separation the organic components of the mixture. The fraotlonatlon system yields a forward flow of liquid ben zene, and an underflow of a mixture of benzene, diphenyl, terpheoyl, etc., while the hydrogen passes forward, saturated with benzene# This hydrogen stream is cooled and cospressed for the recovery of more benzene. The column underfleer is fractionally distilled batohwise at atmospheric pressure to yield distillates of benzene and of saleable diphenyl, and of course, intermediate fractions which are recycled to the still# The still bottoms, containing the higher boiling components of the mixture, are called crude Santowax. -- Some of the crude Santowax Is distilled "straight-over", under vacuum, to yield Santowax R, and some Is put to permanent storage because the supply exceeds the demand. Crude Santowax has also been distilled under vacuum, with fractionation, giving ortho terphenyl of about 94$ purity# Centrifuging of the distillate, or better, fractional distil la tloa, gives a better product# The m-p fraction from distillation, has been allowed to cry stallize In pans, slurried with water at about 85C, centri fugated and washed in the centrifuge with hot water. The centrifuge cake Is then about 95$ para terphenyl. The oil from the centrifuge liquors, separated and dried, is called "Santowax K". See Anniston report #1924, jan.8/45. See page 50 below for a note on the flaking of Santowaxes. All the recovered benzene is recycled and there Is a practi cally continuous make-up of fresh material. This Is believed to ensure the presence of a certain catalyst which is present In the raw benzene, but eliminated by the processing. (See under "Coasaents on Process", page 111 below). DSW 462060 WATER PCB-SD0000050653 NOTE ON CHOICE OP PROCESS See the Anniston Report 1640, Oot.15/^1. F.E.Hubbard, also 1ft*. A. K* Ellenburg's letter of April 2o/49* citing B.I.O.S. tc. reports on the "tube" process. The research departments at Ruabon and Anniston agree In be lieving the tube process say be superior to the lead pot process. Soas pilot plant work has been done In this direc tion at Anniston. See page ZL6 below for the choice of the asthod of recover ing benzene from the offgas. 6 DSW 462061 WATER_PCB-SD0000050654 OOTUETS FOR PROPPOTS 7 At Anniston boom of the diphenyl is flaked for sale, but by far the greater part Is used either at Anniston or at Monsanto, Illinois, for the Manufacture of Aroclors. Monsanto Chemicals Limited do not expect to sell diphenyl* A little also Is used at Monsanto, HI*, for the Manufacture of o. & p. amino diphenyls, . A mixture of crude Santowax and diphenyl nay be chlorinated to give the Aroolar 2565 used as a roofing oo^>ound* Santowax R Is used as the starting Material for Aroclors 5^2 & 5460, also for the hydrogenated product HB~40, which is used as an "extender" in plasticizers, for exai^le In Vinyl film. The Santowax R has some little application In high tem perature greases for valves. Monsanto Limited have given some consideration to the possi bility of usefully disposing of still residues. Dr. Scott, Jhnuary 29/^8, considered that the residues would not be suitable for making firelighters. Dp, Hardy, March 4/^8, passed on Mr. EUehburg's review of the possibility of chlor inating tbs residues to about 60-65# chlorine content to make a cable Joint-box composition. Mr. Ellenburg thinks that chlorination at 250-2o00. to 35*^0# would be about the best procedure, but warns us that the residues are not of constant composition. It was thought at one time that M.O.L. narkets would call for a higher ratio of high boilers to diphenyl than Is normally obtained In the Anniston process, and some thought was given to the possibility of shipping sons of the Anniston excess stock of high boilers to Britain, and calculations of the British laport duty were made. Kr. J. V. Head's note of March 2/k8, however. Indicates It Is now improbable that such shipments will be necessary. Reference Is made to a memoran dum by Br. V. D. Scott of Ruabon, dated June 20/V7, on the long term development of outlets for diphenyl. In the distillation of crude Santowax to give Santowax R, about 15 of the weight of the crude Is left as a residue, and some of this Is sold as "Hontar 9". Some of the hydrogen after cooling and compression, as out lined above. Is used for the hydrogenation of Santowax R to give HB-40. Any excess of hydrogen is vented to atmosphere after being eoopressed and cooled. At one time Anniston put the excess hydrogen into the fuel gas for the furnaces, but the consequent changes in the air requirement of the burners were troublesome. DSW 462062 WATER PCB-SD0000050655 r SCATS OF VORKim About 950,000# diphenyl per month, and 80,000# Santowax R per aonth. were produced at Anniston during the first nine months in September 1949, nine units were In operation, each rated at about 4000# diphenyl per day. In Vorestoer 1950, there were twelve taxlts, ntsabered 2-13# In existence, the original Ho.l haring been taken out. Xewpart needs were estimated by Br, Cooper, Oct. 1/47, to be 1,000,000 3b diphenyl & 240,000 lb Santowax per year. The annual sales of the ortho terphenyl hare averaged only around 4000 lb, and those of "Santowax H" around 10,000 lb. There has been very little sale for the para isomer. r \ DSW 462063 WATER_PCB-SD0000050656 PATKHT POSXgXOH S<w Axmleton Reeeerch Department, Report Ho. 1640, page 6. 9 r DSW 462064 WATER_PCB-SD0000050657 c PROPERTIES CP MATERIALS CONCERNED xo 1. Bengeae M.Vt. K.Pt. B*Ft< Specific Gravity Specific Heat* olid liquid Vapour Cp 78 a 8^1 0.883 to 0.886 15.5A5.50 w %3a S3 (I.O.I.) 0.463 '3P0g * Foraulao for Sp.Ht. at constant pressurex- Cp * 0258 (0,000875 x C) Perry 1st Edn. - 0.266 (0.0006732 x OC) Ind.Eng. Ch., July 1935. Latent Beat, B.t.U./lb: fusion evaporation ig:5 Flash Point F. 2 closed Cleveland 12 Explosive Units In air* Lover tapper % Auto-ignition tent>erature 1.4 8 lffr6P* DSW 462065 Vapour Pressurest Constants in V.Pr. Fornula Log10 Pr* Hg c , *B Kf Range C AB -58 to -30 -30 to * 5 .0 to 42 42.904 44.222 34.172 VM 7*9622 42 to 100 33*295 7*6546 See *1*0 "Design Data for 4th Diphenyl TJhit", Aug.9/39. WATER_PCB-SD0000050658 Properties of Materials Concerned, contd 1. Benzene (contd) Thermal Conductivity, (liquid ft gas) See sheet 3, Design Date fox* Units 4 ft 5$ D, A. Roper, Tor curves. 11 2. Diphenyl See "Description of Process used In Deot.246", Sept.27/46, lories, Sofrrenlco ft Bedew, pages 49 ft ol. See also "Design Data for 4th Diphenyl Chit7 Aug. 9/39, Specific Gravity: 0.992 75C/4C. The Swann Chemical Co. bulletin of 1930 gave thermal data on diphenyl. Viscosity of Diphenyl (fro* Anniston files) P Centlpolses 160 1.439 170 ISO 190 i:u4 200 1.034 220 0.901 250 0.75? 300 0.574 0.45? 0.374 0.313 O.283 Specific Beat (from Anniston files) r B.t.TJ.Ab 200 300 400 500 600 700 800 900 1000 0.416 0.470 0.547 0.616 0.658 ollli 0.690 0.696 DSW 462066 WATER PCB-SD0000050659 L pyppertlei of Hiterlftlt Oonoeroedi oontd .2 Diphenyl (oontd) 12 Another record give* the following values for liquid diphenyl: C Sp.Hoat cals./gran 100 150 200 250 300 350 400 450 0.36 0.380 0.412 0.460 0.518 0.575 0.610 0.625 V&ponr Cp 0.460 (D.A.Hopar, Design Data for Tftilts 4 & 5) Penalty of Liquid Diphenyl (fran Anniston files) C grans,/cc 75 0.9885 100 150 200 250 018470 300 0.8010 nz 450 0.6180 Heat of Vaporisation: 65.4 g.cal/g 258.5C. r- DSVJ A62067 WATER PCB-SD0000050660 Properties of Materials Concerned, eontd 33 3. Santoasuc R contains about l diphenyl ortho terphenyl nets terphenyl 2l para terphenyl 22 trlorthopbei^lezM#qu&te??heayl8 eto. (Anniston Pinal Beport 1942; Upper hold point Lower hold point 13T-1*3C* 45-bOOC. Changed ) 137-l45C Sept.23/47)-- 45-63C The density of the Santcwax nay he assumed to be the sane in both the crude and distilled fora. Viscosity also la the sane. Specific Gravity of Crude Santowax: l40C 175 215 300 1.026 1.000 0.98 0.964 (est. froaa curve of other points). Viscosity of Santowax B 155C 225 32.^ Saybolt Universal Seconds 28.5 " Molten Santow&x fleers about Ilia water at 25C. (A.K.Xllenburg's letter Jta.6/49). Specific Seats (estimated) Solid Liquid 0.4 0.5 latent Heats (estimated) Fusion 63 B.t.U.Ab Vaporisation 140 B.t.U.A^ Bolling Bangsi 38o-450C. Plash Point: 01. open ctg> 191C. sVJ Plane Pointt 01. open osp 238C. Another aceotat gives Plash Point 190Cj Plane point 2200. WATER PCB-SD0000050661 Properties of Materials Oweernad, eontd 3* Santouox R (coctd) Values ossuoed la the design of the Santowox still at Anniston* 8p.Hat - 0.5 Boat of Vaporisation - 140 B.t.TJ./lh Viscosity at 2500 - 5 centlpoises Viscosity at 350C - 2 oentlpaLoes * (, Xllenburg said these values were too high). 14 4. ferpheaylp Crude ortho (SantOMax 0, Crude neta (Sontovax M Crude para (Santoeox P, - p_ eutectic Ternary eutectic "Hold Point" B.Ft. Q 30 m -55JJ0 205C >0 nirture 24oC about 74C. See also Monsanto Technical Bulletin #P-*103, Jan. 4/47 issue, far the properties of the separated crude isomers sold. DSW 462069 WATER PCB-SD0000050662 ( HEAT BALAHCB 15 The boat balance Is discussed in tbo Anniston Design Data, 4th flhit, bjr D. A. Roper, Aug.9/39, pago 6. The heat of reaction of benzene to diphenyl la given aa 59.29 lbC per lb at T^00. . Trans .A*. Chen.Eag, V, 41, p.l44, (1938), gives 61 ibC per lb. This Is calculated to be less than a 3-pot unit. of the heat needed in os*1 WATER PCB-SD0000050663 SAFETY COBSHERATIOKS 16 (Proa Anniston report 1642 of October 1941, section "Health & Tire Hazards, ", and from discussions at Anniston In 1947 and 1950). Health It Is obviously advisable for the operators to minimise their contact with benzene vapours and with the vapour froa the hot lead, but there la no history of trouble froa either of these in the Diphenyl plant. Va are warned that lead vapours may be evolved during the burning out of lead traps and carbon traps. Hhen lower grade benzene was used for Baking diphenyl In 1930-31, the operators in the Aroclor department suffered from skin emotions (acne), thought possibly to be due to chlorine deri vatives froa styrene in the diphenyl# The disappearance of the trouble coincided with the change to better benzene. For a tine the Aroclor operators had to bathe on leaving work and a change of work clothing was provided, but this practice was discontinued. See notes on pages 20 & 82 below relating to the general venti lation of the building. Fire The conversion units contain hydrogen and organic vapours at high pressures and temperatures so any failure of gaskets or of the units themselves nay lead to the esoape of a cloud of inflammable vapour, near to the pot furnaces. The amount of organic material present in the conversion unit at any one moment la, however, quite email, and the flow of benzene to the units can be shut off from outside the building. m the early days at Anniston, fires ware not Infrequent, but there has been none of consequence in recent years. Fuphaals Is laid on good workmanship in the construction and maintenance of the units. There were twelve conversion units at Anniston in November 1950, with their coluxs and other auxiliary equipment, two of the units being In one room, four in another, and six In a third room. The three rooms are separated by fire walls with conBsunicating doors, some of them automatically closing, but held open by chains having a fusible link. There are two DSW 462071 WATER_PCB-SD0000050664 Safety ConIterations, contd 17 Fire, cootd diphenyl btills, each In its own room, these two rooms haring connecting doors Into one another, hut not into the conversion room. The still rooms contain the receivers for the finished diphenyl, the heads and tails fractions, and the still bottoms, hut the benzene receiver is in a pit, behind a fire wall out side the building* The crude diphenyl from the convertor columns flows to two receivers which are outdoors, alongside the benzene receiver* The main store tanks for fresh benzene are in a fenced area, well away fron any manufacturing buildings, and the benzene feed tank and props are In a snaller fenced area, somewhat nearer to the diphenyl plant. Benzene vessels are systematically grounded (earthed)* See the note on page 83 below on the objection to Internal steam colls in the benzene tanks. The department is well supplied with extinguishers, both CO2 and CClh types, and there is a permanently piped Foamlte connection to the sain benzene vessels. There is a drum of CCU* on a stillage on the ground floor of the conversion room for recharging extinguishers, and all the appliances are sub ject to routine Inspection. Cloths soaked In CCI4 are used to stop off pipe openings etc. during repair work and the units nay be filled with C02 during welding repairs. The convertor rooms are not treated as having a dangerous atmosphere. It is not at all unusual to have escaping vapours burning round the cover of the convertor pot (especially now some of the convertors are being run at over SO pslg). If vapour leaks from the pot flange, it may not ignite, especially if the seal between pot and furnace is good, but it is better to Ignite the vapour rather than to have diphenyl subliming about the bulldlng,and lodging on roof members, etc. Welding is done In the building, and of course, same of the vapour pipes are red hot. naked flames are used to deal with blocked pipes In the conversion system. The diphenyl still rooms do, of course, contain large amounts of eoa&ustlble material because of the hatch nature of the operation. The heating coil furnaces for the two stills are side by side in a room of their own, cut off from all other rooms. Those of the still receivers which are Indoors have vent pipes, carried out through the roof of the building, and the outdoor DSW 462072 WATER PCB-SD0000050665 Safety Considerations, contd 18 Flrt contd benzene receiver has ita vent pipe carried veil up Into the air. Each still body has a bursting disc In a pipe leading op through the roof, (see page 9o below) . V Amiston are a little restricted In doing welding repairs to their Santowax distillation systea because of its proximity to the SB-40 plant,which uses hydrogen under high pressure. The receivers for the hydrogen at each stage of cospresslon haveAl-Pb bursting discs. In pipes venting above the roof of the eoapreesor room. Care Is needed, of course. In restarting the gas-fired furnaces. There is a powerful draught tip the furnaces and the pilot Jots my easily be blown out. Mr* Ellenburg, Aug. 18/48, stated that Class I, Grovp D actors are specified for pinping diphenyl and Santowax. Be also stated that the bottom from the diphenyl stills. If discharged at arotnd 400C, nay take fire. At one tine It was the prac tice to purge the stills with CO2 before recharging, but this Is not now thought necessary, and all that Is done Is to allow the stills to cool to about 200C or lower before re charging. Vhen the crude diphenyl, hot fro* the store tanks. Is ptoped into the hot still, there my be a great rush of vapour through the condensing system. Since the Introduction of the restriction orifices after the convertor pots, these pots have been running at about 20 pslg, so that there Is a greater risk of blowing out the gasket on the convertor cover. Sons of the convertors have been fitted with a loose sheet iron ring fence about 6" deep, round the cover Joint, to deflect any possible burst of flam. 71re Fighting Equipment Mr. Barden, August 1950, reported ss follows: "Asple COo and foaa extinguishers are provided. Annis ton has two standard foaa extinguishers on each floor of the convertor section and several In the distilla tion section. On the around floor, available far use throughout the diphenyl aroolar shed was a large hand cart roan appliance. The benzene tanks are permanently piped, to top of eaoh tank and into the stap, for foaa Which Is supplied fraa a central point*. DSW 462073 WATER_PCB-SD0000050666 Safety Considerations, contd 19 Fire Fighting Equipment, contd Each convertor unit has about a 56 lb cylinder of COg per manently connected to a valve leading into the benzene feed line where it runs below the operating floor before rising to the top of the vapourIzer. These cylinders are supported against the stanchions on the ground floor* Ho provision is made for warning then. A drum of CCI4 Is kept cn a stillage on the ground floor for easy recharging of the extinguishers. There are numerous safety showers,, each with the usual yellow painted ring on the floor below then. Hydrogen The conversion units contain hydrogen, and so do the crude diphenyl store tanks, because they are vented into the hydro gen min. The vent plpss from the crude diphenyl tank to the hydrogen min are usually vexy hot. Air oust of course be purged from the systen before any hydrogen is taken to the cospressors, and usually OOg Is used for this purpose (400 cu.ft. for a convertor unit), also to replace hydrogen before a unit is opened for repairs. There is no bar* In allowing the CO2 to go through to the hydrogen conpressors, In fact there is the advantage of recovering Its benzene content. It is no longer thought necessary to do oxygen doterminations on the hydrogen stream. The crude diphenyl receivers are vented into the hydrogen collecting system, and there is a risk that If they are ptaped espty, the hydrogen may escape Into the diphenyl still system. Welfare As already mentioned, only ordinary common sense care is taken against poisoning by lead or benzene, and no special clothing is supplied beyond face masks and leather gloves for men dealing with molten lead. 3iaple "firescreens" of sheet iron on pipe frames are used to shield the man and the equipment during the collection of the sasples for the conversion test. See page 73 below. Hr. Burden reported, August 1950* "High temperatures are fairly general throughout the department particularly above the convertor. All floors are of the vertical strip lattice con struction to improve ventilation, and large portable D3W 462074 WATER PCB-SD0000050667 ( Safety Oonalderatlona, eontd 20 Welfare, eontd "men-cooling fans are used to blow up through these floors whenever a man has to work upstairs. Hormally all operations are controlled from the ground floor, all possible controls and Instruments being located there. The writer regards attentions to this detail as essential If conscientious main tenance of equipment is to be obtained." Repair and maintenance work on the conversion units at Annis ton is made more difficult by the radiation of heat from the surrounding units, especially In the room which eontalns six units. The lead traps, in particular, offer a big radiating surface. The diphenyl building with good ventilators In the roof, is very high, and the open-work steel floors permit a strong tpdravgfct of fresh air to flow through the rooms from the growl floor, which In every oase Is coop lately open on the east side. When molten lead Is being baled about on the operating floor, the ground floor below la roped off. See page 39 below. ( DSW 462075 WATER PCB-SD0000050668 c 21 SA|gT3 SSHT FROM ASHXSTOM TO MOMSAOTO CHEMICALS LIMITED Ths following eagles were sent from Anniston, Hoy.13/47: 0-4070 0-4071 0-4072 0-4073 0-4074 0-4075 Benzol Crude diphenyl Finished diphenyl Santovax R Still bottoms from diphenyl still Still bottoms from Santowax still (also 10 lb Montar Bo*9 requested June 22/48) but only tbs first tiro reached Ru&bon, so the last four wore repeated January 25/49, to London, (and ww received there Feb*2l/49). Tbs foilcaring saaplea also vers sent to M.C.L., on the dates statsd: Jtaae 22/48 Santovax B SoyA9 Benzene as used at Anniston* DSW 462076 WATER_PCB-SD0000050669 C COKROSIOM 22 Benxene and Diphenyl under the working conditions are not corrosive, but It has been found advisable to use stainless steel for the lead pots and connecting vapour pipes. Anniston Inter1* report 2211. dated April 22/48, describes pot failures tor "washing erosion" of the welds and the pot bottoms, failure of the welds, external erosion by lead, and by other causes, and also recommends the use of 310 stainless welding rods, or of 309 rods with subsequent annealing of the pots at 2050YF. Copy #11 of tills report was sent to Mr. D. 8. Eavereroft in London, Jtoly 1/48. See page 57 below for a discussion of the harmful effects which foreign elements In the lead night have on the stainless steel of the pots. In January 1949 It was reported the Anniston experience qp to th&t date showed 309 type stainless ateel to give better ser vice than 316 type. The Texas City Research Report Vo.52 "Special Report on Lead Pot Corrosion", E.Pltxgerald, Teb.17/50, (copy seen at Annis ton, Nov.yAjO), deals with the corrosion of lead pots used to erotic propane at 8OO0C to ethylene (formerly for pyrolysis of ethylbenzene to styrene). It reviews the Information avail able iq> to April 1947* Thermowells had suffered badly, and the pot bodies became pitted around the level of the lead surface and became eroded where there was turbulence of the lead. Laboratory testa showed that lead oxide was highly cor rosive to the stainless steel used. Samples of different stainless steels issuersed In the lead pots Sve inconclusive results, but Bastelloy and Baacrooe appeared have high resistance. Pure nickel, monel, and nlckelehroasliaB alloys with o0> nickel dissolved In the lead in a few hours. V.V.Ifede, Anniston, Apr.22/48, reported poor resistance with 68 nlokel - 18 chromium alloy. Carburatlon of the metal was suspected, with the formation of brittle coating easily removed by the poradlng action of the lead, A patent search revealed that the resistance of high chrome alloys was known. The report ended with a recommendation to run all the pots at a low rate to diminish turbulence rather than to run a few pots at high rate. It was reoosnended too to put carbon In the pots at the outset, to reduce lead oxides, but this recommendation has not been adapted at Anniston. OSW 462077 WATER PCB-SD0000050670 Corrosion, contd 23 Lead Is known to escape son >tlnes through the cover Joints on the pots at Anniston. It * y form lead oxide outside the pet, and this lead oxide aa;r attack the outer stirface of the pot. There is no point in trying to detect lead In the stack gases because there It always sooe lead lying about In the furnaces. DSW 462078 WATER_PCB-SD0000050671 GASKETS Anniston report 1642, Oct.15/41, recommends Oarlock 721 gas kets for tenperature3 up to 500C, and Goetz corrugated steel-asbestos gaskets above 500C. For the cover of the second preheater a 5/16" round soft copper gasket Is used with an asbestos gasket Inside It. This double gasket originated at Texas City, but the asbestos ring there Is outside the copper. For the cover of the convertor a gasket of 1" braided asbestos rope Is used, (see pages 42 fc oj below), but if the distortion of the flanges could be pre vented, the 5/l6n copper gasket would be better. Copper and its alloys hive given poor service In the hydrogen system, apparently because of HoS. Copper gaskets In the compressors have completely failed, and Babbitt metal has been proposed. DSW 462079 WATER_PCB-SD0000050672 PLOW SHEET The attached flow sheets of materials, pages 26 k 27, are taken from Anniston Research Department,. Report HO. 16*2, October 15# 19*11* The sane report glres the following data: - 0.1796 O.S.Osll benzene consumed per 1 lb distilled diphenyl produced (compare cost reoords glrea on pages 118- below, of the present report). The "shut-down* time on the conversion systems averaged 1.6 over a period of 8 months. See also page 28. 25 0SW 462080 WATER PCB-SD0000050673 Diphenyl Process, f-jo-tv Sheet 26 Ct, Mb Feed cut P*bo*#/m- */2 3o # 7~- <? 5 C .^;P - r ?7 VoL = th3 o u orrz i r>q t - P S er A1 /r.c % [-tier ^ -P` GO #/<P 3^0 O-Vj t Vapor Yc/^Z^O/o cA 0</ = lbJSft/%J (2 3*0 `CY fit . Frehiater -= 5<fo *C>/ejc/ t - ? 5?c. Fi ere = 3.0 ft CuHi - tt P= 4ttrc <; .fa 7 ~ 6~ f / Gt a s ec Atcj,i,!c/ P a 3oo*Vi ja ^ m j,rc iSi *>t " --< w / I I', a * 3g *//>, 1 rfc HC'tziq* - 50 ' C op =- / # = 472 0 cT" HQ- 7~r ti Ho -~ 7t Pb - P t4 Lfcc m I Son d fn Srr 2A<y Ar-V; eater T- -750eC/*i4 tr- ? Sec -'O ti OF CbH(, -ULifift- -f 17D P - J&cc> e/ HJB - Ho ~ P6 = 3-6oP >7or ?ft OP Carb o rj Cc nve rt o r 7 ~ 2 HceC /<v| If- - T3 SC' S ft/it 0, fh^t-z V f ^ ~7Xo C. DP = /5y 0 7/ 't r-5 /, CCor V :-t~u' e <0.-- ho! ^ /Sioa Cf H3 - ^`r-OM /Vo ^ 3-off Gxj-bo i r ? * Pb - r P - >4- 5' Jt f,n i 7' i/p/ CT- Co Hb - ??{ o ft dp -- / b3-0 If PB 4^.0 4 fix '- l o tt e//*t -- C ti' H/s: ' P ^ Carbcn. >' ? ft P- /v Pb - 7' g - 5 *s DSW 462081 WATER PCB-SD0000050674 Row SKga-t -pot- o{(f Ses f ~ To wj C o I V ;vy n _ 27 Hi. f A1 r*elt>h *-re~ OS\N 462082 _ /. (0-/4-o ' WATER_PCB-SD0000050675 COMPOaiTHW OP SOW XHTHHHEDXATB LIQDORS 1----wumi 1 .c *?a 3lc S2 Jf n a i A a a 1 o 8 HO JS .I iH 09 8 ro*- s2 A CJTO a ft a H2 <9 1 si w a & a &oo rt <cH a- in H H tty $ m a- i I .. ....... no 4o< O ft a. . IA 2 o ** Vi O ' A, cra\ o ? * a ! - H & 1 2 ilI_Upt4Jr=f(t _ j|I ,S taJ cB4-h> Xa-: * O 4WS* Ve4&| $ 2 a roic*l *(3* SS _ SHOO r&Ql<rC4Q 4CQQr<*1 28 a1 rio Hc a ! Js H ? imn orH > o, - a ri &9 H Orl -Arl*4c4Jo4Ow> H0\ 1 DSW 462083 WATER_PCB-SD0000050676 PROCESS IN DETAIL Benzene Handling Benzene la received at the plant in tank cars and la unloaded "over the top" by pump to 5 large horizontal storage tanka (#1) situated some 300* from the diphenyl building. Water la used to prime the pusp, and of course It must be separated afterwards by draining from the store tanks,or from the recycle tank. From the Storage Tanks the benzene Is punpod by (#2) to (#3) Recycle Feed Tank. Eero recycled material Is constantly made up with fresh benzene. This is necessary In view of the fact that a catalyst, apparently present In the fresh benzene. Is removed In one pass through the convertors. (See under "Comments on Process* pagee311 & 212 below). A float wlthiln the Recycle Feed Tajik automatically operates the pis^p, Item #2. There are also two floats within the tank for high and low level alarms. There Is a history of a little trouble due to the sticking of those various floats $ 3ee notes on Itea #3, page 84 below. The recycle feed tank now In use is a horizontal cylindrical vessel withbulged ends, and the boeizone outlet line to the pimps. Item #4, is taken from the bottom by a Jacketed line (jacket, however. Is not used), cemented through the containing wall which surrounds the tank. This take-off line rises a few inches Inside the tank to trap any water' and solid matter, and the tank has a sump, with a drain valve for tapping off the water, etc. Regular and frequent operation of this valve 13 a very iaportant part of the operation of the plant because a shot of water, pimped through to the vapouriser,may flash off almost explosively, and might break some Joint In the pressure system. Eater gets In from its use. In priming the overhead suction lines from tank cars, and may get In from perforation of the tubes of any of the condensers. The following streams of benzene enter the recycle feed tank: (1) A continuous stream of condensate from the convertor condensers. Item #11, (2) An Intermittent stream from the receivers. Item #27, of the hydrogen cooling and compression system. This stream corns at *bout hourly intervals, under hand control of the pucp. Item #28. (3) The benzene distillate from the benzene receiver, item #21, of the diphenyl still. This distillate comes for ward In batches at Anniston, via the pimp, item #2la, but Hr. EUenburg's letter, Jhn.20/48, indicated that a similar procedure at Newport might cause too groat a DSW 462084 WATER PCB-SD0000050677 Prop-- s In Detail, eontd Benzene Handling, eontd 30 variation in the ratio of fresh benzene to recycled benzene, because the flow of fresh benzene would be rery much smaller at Hevport than at Axmlstan, bat the diphenyl still batohes probably not much smeller. Nr. Slleaburg suggests interposing a buffer tank between the receiver, item #21, and the recycle feed tank, itea #3, so that the stream of recycled material from #21 can be smoothed oat. At Anniston, the operator waits until the float alarm on item #3 Is Vlnglng on lov^ before purging in benzene from the diphenyl distillation on the hydrogen compression, then he steps when the alarm "rings on high". There is a oase on record in whloh the recycle feed tank was overplayed, and the exoess benzene backed tp the distillate line and into the convertor reflux split boxes of the conver sion units, and esoaped down the eoltamts into the crude diphenyl store tank. Pile overplaying was detected by the fall in tem perature of the column underflow streams. There are two (2-stage Gould) (#t) centrifugal puzys for feed ing benzol at approximately 100# presatire from the recycle feed tank to the conversion units, and one or both may be in opera tion depending on the number of units operating at the time. There la a Fisher strainer before the ptays, and there are two filter pots, after the pimps, in the Control room - only one in operation at a time, to allow for olsanlng. These two filter pots are vertical cylindrical vessels about 6" I.D. by 9" deep with bolted-on covers, fit of course for working pressures around 100 pslg. They have baskets inside to support cotton filtercloth, but in Hoveofber/50 the doth had rotted and had not been replaced. It Is thought that glass doth might serve. There are Nordstrom 3-way cooks on the inlet and outlet lines for easy switching of the pots, and there is a pressure gauge on each line, to Indicate the filter resistance. The min purpose of filtration is to protect the flow controllers. Solid matter carrying forward to the vapotirlzer would be siiyly carried through in the vapour stream, and probably end in the lead trap or the carbon trap of the convertor. It is said that a little dust (from the pots) comes forward in the recycled ben zene, and there Is pipe scale from the benzene lines. DSW 462085 WATER PCB-SD0000050678 Process la Dotall, eontd Benzene Handling, eontd 31 The filter pots deliver to a header, serving a row of 13 Poxboro reeordlng flow controllers, one for each unit, plus one spare, the piping being so arranged that the spare controller can be switched In to replace any of the others. These con trollers are In a special room which.is entered iron outdoors and which has a coemmtesting door to the smln control room, otherwise it is completely shut off froa the rest of the plant. Apart froa the benzene flow controllers this roon contains only the meters for the fuel gas to the furnaces, (tnd a controller for water to tbs HC1 absorber for the Aroelor plant). Mr. J. C. Clegorn, the department Manager, mentioned that Rota meters were formerly used, set in a glass box draining to out doors, to take away any benzene leakage. There was a certain advantage In having the visible flow, and he would like sight boxes (fit for 120 pslg) in the present flow lines. The benzene delivery lines frost the controllers are cemented through the wall between this room and the convertor rooms, at about the level of the convertor pot lids, a separate line to each conversion unit. Each line dips to a stop valve within easy reach of the ground floor, inside the convertor room, and then rises to the top of the vapourlzer wit. Ifiasediately after the stop valve there is an inlet for C02 froa a cylinder chained to one of the stanchions of the building. No provision is made for warming the CO2 cylinders, though the ground floor, being completely open to the air on one side, is not very warm. The normal flow of benzene is 170 B5G per conversion unit per hour. DSW 462086 WATER_PCB-SD0000050679 Process In Detail, contd 32 Conversion There were 12 conversion units, Item #5. At Anniston In Koveaber 1950, two of then (unit3 4 & 5) being somewhat different from the rest In that they had a tubular preheater, then two lead pots, whereas the other ten units each had the preferred normal arrangement of three lead pots and no tubular pre heater. (See page 85 below for the reasons for choosing the three pot arrangement). Each of the 10 normal units consists of a (#6a) Vapourlrer, (#5A, 5B, fc 5C), three lead pots In series, (called respectively let preheater, 2nd preheater,and convertor), (#7), lead trap, (#8) fractionating coltsm, with (#11) condenser and (#12) reflux spilt box. The underflow from the column passes through Item (#9), carbon trap, and then by gravity through steamjacketed lines to the crude diphenyl receivers, item (#10). The vapourleer consists of a coll (formerly two colls In parallel) within a stack through which the exhaust gases from the pot furnaces are passed countercurrent to the benzene flow. There Is also (#6B), another stack for the purpose of by-passing the furnace gases. If necessary, away from the v&pourlzer coll. The temperature of the vapour leaving the v&pourlzer Is about 250C, but see below, pages 33*3^If the vapourlzer has two colls In parallel, an attempt must be made to share the flow between them by manipulation of the valves at the exits from the Individual colls. See the notes below, under item #6A, page 91, relating to the possible loss of benzene vapour by perforation of a vapourlzer coll. Each conversion pot is of stainless steel about 22-3A" I.D. by about 6*-0*deep, and contains molten lead. The following figures are given for the depth of load In each pot: Depth of Lead: Anniston report 1642, 0ct.l5/l, 1st Preheater pot 2nd Preheater pot Convertor pot Height of Distributor) Inlet above bottom )- of pot given as ) V-0" 4*-o" 3 *-6" 10" H.L.Barden, Aug/^0. "(confirmed Nov/bO) 2-l" dry dip 2*-6" dry dip 2*-11" dry dip out of 5*-11-3/8" Internal depth There Is little entrainment from the first preheater to the second, and not much from the second preheater to the convertor, but considerable entrainment from the convertor pot to the lead trap, so the depths of lead are liable to vary a little. If It la not desired to open a pot, the loss of lead may be made tq> with molten lead poured In through the neck which carries the thermo well. DSW 4G2087 WATER PCB-SD0000050680 CONNECTION FROM VAPOURIZER TO FIRST LEAD POT O 'O - 2 i lr\* S !I i?+-> V IfA if 'i\ A* WATER PCB-SD0000050681 Propose In Detail, contd 33 Conversion, contd Each pot cover carries (1) a dip pipe to carry the vapours down Into the Bolton lead, (2) a thermo well, and (3), a gas outlet* There are horizontal baffles or splash plates to disrtwtah the entrainment of lead. The first and second pre heaters also hare a connection to an equalizing line with a stop r&re. These equalizing lines connect Into the bottons of the converter columns, Item #8, and are normally closed. They provide a convenient location for pressure gauges (on branches below the stop valves), and could be used to by-pass a pot in emergency, and they can be used to diminish the chance of forcing lead back upstream during an Interruption of the benzene flow. The lead pots hang, hy their rlB8, In a row, each in a gasflred, fire-hrlak lined furnace, the first preheater being between the other two pots. Anniston report 1642 of Oct. 15/50 gives the following approxi mate figures for temperatures in the conversion system: Vapourleer (Bouble boll type) 1st Preheater Benzene in 25C Flue gas In 700C . out 340C (137 reported Aug/4l out 190C CoBbustlon gases Lead temperature Vhpour exit 650-700C 570-600oC (380C,AugA7). 550-580C 2nd Preheater Lead tefiparature 720-740C vapour exit o50-o70C Convertor Lead temp, controlled at 840 5C. (but see "Consents on Process",pages ill- below) Vhpour exit 710-720C. Coltam Top Bottom 80C - sometimes higher--ip to _97C 135-250C - normally about 200C The furnace casings had a temperature of about 100C as Judged by hand, Bovsaber 1950. DSW 462083 WATER PCB-SD0000050682 Process la Detail, oontd Conversion, ecntd 3* llr. Borden reported the following teaperatures in August 1950: Chit Xo. 6 7 8 9 10 11 12 Vaporiser 150-200 320 280 150 120 300 350 Xo.l Preheat 480 530 450 495 480 560 570 Xo.2 Preheat Converter 720 7*5 & 750 750 740 740 740 015 775 600 820 790 Top Coli-- 85 90 83 85 80 85 90 flue gas Inlet All eouples 0/S : -Sheaths burnt out Flue gas Outlet 175 235 322 318 230 190 250 Benzene flow 170 U.S.G. per hour 13 310 540 Sg 80 200 Surlier reoards are glren In a copy of the plant record sheet pegs 51 below. There la a Multipoint temperature recorder for each conversion unit and a recorder-controller far the conversion pot temperature, A s*u(>le strip of the Multipoint chart was sent to Mr,W.M.Cooper. 0ct.3lA7. The pressure drop through the double coll vapourlzer, at 170 USG bet--/boun was-. 8-10 psl. Pressure of vapour entering 1st preheater Pressure of vapour leaving 1st preheater Pressure of vapour leaving 2nd preheater Pressure of vapour leaving convertor 50-60 pslg 38-45 pslg 15-25 pslg 4-7 Plg The pressures vary a little as the depth of lead changes by eatralnnsnt of lead from the pot, and as carbon builds tp in the gas paaaage. The gauges waver a little because of the bubb ling and swinging of the load in the pots. Carbon builds tg> arotnd the baffles and In the pot exit lines during operation, causing a slow increase in back pressure. This Increases the retention tine of the vapours In the system, and causes the conversion reaction t> continue a little beyond the point thought beat, so the operator lovers his pot tempera tures a little to compensate. See Effect of variables, page 114 below. Recently, the back pressure on some of the units has been purposely increased by the Insertion of orifice plates in the lines from the convertors to the colons. In the hope of DSW 462090 WATER PCB-SD0000050683 Process in Detail, contd Conversion, contd 35 Attaining the same degree of reaction with usefully lower pot temperature** so that the pot* would last longer. (She vapour* on these units can be heard roaring into the lead trap). This Is experimental* and experlonce is showing that tbs Increased pressure on the convertor gasket nay cause troubles which outweigh the possible advantages. Mr. KLlenburg's letter of Dec 14/49 states that two orifice plates of stainless steel 1/B" thick are used* one with a 1.31" orifice between lead trap and vapour pipe and the other with a 1.14" orifice* between vapour pipe and oolxaxn. This arrangement gives the required pressure drop without using too snail an orifice. The pressure gauge on the vapour pipe thus reads the pressure between the two orifices. * Mr. Ellenburg gives the following comparison of conditions with and without the orifice platest Gauge Pressures Location of pressure gauge Mo.10 unit old conditions HO.12 unit with orifice plates After #1 preheater After #2 preheater After Convertor After 1st orifice plate After 2nd orifice plate (coluan) f P8l|S -- -- 3 46 pslg 33 21 13 3 Convertor temperatures in the "pressurized" pot were around 785-7900 for 17.5J6 conversion. The lower temperatures nay save a little fuel gas as well as lengthening the life of the pots* and In the early runs* the deposition of carbon in the gas passages appeared to have been greatly diminished. In Hoveatoer 1950 pressures tp to 20 pslg were obeerved after tbs first orifice plate. The following operating conditions were reported In September 19*17: *7he benzene props deliver at pressures of tp to 100 pslg* and the benzene reaches the first lead pot at about 50-60 pslg (fall in pressure through the con trol instruments) * the second pot at about 38-45 and the third pot at about 15-25 pslg* and leaves the third pot at about 4-7 pslg* the drop In pressure In each pot being due amInly to the hydrostatic head of aoltea lead." DSW 462091 WATER_PCB-SD0000050684 Procgg In Detail, contd Conversion, eontd H&tural gas Is need to boat tbs convertors and gtllls and there Is a separata gas mb tor for oaeb unit. Theee asters are Installed In the room mentioned above, page 31* which also eontains the benzene flow controllers. There Is a step cook on the gas min immediately Inside the room, with a long handle ter which It could be operated from outdoors. The min has a branch to each unit, each branch baring Its own step valve, then a pressure reducer. The line then branches again, one branch serrlng the burners on the #1 preheater, under purely manual control, and searing the pilot burners on all the unit8, and the other branch serrlng the control valves for Ho.2 preheater and the convertor, (lee page 107 below). The flue gases from the #2 preheater and from the convertor go downwards to cross flues leading to the bottom of the setting of #1 preheater, so that that unit gets a good deal of Its heat from the flue gases. The #1 preheater setting has gas burners of Its own to supply any additional heat needed and for use In starting up from cold. The tesper&turee of the lead In the two preheater pots are controlled by hand operation of the gas valves, but there Is a high tesperature alarm which would cut off the gas on #2 If the tMperature should run away, say, by leakage of benzene vapour Into the furnace. See also under Instrumentation, page 1CTT below. Typical operation records are given on page SL below. 36 OSW 462092 WATER_PCB-SD0000050685 Process la Dotall, contd 37 gets op Starting a. Conversion Phlt Kr, H. L. Barden, August 1950, reported &s follows: ?The start-up was witnessed of a diphenyl unit fro* cold; after cleaning and minor repairs* The following Instructions apply to an Anniston tait using natural gas* Other fuels say necessitate slight modifications. "Put on one turner of each unit* Throttle hack until warm enough to keep alight. The use of the pilot light will assist in keeping the burners alight. Flue gas is to go up the by pass stack to avoid burning-up the vaporiser coil. All vents on the convertor should be open in the early stages of warming up. As the setting warms ur the other burners are lit one by one, till all are alight and burning steadily. Burners are ad justed (by hand or automatically by controller) to obtain the desired teaperatures. Only the convertor temperature appears orltloal* When temperatures are approximately correct, start OOg through vaporiser, the preheat vessels and convertor to the ooIueb and out by the ooluan vent. Divert flue gas from by-pass to vapourleer to warm up the lattexy and close vents on the preheaters. Pressure on the preheaters should then build np and enable leaks to be discovered. Development of c.20 pslg on go. 1 and e. 10 pslg on Ho.2 takes about $ - & bottle of C0> and it Is assumed that the pressure build-up is a sufficient guarantee that the air has been adequately purged. There should of course be no pressure to speak of in the oonvertor if the column is properly vented. Then, provided teaperattxres (Inclu ding vaporiser) are about right, out off 002 out in benzol feed at a low rate (60 DBG per hour)* As soon as the lead - trap begins to smoke, indicating that benzol feed has reached it, dose the oolxsm vent, put water on the condenser and open valves to reflux split box, benzol run-back and biphenyl line to crude tank. The latter line mist be hot before start ing until sufficient feed of crude diphenyl is available to prevent the line from making q>* "Oautiously Increase the benzene feed by approximately 10 TJSQ per hour every half-hour or so* The governing factors in effectlxgthis Increase are maintenance of preheat temperature (which tends to fall as the benzene feed increases) and the avoidance of high convertor temperatures "The preheater must on no account be allowed to cool to anything approaohlng the melting point ef the lead. Maximum temperature of the oonvertor should be 830-8^0C (800C if operating under pressure)* The benzene feed should also be checked frequently vntil confidence is established because any mechanioal or electrical failure here will cause a stack back of lead into the vaporiser coil." qSVv 462093 WATER PCB-SD0000050686 Process in Detail, contd 38 goto on Stopping & Conversion tfalt It is of course important to prevent Ingress of air to the mlts and to take oare that the lead is not forced backwards through the system by allowing the pressure to fall at the Inlet end of the system. When the gas burners have been out off, the benzene feed Is stopped,' and CO2 Is forced Into the benzene Inlet line to the vapourizer until liquid ceases to condense in the convertor condenser. A full bottle (about 56 lb) of OO2 A* used. During tills period the uncondensed gas stream, consisting of hydrogen and soase CO2, will have been going forward to the hydrogen compression system, to prevent the loss of Its eontent of benzene, and the COg-hydrogen mix ture may be vented to waste after compression and oooling. When the distillation ceases, the system Is closed off fTotn the hydrogen collecting header, and from the crude diphenyl tank,' and the run-back line to the benzene recycle feed tank, and vented through the pipe which passes from the top of the reflux pllt box to a point above the roof. The conversion unit may then be opened up for Inspection and repair. DSW 462094 WATER PCB-SD0000050687 Process in Dotall, eontd 39 Kota on Emptying a Lead Pot ,, khen a pot baa to be emptied, the usual shut-down routlna Is applied, the system allowed to cool a little, the pipe joints and cover joint broken, the apace on the ground floor below roped off for safety, then the cover, with its dip-pipe and thermo well hoisted out by neons of a stirrup bolted to the central flange. The lead In the pot Is, of course, still mol ten, and a little drips off title dip-pipe. The cover Is laid on the floor In a convenient place. The surface of the lead In the pot la usually covered with floating solid debris, and thi is dredged out by naans of the long-handled perforated ladle (see sketch #1) and knocked out in the well of the box frame an the furnace cover. It Is convenient to remove the lead trap, to make room to work, and In most cases the trap ^ Will contain lead which baa to be melted out as described later, page 92 and will contain solid carbon. Iron oxide, etc. The lead molds are made of channel iron, about 2" deep overall and up to 36* long, 6" wide, with sloping ends welded In. Ten of these molds are then placed, five on each side, near the top of the pot, and two coloured men take turns to dip out the molten lead by means of the unperforated ladles, (see sketch 2) and pour it carefully Into each mold In turn until all are full. This Is rather hot work, especially In Summer, and the ladles of lead are rather heavy, but the job Is not too bad. The operator usee the edge of the pot as a fulcrum to lift the ladle, and steps backwards a little, dragging the hdle upwards until he oan get a sufficient purchase on the handle to lift the ladle over to the row of molds. He wears a face mask, and leather gloves, but no other protective clothing. The lead nay spit a little If there Is moisture In the molds, and casual passers-by have to take care not to step Into the molds. The ladle handle Is occasionally wiped with a wet rag to cool It. When ten molds are full, the two men take a breather for about 15 minutes vntll the lead has frozen over (tested by poking with a rod), than the molds are dragged a few yards away, and inverted to diap the ingots. Double handled tongs are used. The molds are then refilled In exactly the same way, and so on until the pot is practically espty, about 60 ingots in all. The pot is then hoisted out of the setting by means of a sling chain, bolted to the companion flange, care being taken that the bolts holding the halves of the flange are still effective, and the pot Is run along until It comes over the hinged part of the floor, so that It can bo lowered to the ground floor. Cre Is taken to close the hinged part immediately the operation Is ended. DSW 462095 WATER_PCB-SD0000050688 Handles of 1" piping 3/16" aetal, holes Sketch Ko.l OSW 462096 NH Sketch Ho.2 L-nj^ /.-TV WATER PCB-SD0000050689 c Process in Detail, eontd 4o Sot* on Pitying a Lead Pot, contd Sometimes a pot la mowed with the lead etill in It, for ample, If the distributor asseably will not break loose because of carbon incrustation, or if a sound pot is to be sored ft?on one setting to another, but such cases are quite rare. The lead would be allowed to solidify before the pot was sored. The usual procedure la to bale out the lead as described abore. The lead Ingots usually leare the nolds quite easily, and they are stacked near the tsiit until the new pot h in place, then they are packed, on end. In the pot, and Belted down, and any lead needed for nake-*o> is added, The.eorer with the dip pipe and thermo well can, of course, be lowered into position only while the lead is molten. DSW 462097 WATER_PCB-SD0000050690 Process la Detail, oontd Hoto on the Maintenance of the Lead Pots (See also tbs notes on item #5# pages below) 121 The practice sheets, pages 12<^ 8rt>elow, give some figures for the frequency of repairs* The first preheater with Its welded-on cover, low vapour velo cities sad relatively low temperature can run for years without attention* The second preheater, and sore particularly, the convertor, suffer fro* deposition of carbon in the vapour passages of the cover, daimge to the dip pipe, thermowell, and baffles by high temperature and by the mechanical and possibly corrosive effect of the lead, failure of the pot wall, usually by corrosion-erosion from Inside, at or below the level of the lead, cracking of the weld between the pot and Its ri* (this weld oarrles the weight of the pot and contents and there is continued gentle buaplng of the lead and vibration of the pot), failure of the cover gasket or warping of the cover flanges, and fro* entrainment of lead in the vapour stream, if lead comes in contact with the outside of the pot, fro* a gasket leak or from, a perforation of the pot wall. It nay cause narked erosion, apparently from the solvent effect of lead oxide on the stainless steel, but this Is not a major cause of failure. If there Is a partial blockage In the vapour outlet. It nay show up as a darker patch on the glowing pipe* The pipe leav ing the convertor needs cleaning about each 3 weeks and that front the seeond preheater about once every 8 weeks. Carbon may also build <g> Inside the cover, round the vapour exit, or In the orifices of the baffle plates around the distributor. The baffle plates slowly sag, apparently purely from high tenperature* A convertor pot usually runs 5 to 9 weeks before needing overhaul* The carbon may be burnt out of a pipe by heating with an oxy- acetylene torch, then turning off the acetylene, but care oust be taken not to overheat the metal when the carbon burns In the oxygen stream, and the welder should take care not to breathe too much of the vapour* There has been sons discussion of possible overheating of the pot walls above the level of the lead, but apparently the splash ing of the lead on to the walls does a little cooling, and of course the furnace draught is downwards In the cases of the second preheater and the convertor. Nr. Burden, August 1930, suggested to Anniston that a stream of COg passed through the hot convertor, might remove the carbon deposit by converting It to carbon monoxide. Anniston have DSW 462098 WATER PCB-SD0000050691 Process In Dotail, contd Koto on the Maintenance of the Lead Pots, contd 42 ado a few trials of this, tor passing two cylinders full of COp through a unit once a week, and though the unit finally choked with carbon. It was hollered that the choking had been delayed by the use of CO2. Although the pressure Inside the convertor la relatively low, the Joint of the convertor over gives a great deal of trouble, because the temperature Is near the yield point of the aetal. The edges of the cover gradually bend downwards and the edges of the companion (backing) flange under the pot lid bend upwards until they aeet, and the gasket no longer holds. The gaskets on the convertors are nade UP from round braided asbes tos rope 1" In diameter. See page 87 below. Leakage of lead from a defective pot will probably be detected first as a drip of lead from the furnace casing. (Hote that It ay drip from the wrong casing by penetration through cracks in the furnace setting, and care should be taken to examine the undersides of all three pots before deciding which actually Is leaking). The pressure gauges on the pots may help to indi cate which pot Is losing Its lead. Leakage of vapour into the furnaces may show up as black smoke In the stack, or as an Increase In furnace tenperature. See also the stack test dis cussed under Item #6A, page 91 below. If a pot Is sound In some parts and defective In others. It may be made good by welding, for exax^le, a new bottom may be welded In. A pot which has seen service In the #3 position say be shifted to the lighter duty of #1 position and so on. See notes under "Corrosion", page 22 above, for an account of the corrosion of the pots. DSW 462099 WAfER_PCB-SD0000050692 Process In Detail, oontd ^3 . The Mixture of vapours and hydrogen leaving the convertor pot passes through the centrifugal lead trap, item #7, where most of the entrained lead separates, and Into the bottom of the fractionating coltm, Item #8, The issderflow of the colusn leaves via f#9) a carbon trap, through steam Jacketed lines to one of (#10) two snap tanks. These tanks are vented to the "off-gas" line from the top of the colums. At the top of the coluan there is a (#11) Condenser and (#12) reflux split box, with an equalising vapour Una between the two. The reflux split box Is adjusted to give a benzene forward flow containing at the aost a few parts percent of diphenyl, and to give an underflow of crude diphenyl containing as low a ben zene content as can be attained. The Anniston report 16K2, Oct.lsAl, states that the crude diphenyl flow contains about A to 8# of benzene. Diphenyl still records seen in November 1950 Indicated that about 17# of the crude diphenyl was recover ed as benzene fraction. See page 53 below. Mr. A. H. EUehburg, Nov. 12/^7, wrote: " The reflux flow is used to control the temperature of the underflow from the column at about 200<>C. The reflux ratio varies according to the tesperature of the condensed benzene. --.-----(which) varies according to the length of time the condenser has been In service. As the film builds up on the water side of the condenser, the cooling efficiency drops off, and more of the hot reflux must be used to obtain the same amount of cooling. These ratios vary from 70/30 to 90/10-----", The vapour line Trcm the convertor to the bottom of the oolxnmt has a branch far sailing the vapours for the purpose of making the conversion test, (see below under "Conversion Test", page 75,) and another branch carrying a pressure gauge. The conversion test Is performed once a day on each unit, usually in the somlig shift. DSW 462100 WATER PCB-SD0000050693 Process la Detail, contd Treatment of Off-gat* The off-gas fro* tbe convertor colunns, consists of hydrogen, saturated with benzene, and It goes rla (#25), a finned cooler with water Jacket, to (#26) a surge tank, from which It goes to either of two compressors (#29) The coapressors work In two stages, 60 p&lg and 300 paig., aad the ecsg>ressed gas from the first stage goes downwards through a vertical interstags cooler (#30) to a low pressure tank (#31) where condensed benzene is deposited, then the hydro gen goes on to the second stage of compression. After the second stage tbe gas goes downwards through the second cooler (#32) Into the high pressure tank (#33\ where it deposits more benzene. The surge tank (#26) has a 55" water seal to a hooded vent pipe through the roof of the compressor room, and there is an auto matic control instrument which maintains the pressure In the surge tank between 11 and 16 or. per sq.in, above atmospheric, by reacting on the unloading valves of the compressors. The benzene which collects in the surge tank gravitates to the benzene receivers (#27) and that which collected in the pressure was formerly discharged by float traps, also into (#27), but the trap seatlngs failed by corrosion, see page 1 03 below, and the traasferrenee is now done at about hourly intervals by means of hand-operated valves. The pressure vessels have gauge glasses, aad the discharge lines have sight glasses. Corrosion of copper aad alleys in the hydrogen expression system is rather amrked aad is attributed to the effect of IMS. See page 101 below. Oxygen determinations were made on the hydrogen at first, but are no longer thought necessary. After any shut-down, the hydrogen is vented to atmosphere and the system purged with *00 cu.ft, of GOg. At Plant "B* therm is a hot wire combustible gas recorder per manently Installed In the hydrogen stream going from the chlorine cells to ecspressors for the hydrogenation plant, wired up to stop the cospressors if more than a trace of oxygen is present in the hydrogen stream, Mr. menburg, Jtoe 30/*3, wrotet dsw462ioi *--------Am you know, our Worthington compressor has two stages aad operates at 150 palg and 300 pslg at tbs dis charge of the low and high stage respectively. These pressures are not critical aad the choice of pressures depends upon the size of the cylinder chamber as well as the amount of gas to be handled-------", See under item #29 below (page 101) for his further remarks. WATER_PCB-SD0000050694 Process In Detail, contd ^5 Diphenyl DigtIllation Crude diphenyl is pushed through steam traced lines from the stop tanks to either of two horizontal diphenyl batch stills (#13). Each still Is fitted with (#13B), an external g&sheated coil, on the side of a fire wall remote from the still, through which the charge Is circulated by means of (#12A), a submerged centrifugal ptnp. The "spread or tern- perature difference between Inlet and cutlet of the coll Is automatically controlled through the gas supply to the burners, and Is automatically recorded. The controller la set for a spread of 20C, but when seen In Korenber 1950 It was running at about 160 to 18 spread, l.e. the heat Input of the burners was falling short of the control limit. See under "instrumen tation", page 108 below, for details of the control Instruments. A typloal still charge (Sept/^7) was 3*00 DSC erode diphenyl plus 450 DSC reeyled "heads & tails" fraction, and the still cycle took about 18 hours. The vapours pass to (#1*) a 20 plate colum with (#15) conden ser, (#16) reflux box working at atmospheric pressure. Cuts are taken based on F.P. determination, (but sea Kr. Barden's consents below). The reflux Is changed during the distillation and the follow ing Is typical of operation (In September/^"): At Start Benzene Fraction (ig> to 80C at top of colum) Heads Fraction Diphenyl Fraction Tails Fraction Reflux 100 25 50 30 20 / Product 0 75 50 Mr. Harden, August 1950, however, reported: "The total reflux period seams to be practically- confined to the still charging operation. The still and column are usually still hot from the preceding batch and the Incoming charge atarts distilling Immediately. Total reflux operation at this stage rapidly cools the column. It cam happen that this initial distillation la too rapid for the split box to deal with It. m such a o&se It Is usual to roller# the box by sending a portion to the heads and tails receiver for return to the still. It Is essential to avoid sending any of this initial dis tillate to the benzeztorecovery system since it is liable to be contaminated with diphenyl and * leed to excess production of high boilers. The charge used at the time of the visit was 3800 DSO af crude diphenyl and heads and DSW 462102 WATER_PCB-SD0000050695 Process In Detail, cootd Diphenyl Distillation, contd 46 "tails from previous run (about 4,200 USO total). "Soms excess pressure is apt to develop at start of distillation, presumably by plugsing of condenser by condensate. If this is observed, a rent is available for temporary relief. This pressure build-\q> was not apparent to the writer but the operator claimed that he could tell by the flow in the split box. "Circulation through the ooll is started before light ing burners. As the systeei Is usually hot. the burners oan be put full on quite aulckly. The burners are controlled by the "spread" between inlet and outlet temperatures. This is normally set to 20C. but is re duced to 10-15C when distilling heads to minimise carry over of good diphenyl. ' "Officially, the change-over from heads to good diphenyl should be based on Cr.Pt. (68.4C). In practice it is usual to change-over arbitrarily after one hour's dis tillation because the time spent in testing saaples means unnecessary diversion of product to heads for redistillation. The arbitrary tine factor obviously depends on distillation rate and my not apply on another installation. "Similarly, the tails fraction is cut out as soon as tem perature (see specimen chart) indicates need, as delay in f.p. determinations results in contamination of good diphenyl. Furthermore, it has been found that f.pt. can be O.K.. but boiling range out of specification If the cutting out of tails Is delayed. ^Scaeral rulest- Use top of column temperature to out Into heads. Use tlim and coil Inlet teapereture to out into diphenyl. Ube coil Inlet temperature (345-355) to out into tails. Ube top of eolusa teaperature to out off tails. " The fare rm is benzene, which is run off to (#21) a small buffer tank (situated near the simp tanks), from which It is periodi cally ptag>ed to the recycle feed tank. DSW 462103 WATER_PCB-SD0000050696 Process in Dota.il, contd $7 Diphenyl Distillation, contd Each still has three other receivers* (#18) Beads and Tails, which are punped back Into the still, by swans of a subraerged pupg>, for the next batch, (#19) Diphenyl Receiver. From here the finished product nay be pusped by weans of a suboerged pusp to storage tanks, to the Aroolar plant, to tank cars, or to a saall flaker. (#20) Bottoms Receiver. Crude Santowax la punped here at the end of tide distillation, tiling the subosrged pimp of the still. This mterlal way be pimped to the Santowax still or drawn off Into drums for permanent storage. Excess material may be pimped into a pend. The distillate flows through (#17) a small open box from which It may be sa&pled and diverted to either the diphenyl or the "Beads and Tails" receiver. The benzene fore runnings by-pass this box. Each of these three receivers Is fitted with a steam coll and submerged pimp. A typical operating record Is given on page52 below, and a uswary of records on page 53 This record Includes figures for the reflux ratios used. A saaple chart from the Multipoint temperature recorder was sent to Mr. V. M. Cooper, In London, Oct.31/47* Anniston recycle the cooling water In their condenser,at atmos pheric pressure, ooollng It In a separate heat exchanger. This Is to prevent sealing of the condenser by Impurities from the pood water. They rely simply on hand control of the cooling water to the heat exohanger to give sufficient condensation without oauslng freezing of diphenyl In the condenser. The temperature of the recycled water need not ordinarily be above the freezing point of diphenyl, but It Is allowed to rise to 80-85C during the collection of the "tails* fraction. Instru mented control of the booling water was abandoned In favor of hand control. The benzene distillate Is examined before being punped to the recycle feed tank, to mates sure It does not contain more than 2-356 at most of diphenyl. The operators Judge by a greenish colour which appears if there 1b much diphenyl, and by evapora ting a few drops between the fingers, but a freezing point test would be the safer criterion. The diphenyl may have filched over at the start of the distillation, as described on page above, or ay have come over through Inefficient fraction {insufficient reflux). DSW 462104 WATER_PCB-SD0000050697 Process In Detail, contd 48 Mpheoyi Distillation, contd The build \g>, in the heads fraction, of Materials other than benzene and diphenyl Is rery slow (only about 0,065^ on the diphenyl according to Anniston report 1642^ but as the frac tion Is recycled to one batch after another, styrene and other contaminants do slowly accumulate In It, (see under "Flow Sheet", page 28 above, far typical values; see also page 116 below), and the out between benzene and diphenyl beeones less sharp. Occasionally therefore a heads fraction "cut" from about 900 to 200C-amounting to about 10 gallons. Is side-tracked, allowed to cool to deposit diphenyl, which is drained off and returned to the system, and the liquid con taminants discarded* Mr. Ellenburg, jan.14/49, In answer to a suggestion from London, agreed that the heads fraction Might be accumulated until there was enough to charge a still. Instead of being recyoled each tine, but he advised against that course on the grounds that a larger store tank would be needed, and that the Anniston method is general^ more conveni ent. Any lspurltles boiling lust above diphenyl will go into the tails fraction, but this fraction is slsply recyoled so that the iapuritlea finally go out In the crude Saatowax. Kr. Ellehburg advises against handling or shipping diphenyl In drums because of the difficulty of re-melting. Anniston flake any diphenyl not handled In tank cars or by pipe line. London, Peb/49, did not expect to sell any diphenyl* The diphenyl still bottoms (Crude Santowax) are kept molten In steam-heated tanks if they are to be processed further. The excess production of bottoms at Anniston Is put into old drums, and more of It has boon puaped away Into an excavation. If the stored material Is needed again. It Is put Into a melting pot and then moved by puling. At one time when smaller (direct fired) stills wore used, Anniston purged the diphenyl still system with CO2 between batches, but discontinued this as being unnecessary with the present stills* The still la ainply allowed to cool a little after being discharged, then a little care taken In punplug In the new (hot) charge of crude diphenyl* Still bottoms, discharged at about 400C have been known to take fire* The circulation through the heating ooil Is not started until the still has settled down after the charging. DSW 462105 WATER_PCB-SD0000050698 Process la Datall, oontd 49 Distillation of Santowax This Is carried out under about 50 * of Hg absolute pressure at the receiver, using a DeVine Vacuum Ptm*> (#40) The pre sent equipment was originally *t up for the separation of erode Santowax Into Santowax 0, X and ?, which are erode ortho, neta, and para laoners of terphenyl. For S&ntowax R only straight over distillation Is required, with a single receiver (#39) Instead of the series of receivers Installed In the present plant* The still obarge Is received at about 150C In one of two Charging Tanks (#35A and #35B). These are interconnected at a low level and really constitute one vessel* They contain heating colls and one has a submerged ptnp (#34d), by which liquor Is circulated through (#34CJ an external preheater coll on the side of a firewall remote from the tanks and still. The ooll Is heated by gas burners and Is suppptrted in a steel flue with firebrick lining. The material is pre heated to 230C* The same puzg> Is used to transfer the hot liquor to the (#35) still which has a similar heating arrange ment* The still capacity Is 6-7000# of erode; The still has a 10" vapour outlet (and explosion diso) to (#36), 12 plate column. This colunn has (#37), a condenser consisting of a single Jacketed tube in the form of a zig-zag. (See pages 104 & 105 below) This condenser was used In place of the original shell and tube condenser because the latter often became Mocked with frozen material* There Is a reflux split box (#38), but reflux Is used only 1m the (rough) separation or erode Santowax Into the o, a & p Isomers* The Santowax R distillate Is run from the receiver (#39) to the buffer vessel (#39B) and pimped from there, by (#39A) to the Aroclor plant, the Santowax R flaker, or to the HB-40 plant, as required* The still bottoms are sold as Montar 9, any excess production being diaped* Er. Hardy reported, Oct. 17/^7: "A total forward flow is maintained on the split flow box throughout the Santowax R distillation. This Is nece ssary In order to get over some of the higher boiling constituents which boll above the terphenyls* These materials must be present to bring the hold, point of the Santowax within our specifications "Our original Santowax still had no reflux at all and the new still (1943) Is supposed to duplicate the product of the old still". DSW 462106 WATER PCB-SD0000050699 froe--g In PeUlIj contd Distillation of Santowax, contd 50 Distilled Santowax Is a reddish brown liquid crystallising to * yellow solid at room temperatures. The distillation is controlled by "Sold point" determinations -- there are two such points and the specification for the pro* duct requires; Upper hold point 137-1*3C (changed Sept.23/47 to 137-1*50C) Lower hold point 45-6oC (ohanged Sept.23/47 to 45-63c) A sasple strip of temperature recorder chart for various points on the Santowax still system was sent to Mr. Cooper, Oct.31/47. On January 6/49, Mr. A, H. XUenburg wrote: "At one tine, distilled Santowax was allowed to solidify la drums, sad shipped, but It was noticed that the pro duct tended to segregate as It solidified. As a result, the outer layer of the drums contained the high melt ing portion of the Santowax R, and the center of the drug was sade tg> of the lower melting fractions. "in order to obtain a uniform product, so that a whole draw would not hare to be melted to get out 10 pounds, the Idea of flaking the Sentowax was proposed. A chill wheel is used to cool the material, and & blade scrapes the material off the rolls, from which It falls Into bags or lined barrels. (Mote, however, that Santowax has two "hold" points In Its crystallisation curve, see page 13 above, and gives first a plastic mass at about l40C, then a waxy solid at about 50C, a fact which affects the flaking process. E.M.) " The ltsagjy material results because the Santowax la not chilled sufficiently before placing In the bags, and a slight fusion of tbs flakes results. However, the partial chilling Is enough to give a uniform solid product. The crude Santowax could be handled a similar way If necessary, because Its characteristics are similar to those of Santo wax H. A big tank for storage of the molten crude is, of course, the easiest solution to a storage problem". Anniston, Jan. 14/49. pointed out that the Santowax flaking could be done on the same drum as the diphenyl flaking. See also the note under Specifications & Test Methods, page 67 below. oSVJ 462107 WATER PCB-SD0000050700  rr-*g T Ht 1r- ; ! ii!! h* t N:. :i ' CO :; c*- < iJ * 5 *1 l u I Vv M ''j r\' M V. \ . -5v c \ 'V N s, > \ _sL < ;J > r. s'* c 0v SI ^ . c \a N. i w^ *v 0 % u. ^ N W 4 ZU-"^L_ *J *Zbe3t E i I n_ae. zo 'v 1> ; - V ' ) v <*' O ^CN , >. . . r~+ cY r**. rs ' ^ OL ' - r\ , Of ' rv !\ . . w fr - 'n >->. i >.'3'K X v* ^V, A T-) *-> H /v^ r1 r^ VrS* I--^v/N">-SV ->*> ^-- ^ Qv; ^___ *s^Q, -io> -o- _<5; - c* ChO'^(P*'Off> c^. TTya^ra^fyyyt WATER_PCB-SD0000050701 o> Diphenyl PlrtlUlatlon Reoord Dirtilllation Ho. Still Ho. 7 P _______;Date Started Operatar Start Operator Finish Started Charging Still at / -Ui- -L /J //; /f7^ 19* _ehirt Shift CHARCK: B&T Tank; Inched bofore ^ Inchoa after gallons oharedd 7 X/S? Ho. 1 S.T. Inches before,Inches after Q gallons charged-?' 7 ^ U-- Ho. 2 S.TT inohesebefere Inches kfter gallons charpod No. 3 S.T. inches before ,<7 inchos after / 4 g.u-'.ons chnrgol No. 4 S.T. inches before Benrol Still " M inches after Inches after g&Z j.ono cha-ge-1 gall one chavg? 1 Total Gallons Chargod 4-. o b O Circulating at /?: Fire> on at /2-: i-T Fir.; P X\ r-TTINGS: l. On // r - ' * '1.Y1- 5. on _ 2 On ' ' at J) , 6l On 5- On S C At '-** "V. AT. 7. On h. On ? O t 8. On _______ b ^ _ , . ** - at at at REH7.0L: Ran " or gal. HEADS: Started orer at -3.` Ran OL A > M Finished at U-- A -M. or /ns* lbs. TAILS: Started orer 2 , Finished at 'J_ Ban/ __________________________" or _,//^ ____lba. DIFHENTL: Total In Diet1111ation y- HIGH BOHJB: Groes Weight Mer Weight y// _lbe. 9b. of Brusaa lba. " of,7$2L lbs. DSW 462109 WATER_PCB-SD0000050702 ( Process In Dotall, contd S*M-ry of Records of Diphenyl Still Working 53 (Taken November 1950 ft*om the plant sheets as on the preceding page) New Material to Still U.S.Gallon 3485 3394 3394 3485 3869 3394 3485 3485 3576 3485 3485 Benzene Fraction off U. 3.0. 437 598 667 506 690 690 874 598 391 598 598 Heads Tails lb lb (8 Ib/OSO) 2024 964 1928 2121 1157 1157 2217 1446 2121 1157 2217 2121 1253 1350 2024 231* 2314 1253 1157 1060 2410 1350 Diphenyl Bottoms lb lb 18,785 17,459 15,470 17,680 17,238 17,017 16,133 16,796 15,249 19,448 18,785 4400 5400 4400 5000 4600 4600 -- -- -- 48oo 4200 DSW 462110 WATER PCB-SD0000050703 C SPECIFICATION AHD TEST KBTHOEB 54 Raw Material* Benzene In September 1947 Anniston were In a favourable position In respect of the quality of the benzene received. By a special arrangement with the suppliers in Birmingham, Alabama, they received material usually running 5*390 Gr.Pt., (on the sample "as Is"), and at least up to "nitration Grade" on all other points except possibly sulphur content. The arrangement was so good that Anniston were able to rely almost entirely on Birmingham control of the quality of the deliveries. The only specification was that lulled In the description of the mate rial as "industrial Grade Benzene", and the deliveries were so much above specification that the specification was inopera tive. Because of this position Anniston do not perform routine tests on benzene deliveries made to the diphenyl plant. There are only very few test reports on file, but the following which were available were believed to be representative of the mate rial usually received: Tank # GATX 3! A6l Date Fr.Pt.C. ,, Sp.Or. 60F. Dist. range 1st drop 5* 79.7 80.2 50* 80.3 95* 80.4 dry 80.4 Colour Water white Wash test #1 Total sulphur * 0.05 Acidity mg NaOH/g HgO p.p.m. Paraffins GATX 21357 79.7 60.2 80.4 80.5 80.5 w.w. #1 0.05 ? Aug^*7 0.8825 O.90C 79.5 80.2 80.4 80.4 80.4 GATX 19403 June 17/48 0.86i.446 1.V ( corr.) 79.4 80.1 80.3 80.4 80.4 w.w. #1 0.0004 700 nil The freezing point method used was that of J.I.E.Ch.Vol.10, p.1010, (see page 58 below). For the other tests, the latest A.S.T.M. methods were used. In order to check the Anniston method of determination of the freezing paint of benzene, against the Ruabon method, and to make sure that the Anniston material really had such a high test, by British test methods, a sanple from rail tank GATX 21357. Nov/47, was sent to Ruabon. It was found to have a crystallizing point of 5.4 c by the standard Ruabon method. DSW 462111 WATER PCB-SD0000050704 Specifications & Teat Methods, contd 55 Sasples, representing material available In Britain, bare ' been examined at Anniston: Sanple NB279 Taken Jta.24/48 Saaplo 60768 Taken Jfcly 27/48 Fr.Pt.C Sp.0r.60F Dist. range 1st drop 5* 5<* 95* dry Colour Wash test Acidity mg BaOH/g HoO p.p.m. Paraffins 5.16 0.886 79.6c 80.2 8O.3 80.4 80.6 w.w. #9 0.0016 458 0cn 00 5.3! 0,8 1.2 B:5 80.2 80.6 w.w. #4 o.oo4 621 nil (corr.) Anniston's eoxsaents on the British s&sple HB279 were that It was not UP to the quality of the Anniston supply, it had a high acid wash colour,Indicating Incompleteid treatment, and the presence of unsaturated or hydroxy oasqpounds, but that It should give good diphenyl If the fractional dlstllla tlon of the diphenyl were good enough. The sample left an odour of tar at the end of the distillation. Anniston's consents on British sample 60768 were that its acid wash test was Just at the limit of the Anniston specification (for "industrial Benzene") again Indicating lncooplete add treatment of the material. Material to this sajg>le would pro duce good diphenyl, though a little extra attention to the diphenyl distillation might be needed. The effects of lapurlties in the benzene were discussed In Anniston reports 1640, p.19, & 1642, pages 1 & 2. For a time no very high standard was set, but In recent years, trouble with the stability of Aroolom used for electrical work has renewed the interest In this subject. See pages 111- below. In a "short form report", Ko.2152, dated Hovei4ber 17/47, Hr. A. M. Ellenburg of Anniston reviewed the quality require ments for benzene for making Biphenyl and Chlorbenzenes. Be found the available evidence to be Inconclusive, and reoonaended maintaining a limit of 50C minimum for Diphenyl and 5.2C for Chlorbenzenes. IT. X. E. Hardy, Doc,23/47, advised caution In accepting a lower standard, far benzene, which contained paraffins, would DSW 462112 WATER_PCB-SD0000050705 c Specifications ft Test Methods, cantd 56 yield styrene and apparently rlnyl biphenyls. Sere again the chlorine derivatives wore blaaed for skin trouble in the Aroolotr plant* HC pointed out that though the chlorinated products my be stabilized by being held at high temperatures. It Is not certain that this process will go far enough. On April 28/H9, Kr. KLlenburg Mentioned the beneficial effect which high tesperetures In ooks crons are known to bare on tbs purity of the benzene (freedosi fre* paraffins). See also the discussion of Effects of Variation in Benzene Quality/ page 111 below. Test Methods for Benzene Vr.Pt, As Mentioned abore. Amis ton use the Method of md.Bng.Chem.lO,, p.1010 (20 ce sasple in the "as is" condition, put in a test tube in an lee bath, and when the solid benzene begins to separate, the tesper&ture will ramin constant fnr a tlMs. This teapereture is taken as the solidifying point. Accurate 0.05C. Other Tests Anniston use A.S.T.M. Methods, but on Fob. 15/^9j Amis ton re ported that the British (N.B.C.) form of the acid wash test was simpler, especially in respect of the preparation of the standards. On applying the N.B.C. test to their own benzene, they found It to glre a colour Matched by only 0,1 g dlchroaate per liter of water as agslnst tho H.B.C. standard of 1.0 g. The saae benzene gare colour jfl on the A.S.T.M. acid wash test. DSW 462113 WATER PCB-SD0000050706 ,Specifications & Test Methods contd 57 Lead for the Pots So specifleatIona exist on lead used In the diphenyl lead convertor units other than that "chemical lead Is to he purchased. Apparently the effect which impurities in the lead say have on the stainless steel. Is sore feared than any effect an the conversion process Itself. Kr. V. V. Made of Anniston have the following Information: "A typical analysis of chemical lead &s stalled by the St. Joseph Lead Coapaay, 250 Park Ave., Hew York, New York, wasi Fb: Cu: Ast Sb: Snt Pe: Zn: Cd: 99.9272* 0.0608 0.0001 0.0001 0.0001 0.0003 *** Co: Ag: Bi: 0.0048 0.0066 nil "Elements that should be particularly avoided, as observed la practice at Anniston, and as recommended by manufacturers of stainless steels, are, in order of importance: tin, zinc, bismuth, antimony, and cadmium. Tin, zinc, and bismuth should not exceed 0.0002*max., antimony and cadmium, 0.0005* max. "Caution should be taken in plant operation to avoid contamination by any of the above elements." Spectrographlc analysis has been found useful far checking the purity of the lead for the convertors. We are warned to look out for impurities which may be picked UP In any reused lead. MCL have found a source of supply of lead apparently up to the Anniston standard except that the bismuth content is 0.001* and the lead percentage only 999955. Anniston consented that they did not think 0.001* of bismuth would cause any serious trouble. DSW 462114 WATER_PCB-SD0000050707 Specifications and Tert Kefchodc, cont'd 58 Finished Products Diphenyl code 3380-500-74-0p ~ TTeounlcal) authorised July 16/45, (Sampled from the rail tank, la supercedes Kar. 2/38. the ease of tank oar sales) Production Dept. Jfo. 09 Appearance yellow to pale amber solid. Colour Pr.Pt. equal to 2.0j aqueous FeClo s&Z: 66.r7rO min. J Dist. range 2.5C max. including 255C. Freezing Point of Diphenyl Standardised Anschutz thermometer 40-100C. zsounted vertically near shaft of a email electric stirrer? Bulb about 1/4" above stirrer bled, and 3/8" frera th shaft. Ragnlflor clanged to thoracise-ter. Kelt and nix th =-*Ie. Pour about 120 o.c. Into 250 co Griffin bsaksr nulling cure thermometer Is not above 100C. Set be&ksr u.r?der stirrer so blade 1* ^-l" above the beaker, and the thermometer lsmerced to 40C. nark or more. Run stirrer at moderate rate through test. When cry stallization begins, the fceit$>erature will rise quickly about 0.4C., then becoaie >. onstant for 2-5 lain* before slowly falling. Highest reading taken avoiding parallax. Dlst. Kenge of Diphenyl Plant "B" method slightly modified by R.Sudhoff, Har.8/37. 200 cc Pyrex (Coming 1600) distilling flask. Feck wrapped with wet asbestos paper allowed to dry on. 200-275 range precision Taylor type thermo, engraved stem 0.2 division. 1 - 4 m.'i. Both bulb and expansion bulb to be below the side tube. 6" auxiliary thermo. 0-150, attached with bulb 2/3 way down tho emergent eteia. Cover both thermos oS\N 462115 WATER PCB-SD0000050708 Specifications find `lest I-^thcrts, oont;d Finished Products, cont'd with 3/4" gls.sc tut. Correct for emergent stem, fend b&roristriC end' thervicnetrio e.- .-or. Condenser as for Barrett bensol distillation, i.e., 24" x im glass tube @ 75 anc.l Jn copper trough. W> r at 75-5C. Auxiliary burner to prevent freezing at exit. Drift shield 5* I.D. x 6* high. Flask on 6" x 6* piece of 1/4* transit board with 1-1/4f hole. Top of auxiliary bulb of theraossetcr level with bottoa of side outlet. Cork stopper flask to con denser. Weigh 100 g. acujyle into flack. Hc-at with Bker burner till ebullition startr. When vapour ring reaches into the neck, but not to the side c.rn, cool a little. Repeat 3 tiroes to get tberronstor and flask neck heated. Distil 2 drops/oeccnd into bnfJcer on a balance. Road: . 1st drop 1% 32 52 502 902 952 9^2 972 Ery read auxiliary thersFoiaster fit *. Record weight in receiver 1 ninute after removal of flane. Eky points observe when last drop ie ves bottom -- needs two observers. ^760 - Barom. reading } x 0.036 = (1st drop - average of auxiliary rsaolngs) x 0.000153 x no. of degrees above the bath * Tg. Add and Tg to observed readings. 5* OSNN 462^6 WATER_PCB-SD0000050709 1 t Specifications and last Methods, <r>nt'd Ol! Finished Products, cont'd Diphenyl Colour Melt 120 g diphenyl In low form Griffin 150 co beaker. Conpare @ about 8oC. with ferric chloride sonnies In bottles of same diameter as beaker. Standards 25g- FeClo^T^O 0.625 g. C0Cl2*6lL,0 in 200 cc JIgO * 15 cc HCI (S.G. 1.18), make up to 250 cc. Make Standards 10 cc stock per 200 cc 20 50 60 Standard Fob. are . Saqple 0-4072 of Finished Diphenyl was sent from Anniston to Ruabon, Oct.13/47. 1 1$ 2 OSW 462117 WATER PCB-SD0000050710 Specifications fc Test Hathods, contd 61 Mr. Barden, August 1950, collected the following test results at Anniston: "Typloal Analyses: Diphenyl Date Colour ?.Pt. idst. Range 1 drop l 3 5 10 50 90 95 96 97 dry 31ft 1.0 68.8 0.6 254.8 254.9 254.9 254.9 255.0 255.1 255.1 255.2 255.2 255.3 255.4 25ft 1.0 68.85 0.7 254.7 254.8 254.9 255.0 255.0 255.2 255.2 255.2 255.3 255.3 255.4 Lot 246 20/r 1.0 68.9 .5 254.8 254.8 254.9 254.9 255.0 255.1 255.2 255.2 255.2 255.3 255.3 18A i 08.9 0.6 255.0 255.0 255.1 255.1 255.1 255.3 255.3 255-4 255.5 255.6 255.6 247 18/r e 68.9 0.8 254.7 254.9 255.0 255.0 255.1 255.2 255.3 255.4 255.4 255.4 255.5 12/5 (1950) 1.5 68.8t 1.0 255-2 254.8 254.9 255.0 255.0 255.2 255.2 255.3 255.4 255.5 255.7 Date Colour F.Ft. Diet. Range 1 drop 1 3 4 10 50 90 95 96 97 dry 25/4 1.5 68. 8 1.2 254.8 255 .0 255 .0 255 .0 255 .1 255.2 255.4 255.6 255.6 255.8 256.0 27/2 1.0 68.8 2.5 254.9 255.0 255.0 255.1 255.2 255.4 255.6 255.8 256.0 256.4 257.4 (1950) Test results collected in 1947 are given on the next two pages, 62 & 63, DSW 462118 WATER PCB-SD0000050711 TECHNICAL DIPHENYL 3HIP HERTS *p c CO o Uk/ b- On 0 o o 000 %H 2 H rH oS & o o u 0 C (x, e o o oOo 3 o o rH rH H 1-1 rH ON H CO COo ON 0 tn NO m in tn in in in in in CM CM CM Cl Od oo rl rH ON C in in in in CM CM t-- CO ON e c- b'* CO 3 COo voc $ C\ t- t-o VOe VOe f-o t- tne ON NO t~c ir. NOc NOo in tn oc $ ON in * NO Hr * m tn in tn e lA in m <* * si- .*? ce in 9 rH m w cry * Cd 9 na o ' rH m $ CMo CM C r-S Cl CM Pi CMc V*. CA CJ o rv o rH *' W6 CM 0 O rH oC e o Vt CNri H c O Cv rH o o rH 99 *H rH m m tA in in in IT* ir. in LTi in in in Cl CM CM cv OJ CM CM CM rH rH -rH rl r--l o A O* o CN On o * o m in in in in oi- J* r( O (i in tn tn UN in in in OHO CM CM CM Cd CM CM CM !o P-c ON> On o ONc On c ON ON ON * z> CO CO cc s CO 00 00 fc VO NO vo NO vo NO -! o o on in VO -oc in cn b- CO rH Ox fc *r rH in vo m tn NO vo JC fcj K K K KK Li pj fe o K O < it; o is o is o Soc o 35 X C? ac R X X t~ V o 4-> a S' n t- t- t~ t- -i N A? N < "V O t- ON t"- c r-i >/ H O H e PC7c Oc. G, CO <n n CO OS>N462^9 62 WATER PCB-SD0000050712 4* ft c nC ws* VO t- 00 CO 0 c o H o O o o o in H H H H r-t o O o b| o OV t- Ov Ov VO vo Oft C+k* n m CM in m CM cn in CM in m m tn m u.'l CM CM CM XI & X) 4 CO CO 0 in in* 63 * t in o\ VOo VO e VO t- t- -=*" * t- CM in e in. $ o\ m o mo VO vo CM ft- 0 O in <4o ** in in in cn m CM $ rt 'V o C4 cn CO ov CM W in :-j o Ooi cnc cr, * CT\ 0 rt 0 % CM O m m COo rS c o rj r-S Gi CM CO C P m 4m -l CM -l H PPP O cm o CM --t CO Ov ee act n in in in ft ft r)r(3 n m in in tn in Q CM CM CM CM CM CM p C?V OV 9 C\ Ov e 0\ 0 o c* CO CO CO CO Ov fc O VO VO VO VO VO p Q n X) eto- o -< OV Ov cH H i-i r"f rH rH -ft N CM <-1 c P g Qo, ftt- ftt- \ in \ft rl CM 0 & D, co co sw 462120 WATER PCB-SD0000050713 V Specifications & Test Methods, contd 64 3AHT0WAJ R The properties of Santowax R as given In "Monsanto Cheo&cals and Plastics" - 27th Edition - are typical analyses* Tentative Sales Specification Color 3.5 - 4.5 NPA Solidification Hold Tesperatures: Upper 137 - 143C Lower 45 - 69c Distillation Range: At least 8o$6 between 350 - 400C funcorr.) This property is not aeterained on every lot. See also under "Properties of Materials", page 13 above. DSW 462121 WATER PCB-SD0000050714 65 Specifications and Test Methods {cclit. ' d ) Analysis of Santowsx R DESCPIFTION Santowax R Is a yellow solid hydrocarbon mixture of micro-crystalline structure. The main constituents are ortho-terphenyl^ reta-terphenyl with smaller quan tities of diphenyl and tri-orthophenylene present. The molten product is chilled cn a flaking rcll and bagged or put in paper lined barrels for shipment. LABORATORY TESTS: 1. Color: 3-5 - H.5 NPA Test Method; A portion of the sample is melted slowly over a low heat. Care should be taken not to overheat the sample. Some darkening due to decomposition 5.s possible if the sample is heated above 250C for any length of time. The molten sample is poured into a clean 15 mm test tube and the color is compared to the standard disc comparator of the National Petroleum Association. * 2. Hold Points: Upper hold point 137-1*3C Lower hold point 45-6CC Changed Sept.25/* to: 137-145 ^15-63 Method: A 250 cc beaker is filled about three fourths filled with the flaked. 8&Bg>le and placed on the hot plate to melt. After the molten sample ha3 attained a temperature of 175-200C, the stirrer and thermometer are placed In the beaker. The material is allowed to cool while stirring until crystals appear in the liquid. Then the thermometer is watch closely to see if the temperature holds as the sample crystallizes. This temperature where the thermometer reads constant for a few minutes is known as the upper held point. The stirrer i3 removed and the sample is now al lowed to cool to about 80C. The sample is then stirred by hand with the thermometer. The chilled material on the sides of the beaker must be inti mately mixed with the molten sample in the middle at all times. The temperature at which the ther mometer stops after it goes down and rises to 1 to 3 Is called the lower hold point. DSW 462122 WATER PCB-SD0000050715 66 SpecificatIons an-1 Test Kethods (cont'd) Analysis of Santow&x R - Page 2 Standard therr c :,etcrs p.re v.se>i in this test. No correction Is made for etan exposure. 3. Distillation Range: At least 80 per cent between 350_400c (uncvrr.) Method: ASTM D-20 (modified) The apparatus is designated in the ASTM method and consists of a flask, condenser tube, shield, receiver, and thermometer. After assembly of the apparatus is complete, the sample is melted and stirred to insure proper blend ing. One hundred grerns of the sample is weighed into the flaak, the flask is put in place and heat is applied so that the first drip comes over in from 5 to 15 minuteB. The distillation shall be conducted at the "ate of 50-70 drops per minute and the distillate collected in a weighed receiver. The temperatures recorded shall be at the first drop and at each 10 degree interval between 350-400C, the per cent distilled over shall also be noted. This distillation range specification is subject to change as aor-e history on this test of the prod uct is obtained. 4. Sediment Tost: There shall be no insoluble impur ities in a molten sample of the product. Method of Sampling: It will be very difficult to obtain a good sample for this test and, therefore, the best means of having a clean product is to el iminate the sources of impurities and keep a check on these sources. For the above test, a 3 or 4 pound sample could be melted at hourly intervals to see if impurities were present. os^J 4^2A23 WATER PCB-SD0000050716 (_ Specifications anc vest Methods (cent'd ) 67 Analysis of Santorax R - Page 3 For a fool-proof Kvthod, a mechanical means would be needed at the flaking r<lls to remove the metal scraps which are the main contaminant at present. If the rolls were capable of cooling the product to 30C, so that a cold, brittle product could be obtained, the metal could probably be removed by a vibrating screen of 4 mesh size. A magr.et might also be capable of removing the metal scraps. At present the flaker is being resurfaced and a new blade is being installed to prevent the clings from coming off the ro'ls. DSW 462124 "WATER PCB-SD0000050717 c Specifications and Test Kathods, cont'd 68 Determination of Sulphur In Santomx Tentative Method: 40 g. sample, 20 g. KIckel formate, 150 cc H-B. 40 (hydrogenated Santowax) In 1000 cc boiling flask. Heat 6 rate of 30-40C per minute to 2S5C. Water of cryst. of formate cohos out up to 200. Formate begins to deconf>oss around 240, and rapidly @ 270-275 Heat to 285 and cool to 20-80C. Add 100 cc 1:1 KOI and some boiling chips. Reflux till Hi has dissolved, (foams), with Meyer or similar apparatus attached to condenser, 50 cc annnoniacal CdCl2 (l/ of CdClg ^ the bulbs. Wash CdS to beaker or flask. Hake acid to ; benolphtbaleln with concentrated HC1, and dilute to about 250 cc. Add 10 cc H/l0 Iodine and 5 cc concentrated HC1 added. Leave 10 minutes. Titrate iodine with K/lO thio uring stsroh. Calculate to % sulphur In Tantonax. Do blank. Determination of Chlorine In Santovax 30 g. Santowax and 1 g. sodiua In dry Hi crucible, 80-100 cc capacity. Cover. Heat over low fl&Ee, until Santowax has gone, than to dull red. Remove cover and allow excess Ba to bum under hood. Cool, wash well with HjQ. Make up to 250 cc. Determine Cl grevlmstrlcally in 100 cc aliquot. Do blank by dropping Ma into water. DSW 462125 WATER.PCB-SD0000050718 ( Suscificatioaa &!*d Tost Ratliodc, i ^ t. .-r^sa----r u n i lit - - . - '. * ~ 5WS~***-- 69 S&ntowax R o-ijOI Bo Specification, Data Sep.l5A7 Sop. 22A7 Lot Bo. 47 48 50 51 52 53 Upper Hold Lasear Hold Colour rt. c0 Pt. C. KPA 144.2 60.0 38 144.2 61.0 3.8 142.0 60.0 3.6 142,5 60.0 3-5 142 oC 56.0 3.5 142,0 59.5 3.5 Sug>le 0-4073 of Sastovax H &s sent fro* Anniston to Button, 0ct.l3A7* OSW 462126 WATER PCB-SD0000050719 Specifications and Teat Methods (cont'd) Finished Products (cont'd) 70 Santowax, (crude individual i 3 cruer s} Only small amounts of the individual isomers have been made. Test methods used at Anniston are given below. Analysis of Santowax 0 DESCRIPTION; This material consists of crystals coated with a eutectic oil. The color is pale yellow due mainly to the adher ing oil. The product Is about 97-93 per cent ortho terphenyl with traces of diphenyl and meta terphenyl which form the oil. Tills product is distilled, mixed In the catch tank and run into drums to crystallize before shipment. LABORATORY TESTS: . 1. Color; 1.5 NPA Method; The sample Is mc-ited with precautions not to overheat and poured Into test tube. The color Is compared to the standard disc comparator of Nat ional Petroleum Association. 2. Hold Point: 53<>C - 55C. Method: A portion of the sample Is melted to fill a 250 cc beaker one-half full. A stirrer and a thermometer (standard!red) are Inserted In the beaker and the material is allowed to cool while stirring slowly. When the temperature gets down to 55C, add seed crystals slowly until sample begins to crystal lize. Record highest temperature reached after ther mometer starts to rise. 3. Distillation Range; 10C maximum range. This range should Include 335C (corr) Method: ASTM D-20 (Modified) Sane procedure as for Santourax R. DSW 4b2127 WATER_PCB-SD0000050720 71 Speciflcatlons and Test Methods (cont'd) Finished Products (cor.t'd) Analysis of Santowax 0 - Page 2 The temperatures recorded should be at the first drop and at each 10 per cent of distillate up to 90. Then the temperature should be taken again at 96 per cent of distillate. Reces3aiy stem temperatures should be taken to correct for the emergent stem. Correction should also be made for the atmospheric pressure. The distillation range test is subject to revision when more history or. this test of the product has been obtained. 4. Sediment Test: No insoluble impurities present in the molten sample. Kethod of Sampling: The material shall be fil tered through 100 mesh screen as it is leaded into drums. A sample of the molten product shall be inspected for colids. The drums also shall be inspected for cleanliness. DSW 462128 WATER PCB-SD0000050721 ,Specifications and Test Methods (eontd) Finished Products oontd Analysis of Santowax "in" This material Is a yellow solid of crystalline structure. Tbs purity is approximately 95 per cent rasta terphenyl with para terphenyl making up the reminder.. The pre sent method of separation yeilds a neta-para eutectlo at 3,3 per cent para which accounts for most of the para contaminants. At present this product is crushed to Imps one Inch or smaller, and shipped in paper hags or paper-lined barrels. laboratory Tests 1 1. Color: 1.5 - 3*5 HPA Method: The saaple is melted with care not to dis color by heat and poured Into a 15-am test tube. The color of the molten sawple is cospared to the standard HPA disc caaparator. 2. First crystals: <K)C. Method: A 250 cc beaker is filled about two-thirds full of the molten sasg>le and placed under a stirrer and a standard thermometer. The material is allowed to cool with stirring until the temperature reaches about 90C. Seed crystals of the saaple are added slowly as the cooling continues. The temperature at which new crystals cone out of the solution is reoarded as the "first crystals" point. 3. Hold Point: 83-85. Method: Stirring Is continued in the above sasple while the teaperature goes down and then rises 1-3C to the hold point. The highest temperature at which the teaperature remains constant for a while is re corded. '! 72 DSW 462129 WATER PCB-SD0000050722 Specifications and Teat Kethoda (cent!d) Finished Products (cont'd) Analysis of Sg.ntoyax ~,M"- Page 2 4. Sediment Tests No Insoluble Impurities are present in a molten sample cf the product. Method of Sampling: A sample of the crushed mater ial as it is prepared for shipment shall to rrelted for this test. The product can be-filtered before it is remolded after removing the water. Then it can be crushed to 1/2." size for shipment Under the present open air method of handling this material, the dirt la especially hard to keep out. os.V'J WATER PCB-SD0000050723 Specifications and Teat Methods (cont *d) Finished Products (cont'd) Analysis of Santowax P DESCRIPTION: This material is a white solid which crystallizes in thick plates. The purity Is 95-98 per cent with small amounts of mete terphenyl and tri~orthophenylene as the contaminants. LABORATORY TESTS; 1. Color: 2.0 Iv'FA Method: The sample is melted and poured Into a 15-mm test tube. Of the standard disc comparator the color is Judged with KPA standards. 2. Crystallization Point: 209-213C Method: The cooling in this test is started In a mineral oil beth heated to 230C. The sample Is melted and poured into a large test tube. The test tube is placed in the mineral oil bath and cooling begins. The test tube contents are stirred as seed crystals of the sample are added. When new crystals appear In the solution the thermometer which Is in the test tube is read. This temperature la the cry stallization point of the para terphenyl. 3. Sediment Test: Ko insoluble impurities are found in molten sample of the product. Method of Sampling: A thief sample of each barrel of material shipped shall be examined for Impurities. This material is aleo hard to keep clean under the present means of handling. DSW 462131 WATER_PCB-SD0000050724 V Control Tests 75 Coptora Ion Stage t j Conversion There la a connection from the pipe between the lead trap and the bottom of the ooluan on the convertor, leading to a snail, water-cooled condenser. Saaples of the vapour nay be drawn off at any tine, condensed, and a careful measure ment of the density of the condensate made, as a measure of the jf conversion. The Amiston operator makes this test once a day on each unit, usually in the morning shift. Each unit has a iwupllng condenser, as shown on the attached sketch, permanently connected, and there la a steel sheet "fire screen" to protect the operator and the equipment from the radiation from the conversion unit (especially the lead trap). The operator starts the water flow through the condenser and opens the saspllng valve enough to give a distillation rate of about 500 ml/hr as Judged by eye. Ho .purging with OO2 Is done. Some of the condensers have a screw cap fitting with the oendensate and vent tube passing through, others simply have a two-holed rubber bung. In either ease, the easple bottle must make a good vapour-tight connection. The use of spring clips to hold the bottle against the stopper has been suggested. The sample bottles are of glass, square, 3i" x 3" x r on the straight side, with screw necks, and the operator oarrles 12 of them, one for each unit (clearly numbered) in a wooden rack, puts one on each unit, leaves them for about an hour, then turns off the vapour and water, removes the bottles, puts on their screw caps or blank corks, and carries the twelve of them to his test bench where he has a twelve-holed water bath, at about room temperature, and leaves them until their temperature Is somewhere between 3% and l6C. A portion of each sample In turn Is poured Into a glass hydro meter Jar and Its temperature and density observed. The hydrometer Is graduated 0.880 to 0,950 Sp.Gr. in 0.0005 divi sions and should have been calibrated. The tesperature should be determined with an accuracy of 0.2C, and even at that the test Is not highly accurate, but it Is good enough for routine control. . Tbs % conversion, corresponding to the density and tesperature. Is read frost the attached nomograph. The saaples do not crystal lise, but are must, of course, be taken not to allow serious loss of benzene by evaporation during the saspllng and testing. The easples are quite hot when first collected, and there Is a distinct smell of benzene (and HjjS) from the vent of the test condenses1- while the easple Is being taken. There Is, of course, a slow stream of hydrogen escaping through the bottle. DSW 462132 WATER PCB-SD0000050725 WATER PCB-SD0000050726 0.890 Hssy.Bre conversion chart Froa Final Report No. 1712, "Study of the Diphenyl Process, Part HI" By F. T. Miles October 19, 1942 0.900 (X !_ 0,910. I ( r 0.9211 Specific Gravity 54 52 3 0 30 1 x 8 8 ' . i_ 6 26 T x 4 24 * f 2 22 i_8 IB 16 Temperature C. Monograph far Calculating Conversion of Benzene from Speelfic Orerity In the example shown by a broken line, specific gravity of 0.8977 at 30.0*C. shows 18.65 per cent conversion. DSW 462134 23 I 22 h P Y~ " 21 I- P r 18 ___ 17 fh 16 ____ 15 i Per Cent Conversion WATER PCB-SD0000050727 76 Control Tests contd 4> Conversion, contd This f conversion test is not done until a unit has settled down to steady running conditions. IT the f conversion of any wit Is distinctly different froai lQf, the temperature of the conversion pot will be altered, 2 to 3C at a time, to bring the conversion back to normal. Analysis of Crude Diphenyl Melt 100*150 g. In a 250 diet, flask with 10" ooltm. Into weighed receiver to 275C. < dinhetrrl = ^ weight distillate x 100 weight sample Distil Test for Benzene Leaks Into the Stack Oases See page 91 below. Distillation Stage "Bold" points and crystallising points are used In the control of the distillation of diphenyl and of santowax, though in both oases Anniston run the stills largely by "dead reckoning", plus colussi top teejperatures in the ease of diphenyl, because of the time required to take the samples and make the test. (Sea page 46 above). The test bench In the department Is, however, equipped with the mechanically stirred crystallising point apparatus described under "Specifications & Test Methods" on page 58 above. See seaple G-4071 of Crude Diphenyl, saaple 0-4074 of Still Bottoms frost the diphenyl still, and supple 0-4075 of Still bottoms from the Santowax still, sent from Anniston to Ruabon, Oct .13/47 Sables 0-4074 & 0-4075 were lost, and repeat saaples were sent J&n.25A9 to London. Benzene recycled from the distillation stage should not contain more than a few parts per cent of diphenyl. The operator can Judge by the colour of the benzene and by allowing a few drops to evaporate on his fingers, but a freezing point test would give a bettor criterion - ' DSW 462135 WATER PCB-SD0000050728 LIST OP DRAWXM3S 77 There are over 350 drawings on file at Anniston, but copies only of those thought to be of interest to MCL hare been sent to London at different times,as listed below. Those narked "A" In oolum 3 vrere sent by Mr. V&gel of Anniston, to Kr.W.K.Cooper In London,by airmail, Oot.lO/47, a second copy going by sea mall to Mr. ft. Buis In London, Oct.13/47. brewing fiSL:______ Dgggrlptlon Date sent Sent -tP.. Conversion Unit AA-5581 Cast refractory tops for linings A 9C-5584-3 Cast steel door on shell of fur nace A 9C-7817 Lead pot, flange, & thermowell July 21y^0 OVM 90-7845-5 Stacks on VapourIt or A 9C-7847-8 Pot #1, details,with later revisions A 9C-7937-4 Baffle for distributor A& July 21/50 OWM AA-7955 Cast steel flange for distrib. pipe A 9C-8131-4 Details of column A 90-81*3-3 Traps for colunaa A 90-81*7-3 Stainless piping for convertors A 9C-8151 Brick lining, #6 unit A serves for references to 81*7-1 & 81*7-2 9C-8153-3 Details of mounting of burners A 90-8159-1 Manhole Into stacks A 9C-8166-4 Pot cover 9C-8166-2 Pot cover,Baffle locations, tmt its 11, 12, & 13 JUly 21/5 OWM A OSW 462136 WATER PCB-SD0000050729 List of drawings, oontd 78 braving So,____________Description Cate sent Sent to MCL to ResiarkB Conversion Chit, contd 90-8194 90-8205 layout of piping, diphenyl traps to svoqp tanks Layout; benzene feed tanks & pusps Hov*13/50 see letter DWT EM-DWT Hor.13/50 for aaendBents Hov. I3/5O DWT do 90-8208-2 Layout; feed piping & asters Nor. 13/50 DWT do 90-9015 Layout units 11, 12, & 13 90-9019-1 Preheater pot 9C-9019-2 Distributor A A July 21/54 0WM 90-9022 90-9024 Convertor shell, units 11,12,& 13 Yhpourizer coll details A A 90-9025 90-9026 Vfepourizer shell Vapourlzer assembly A Apr.1/48 DSH Revised to Kov.14/47 90-9054 Oes piping to burners, & bill of Materials A 90-9055 Vertical condenser & split flow bojc A 9AA-9058 90-9069 Expansion flange in steaa Jacketed Piping Aug.10/50 DWT layout of benzene nmback Kor. 13/50 DWT see consent EM-DWT Hov.13/5o 90-9070 9C-9075 Hg piping to coolers Nov.13/54 DWT Benzene line, coolers to feed tanks Nov.13/54 LWT DSW 462137 WATER PCB-SD0000050730 List of drawings, contd 79 Drawing So. Description bate sent Sent to MCL to Remarks Ho. 2 Diphenyl Still 90-7870 90-7871 Fractionating eoltaan Layout Diphenyl still #1 A A 90-7872 Condenser A 90-7873 Still A 90-7875 Diagranssatlc flow sheet. New still. Included In 0Design Data for Diphenyl Stilla, Sept/38. June/46 90-7877 90-7886 Hiboiler tank Swivel split flow box A A 90-7888 90-8152 90-8192 Diphenyl still condenser details July 29/48 DSH 1500 gallon catch tank A 6000 gallon catch tank A 90-8193-1 Layout of crude catch tanks A 90-8210 Layout #2 still A 90-8211 #2 Still heater shell A 90-8212 #2 Still heater assembly A 90-8213 #2 Still catch tanks A 90-8214 #2 Still heads and tails tank A 90-8215 #2 Still: piping from siap tank to still A 90-8223 Rotawelr A 90-8224 Rotawelr details A 90-8254 Layout of piping, plan & eleva tion. Diphenyl still A DSW 462138 WATER PCB-SD0000050731 List of drawings, oontd Drawing Vo. Doscript Ion 80 Date sent Seht| to MCL to Remark* 90-7360 Arrangements for burning gg (obsolete), not sent to MOL, 90-7366 90-8040 90-8165 Hg coapr; irlrlng 4 control Hg condenser piping wiring 4 piping diagram discussed In AME'a letter Jhne 30/48 May 24/48 DSH * !A A Bee note on bc-7366j 90-8188 90-8279 Layout & piping Layout Hg piping from units A4 Wot. 13/50 A toe: w Hot. 13/50 EM t DOT for amend ments HYB-14836 Grissom Russel Company Condenser (Pinned) A * See AME1 e letter June 30/48 for corrections to thlB drawing, also E.O.Thoenas, May 21/48, explaining that it relates to one oosgaressar, while 90-8165 refers to 2 and 90-9095 to 3 90-9095 Hot seat to MCL Santowax Ihstillation crfcrurfcfT n c 90-8551-1 High boiler distillation Juno 1946 EACH 90-8551-2 Elevation 90-8552 Hlghboiler distillation A June 1946 EAok 90-8552-2 Pictorial Elevation A General 90-7850? 9C-7851J 90-8140 Dlagrasmatlc Flow Sheets. Inclu- June 13/4^ ded in "Design Data, 4B Diphenyl Unit, "Aug. 9/39 #6 Unit. March 25Al June 13/4 EA0 T DSW 462139 WATER PCB-SD0000050732 Xlst of drawings, oonfcd 81 DSW 4G2140 WATER.PCB-SD0000050733 GENERAL M0TB3 OH EQUIPMENT LIST 82 Reference should be made to Anniston drawings and design data for further details of the different pieces of equipment. The following list Is Intended only to giro a general description of the equipment, and a few coeanenta picked up during various visits to Anniston. The equipment list is valid only for units other than #4 and #5 The design was developed at Anniston and some modifications were introduced when units #4 and #5 vere planned. These changes were found to be undesirable and the design of subsequent units. Including #11, 12 and 13# (the last 1nstalied oilowed the ori ginal pattern. See page 32 above and page 85 below. All benzene lines are steam traced and lines carrying diphenyl or hi-boilers are steam Jacketed, the Jackets being liberally provided with expansion Joints. (See AnniBton drawing 9AA-9058). It Is important to carry the steam Jackets right up to the flanges. The diphenyl conversion and distillation units were indoors, the store tanks and the Santowax distillation unit outdoors, with some shelter for the heating furnaces, vacuum puny and Instrument panels. See the notes on "Gaskets" page 24 above. The conversion and still buildings are high, with effective roof ventilators, and are open completely on the East side, up to the level of the working floor. (Pot rim level). See also pages 16- above. The floors all are of vertical steel grating and Anniston are very emphatic about the advantages of this. flanges on the conversion system piping are of the standard ammonia pattern tg> to 3" pipe size, and are special cast steel tongue and grooved flanges of Monsanto design^.bove 3". Crane clasp gate valves up to 300C, no valves above 300C except cast steel gate valves on the diphenyl still. Materials of construction for various Items are detailed In Anniston Research Department, Report No. 1642. See also pages 22 ft 4l above and 86 ft 88 below for notes on the choice and performance of different stainless steels. DSW 462141 WATER PCB-SD0000050734 Equipment List, oootd 83 #1 Benzene Storage Tanka Eire horizontal steel tanka. See also pages 17 & 29 above. #2 Ptnm> for supplying #3 Controlled by float In #3 which exits off the pimp when the (5000 TJSO) tank, Item 3, contains about 4,500 gallons, and re-starts It at about 4000 gallons. #3 Recycle Peed Tank See also page 29 abore. Efawlng 9C-8205, but see amend ments In letter EH - DOT, Hot. 13/50. The In use In Horember 1950 was a horizontal cylindrical steel Tassel with bulged ends, about 7*-6" diameter by 13r-0n straight side. Capacity about 5000 XJSG. Set in side a concrete retaining wall outdoors at some distance from the flow control room. Pipes to and from the operating building run OTerhead under a steel sheet shield with a steam tracing line. Fresh benzene In from #1 via #2 under float control; recoTered benzene In continuously by gravity from the reflux split boxes, item #12, of the conversion units, and semloontinuoualy by pimp from the benzene receivers. Item #27, of the hydrogen cospresslon room. Benzene In (after exami nation), batcbvlse by pump from thetenzene receiver, ltea #21, of the diphenyl still. This batcbwlse return Is dis liked, for Newport conditions because it would cause wide variations In the ratio of fresh to recovered benzene, and it has been proposed to Interpose a buffer vessel so as to smooth out the flow. Benzene out, by a bottom outlet with an "upstand" in the tank to retain water and solid matter, to the feed pimp, ltea #4. The use of a bottom outlet helps to prevent air locking of the pimps. The line to these props Is Jacketed for steam,but It had no steam connections in November 1950. It is cemented through the retaining wall. Drain, from a simp at the lowest point of the tank for removal of water. This drain requires attention several times a day and should be conveniently looated so that the operator can see when the water has all run out, and the benzene is coming. There Is no special provision for sampling, though a saaple might be drawn through the drain valve, or taken ^froa^ne of the branches on the line from the feed pimp^ DSW 462142 WATER PCB-SD0000050735 Equipment List, contd 84 #3 Rooyolo Peed Tank, contd Vented to the hydrogen system by means of a l" Insulated line, (the tank oontents are usually warm at Anniston, partly fro* the stream of recycled benzene.) There Is a if" branch on the top of the tank, with a gate valve through which the tank might be gauged. This Is very rarely done. Float controller stopping the pur*), Item #2, when the tank contains 4500 TJSG, and restarting it when the level has fallen to 4000 030. Floats for high and low level alarm. (There is no formal drawing of this arrangement). These various floats are now made of glass, stainless floats having failed, and stainless guide rods are recommended because sticking of the float my cause bad trouble. The domes over the float controllers are Insulated and steam traced, but the tank Itself Is not Insulated. It Is, of course, electrically earthed ("grounded"). Anniston experience Is unfavourable to the use of Internal steam colls In benzene tanks because of the risk of water getting Into the benzene If the coil falls, and of ben zene loss via the coll. Even if the coll Is taken out over the top of the vessel, loss is possible by siphoning. #4 Feed Pung> to the Conversion Uhits Anniston report 1642 of Oct .15/41 describes the feed pimps as Wes co Ho. 507 (one then for each unit) with band-operated by-pass to set the delivery rate. Stainless shafts with "HI Resist" lnpellorg and liners, 1800 RPM. In November 1950, one Gould centrifugal pimp with an In stalled spare (and room far one more; was serving the whole 12 units through a row of flow controllers, one for each unit. It was of the 2-stage centrifugal type, 3400 RPM, (but this high speed Is disliked by the operators), gland packing of fiber-reinforced rubber rings, with the gland nut slsply tightened carefully with the band. * Delivery about 170 OSG/br/unlt at 100 palg. A steady feed rate is important because fluctuations In the rate cause variations In the pot tenperatures. (The flow of fuel gas to the preheater pots Is set by hand, and con sequently Is not automatically readjusted to meet such . variations. The convertor pot temperature Is automatically controlled, via the fuel gas supply rate, but naturally the control lags, because of the sensible heat of the lead, the pot, and Its setting). * These rings are made far the stems of water valves, but they have been' found to serve better t-fcan other packings in the benzene pumps . Tbs slight softening 4 swellng of bb/s rubber 1 r. adra, -- "vnis . 1 DSW 462143 WATER PCB-SD0000050736 Equipment List, contd 85 #4 Feed Punps to the Conversion Phita, contd There Is a Fisher strainer on the line from the feed tank, item #4, and the purops deliver to twin filter vessels in the flow control room. See under "Process in Detail" above, page 30. #5 Conversion ITnlta Drawings. See list under "Diphenyl Process Drawing List" subtitle "Conversion Uhit", page 77, especially: 9C-9019-1 9C-9022 AA-7955 In November 1950, Anniston had 12 units, of which two were of the two-pot type with special tubular heat exchangers, to transfer heat from the converted vapours to the in coming benzene vapour, and the remainder were of the 3-i>ot type. There were numerous minor differences from one unit to another, representing different stages in the design history of the process. The 3*i>ot units are preferred, and will be described in this report, because the design of the Newport plant is to be based on the 3~i>ot system. A 3-pot unit consists of three essentially similar stain less steel vessels, about 6*-0" deep by l'-10-3/4" I.D. (see drawings listed above), each set in its own gas-fired furnace. The benzene vapour passes through the three in series, called respectively 1st preheater, 2nd preheater, and convertor. The 1st preheater is in the middle of the row of three pots, with the 2nd preheater on one side of it and the convertor on the other side of it. The furnaces are of firebrick, with steel casing. Each setting has eight gas burners, horizontally directed tangentially into it, four at one level and four at another, evenly spaced round the circumference, and there is a pres sure gauge (about 22 p.s.l.g.) on each burner, between the control valve and the Jet. The flue gases from pots 2 & 3 go downwards to cross flues leading to the bottom of the setting of pot #1, then up Inside that casing, and out to a flue leading to the bottom of the setting of the vapour!zer, item #6a. There is a steel door, with stabbing hole, at the bottom of each setting, for running out lead In ease of leakage from a pot, and there are access doors In the flues where needed. ; DSW 462144 WATER PCB-SD0000050737 Equipment List, contd 86 #5 Conversion gbits, conM By ste&nfl of Gaspers, the flue gas a my be by-passed away fro* the vapourlxer by passing them Into a Tent stack* The temperatures of the three pots are discussed under "Process In Detail" on pages 33 and 51 above, and It sill be seen that the #1 pot Is at a such lover temperature than the other two. Actually the waste heat fro* pots 2 and 3 Is almost enough for #1, but the burners In the #1 setting have a little work to do, and of course, would be needed If the pot had to be re-started fro* cold. Benzene fro* the recycle feed tank, lte* #3# la forced by the pumps. Item #4, against the back pressure of the whole conversion syatea, to the top of the vapourlxer, lte* #6a. It Is vapourized in passing down the colls, and the vapour bubbles through the lead In the three pots In series, then passes, via the lead trap, lte* #7, to the bottom of the fraotlonatlng eoluan, item #8. Bach lead pot has a dip pipe reaching nearly to the bottom of the pot, and the upper part of the pot has circular baffles to diminish the entrainment of the lead. A great deal of experimentation has been devoted to such problems as the design of the dip plys and of distributors, (see below), for use at Its lower end, the design of the baffles, the choice of stainless steels for the differ ent parts of the equipment, and the choice of gaskets for the pot covers. Something of this can be seen fro* the "Vote on the Maintenance of the Lead Pots", on pages 4l- above. Bo attempt is made here to review the whole history, but only to describe the arrangement now used, with present-day ideas for Improvement. Tforlous old patterns of pots etc are still giving good service at Anniston. The pots, and the fittings Inside them, exposed to the hot lead,A(but not the covers), are of stainless steel. See pages 88 & 89 below for the choice of the type of stainless steel, also page 22 above. Anniston use steel covers for the pots. The #1 pots have flat covers, of steel plate welded on to the rim of the stainless pot. Some of these plate covers have heavy radial fins welded on for stiffness. The #2 pots have heavy oast steel domed covers, bolted on, with heavy cast steel (divided) backing flanges under the pot rims. The backing flange has four lugs, fitting Into sockets on the furnace roof and supporting the pot so that the rim Is several inches above the top plate of the furnace casing. DSW 462145 WATER PCB-SD0000050738 Equipment List, contd 8r #5 Conversion Units, contd The gasket In this ease Is 5/16" round soft annealed copper, with a square section asbestos gasket Inside it. The Jrd pot is similar In construction, but has a gasket of l" braided asbestos rope. Mr. Barden took a sample of this In August 1930. This thicker gasket is used beoause the edges of the flanges pull together in service at the high temperature, and they can remain in service longer with a thick gasket. To make the gasket,a pleoe of the asbestos rope Is out, long enough to lie Just inside the bolt circle, plus a few laches over-lap. The rads are overlapped, and the groove between the overlapped parts 1s stuffed with strands of asbestos rope, and the overlap bound with a few thin wires. Liquid glue is used to help in making the overlap and In holding the ring in place between the flanges during assembly. In us^ the Joint ring becomes completely flattened and hard. The use of restriction orifices after #3 pot, as discussed on page 34 above, has thrown much greater pressure on the Joint of the pot and the asbestos rope gaskets are giving much more trouble. Vertical sheet iron fences, about 6" deep. In two halves bolted together, are being loosely fitted around the cover Joints of some of the conversion pots to deflect any possible burst of flame due to a fail ure of a gasket. London's proposal to use thicker covers and backing flanges, and to make them of stainless steel having a higher tem perature yield point than the plain steal ones, is liked at Anniston, because It offers a chance of using copper gaskets on the conversion pots as well as on the preheaters. London's proposal to use slotted bolt holes is disliked as tending to favour the bending of the flanges. London also suggested supporting the pot rim on a continu ous ring instead of on the four lugs, but this arrangement would appear to make the cover bolts more difficult of access, to rob then of the cooling effect which exists under the Anniston arrangement, and to make the raising of the pot a little more difficult. At Anniston the ooapanlon flange under the pot rim is In two halves, bolted together, as mentioned above, faming a circle, with four lugs at right angles to one another. Fairs of short pieces of angle iron are welded inside the box frame which composes the furnace roof, these pairs forming guides for the lugs on the companion flange. Steel packing pieces are placed first on the bottom of these guides to give the necessary elevation of the pot. then DSW 462146 WATER_PCB-SD0000050739 Bquipment List, contd 68 #5 Conversion Italts, contd the pot lowered In and a cured by steel wedges driven In alongside the lugs. The pots should not be rigidly con nected to the furnace casing because the continuous vibra tion of the pots would be coassunloated to the brickwork of the furnace, and tend to sake it crumble. At Anniston they fit fairly closely Into the round holes In the furnace roofs, but in some eases steel sheet sectors are used to bridge the gap. in any case, the Joint is made good, either with clay, or more recently with a paste contain ing asbestos fiber (insulating compound 66), This helps to keep the pot rims a little cooler. The bolts in the pot cover Joint and in the adjoining pipe flanges, are of ordinary steel%at Anniston, and are never re-used, partly to save time and because they become stretched, but mainly because working conditions can be very uncomfortable because of the radiation of heat from other units. Howport, using stainless bolts, and having only one unit should be able to re-use the bolts. It is suggested to use nuts of dissimilar metal, to diminish galling, m cutting out the bolts on the pot cover, the Anniston welder uses an oxy-acetylene torch, and cuts through the base of the nu t, flush with the upper surface of the cover, and the stub of the bolt is then knocked out downwards, the pot cover being high enough over the furnace roof to leave room for this and for the insertion of the new bolts. On the pipe Joints, the bolts are out between the flanges. The cover bolts have square heads, but hexagon nuts, and they go through the flanges, tail upwards. Small squares of steel (old nuts will serve), are welded under the eoapaalon flange, between the bolt holes*to pre vent rotation of the bolt. The nuts are run an with an air-driven spanner, for speed, and tightened up after the pots are hot, with a 4' long ratchet spanner. Anniston advise strongly against attempting to use a tongue and groove Joint on the pot covers because of the risk of dis tortion. The attachment of the stainless pot rim (by welding) to the pot itself appears to be a line of weakness because the lead in the pot is very heavy and is gently busping and vibrating whenever the vapour is passing. Crack* at this point are not unknown. Anniston have increased the thickness of the rim of the pot) formerly 5/&n, now 1". British makers of stainless steel agree that 309 stainless steel tends to be brittle at the pot temperature, but that 330 does not. HCL are experimenting with pots fabricated, by welding, frost centrifugally oast sections of 330 stain less, welded either with 309 or with special electrodes. DSW 462147 WATER PCB-SD0000050740 Bqulpnaot List, ootxtd 89 #5 Conversion Calta, contd (See the letter fro* Michigan Steel Casting Co. to Annis ton, May 31/50). Anniston discussed this design, Aug.1/49. Heat treatment, after fabrication. Is recoanended for both 309 and 330 stainless. Inconel, Eureka, nickel, and allays rich in nickel are said to give poor service. Anniston reported, Aug.l ft 5/^9, that cast therao veils of 309 had proved to be brittle. Centrlfugally cast dip pipes of 309 were better than rolled, or sold cast, but ware brittle or becaae brittle In service. Centrlfugally oast dip plpes*have recently (1950) given excellent ser- >of 310 Ice. It Is noted that In this Method of casting the dross sta.in tends to fora on the Inside of the pipe, but no barn less appears to have resulted. Therao veils of east 330 have steel) recently given very good service. Carbon tends to build ip in any crevices in the Interior of the pots, and to exert a bursting pressure. Cast covers should be free fro* pinholes, blowholes and slag Inclusions. Anniston varn against voiding cover strips on the outside of velds on the pots. Outlet branches on the covers should not extend dovn Into the pot, and the Joint should be voided inside, to fill up any crevices. Sharp edges of Metal also tend to attract deposits of oarbon. Any distributor at the bottoa of the dip pipe Bust be such as will not increase Isplngesent of the gas on the pot vail, because of erosion. Anniston are using a sijg>le open-ended pipe vith a renewable liner for the last few Inches at the bottoa and four 5/3" holes drilled radially through the well about 4" fro* the bottoa. At a Meeting at Anniston, Mar.16/%9, a proposal was *ade to try a deflector which would discharge the vapours tan . gentlally in a horizontal direction. This It was thought would give the lead a swirling Motion, diminish vibration of the pots. Increase the contact .tine of the vapours, iHTtninh the size of the gas bubbles, and diminish the entrainment of lead. Harrow orifices must be avoided In the design of any deflector because of the building up of oarbon. This deflector had not been tried at Anniston, but London asked that the pots to be supplied for Hewport should have a deflector. Standard dip pipes vara, however, shipped in Hovead>er/*!>0 to avoid delay. DSW 462148 WATER_PCB-SD0000050741 Equipment List, eootd 90 #5 Copyera loo Units, eontd Texas City, Aug, 10/0(9, sent HCL a drawing of the distributor they had developed for pots for propane cracking. There la no heat Insulation used anywhere In the conversion system, except that the vapour pipe from second preheater to the convertor Is Insulated cs some of the units. It Is desirable to avoid loss of heat anywhere along the Hue of flow, up to the top of the convertor pot, and at one time the pot lids were Insulated. It was found, however, that greater distortion of the pot covers and bolts resulted. The Insulation of the vapour pipe from the second preheater to the conversion pot Is not an easy task. In sons cases, asbestos fiber. Inside a sectional (lobster back) oaslng is used, and In other cases a special Johns Monsvillo powdered product In a dust-proof steel oaslng. This Insulation and easing adds to the weight of the pipe, and may oauae sagging, and all these long pipes, from 2nd to 3rd pot, whether Insulated or not, have support stirrups, which, however, appear to be out of use. In addition to the vapour inlet and outlet and the thermo well, each pot cover has a fourth opening. On the conver tor pots this Is blanked, but cm the 1st and 2nd preheaters opening Is used to connect the balancing line. These connections run in 1" pipe vertically for about 71 -0",and then Join Into a horizontal header running to the bottom of the eolum, Item #8. There Is a pipe coupling on this line about 2'-0" above the pot cover, then a branch for a pressure gauge, then a stop valve, which Is normally closed. These gauges show the pressure In the space above the lead in pots 1 and 2, and the lines nay be used to by-pass a pot In emergency, and to prevent lead from being forced backwards Into the vapourixer. There is a pressure gauge on a branch In the vapour line from the convertor pot to the eoluomi, Also a branch for the saapling condenser des cribed wider "Control Tests" on page 75 above. There la a heavy I-beam about 18" deep, with 6" flanges, and with a chain block . . permanently Installed, running along It , one for each row of 3 pots, with a hinged section of the floor to permit lowering a pot or a lead trap to the ground floor. When a pot la put out of action opportunity la taken to clear debris from the bottom of the furnace. DSW 462149 WATER_PCB-SD0000050742 TO SHOW EQUALIZING LIM B ON LEAD POTS U 4 *o<A J I -f* I Equipment List, eontd #6a Y&pourleers Drawings. 9C-7845-5 9C-6159-1 9C-902* 9C-9025 . One per conversion unit. Each consists of a single coll or of two oolls In parallel. In a brick lined steel stack through which the flue gases pass. Benunepasses downwards through the oolls countercurrent to the flue gases. The rapourlsers first Installed at Anniston had two oolls In parallel, one inside the other with a common Inlet header, and a control valve on each exit, leading to a 7 pipe Into the first preheater. This arrangement was found Inconvenient, and has (except in the case of one unit) been abandoned In favour of a single coll. There was some doubt about the proper shoring of the flow between the two colls, and considerable trouble with car bon blockage. Anniston report 1642, Oct. 15/41* states that corrosion of the vapourizer oolls may occur by con densation of moisture from the flue gas, near the point of entry of the cool benzene. A small leakage from a vapourizer coll into the flue may easily pass unnoticed, because the benzene will probably not ignite in the stream of flue gas, and the leakoay cause serious losses before It Is discovered. As a safeguard against heavy losses a saaple of the flue gas Is aspirated every day from a point In the flue, insacdl- ately above the vapourizgp coll on each unit, and Is passed through a nlxture^equal parts of sulphuric and nitric acid. For this purpose a steel pips is per manently Installed In the vapourlser flue and all these pipes are carried down to a test bench on the ground floor. The sulphuric-nitric mixture Is put into small rubber stoppered bubbling bottles, which, for safety, are sheathed In simple steam-pipe sockets. The bench has a siaple water aspirator, and it is the night opera tor's duty to test all units. He slxply aspirates the gas through the add for about 10 minutes, then smalls the sold bottle. Blockage of the vapourlzer coll is rare (except by lead forced back from the first preheater). Traces of Impuri ties left by evaporation of the benzene are singly carried forward In the vapour stream. The flue gas must, of course, be by-passed round the va pourlzer when the benzene flow is stopped. A coil should give over a year's service before needing repair. 91 DSW 462151 WATER_PCB-SD0000050744 Equipment List, oontd By-Pass Pine Craving. 90-78*5-5. . One per conversion unit. Fitted with a dasper. To allow . flue gases to by-pass the vapourlser as neoeaaary. 92 #T Lead Trap Craving 90-81*3-3. One per conversion unit. This Is a centrifugal type separator for renoring any entrained lead. It Is a elosed, welded vessel which suat be taken out at inter vals to allow any lead to be melted out. (hi sons of the Anniston units, the lead trap Is a rela' tlvely snail steel vessel, so placed that lead can drain back frosi It to the convertor. In the newer units, the trap is larger and stands on a steel plate on the opera ting floor, which Is nearly at the level of the flanges of the lead pots. These traps take the form of flatbottomed cyclones 3*-6" high, 2*-0" diameter, with a oentral pipe 6" diameter, for the gas exit , and the usual tangential Inlet for gas, high up the side. It has been found best not to oarry the central pipe more than about 1*" down Inside the trap, otherwise its bottom end becomes choked too soon by lead and by solid debris (iron oxide, carbon, and shots of entrapped lead). In a trap, seen Hov/^0, the layer of entrained lead was 1*" deep, and there was considerable solid debris as well. After the lead bad been molted out this vessel was out open to permit the shortening of the Inner tube, and opportunity has taken to dig out the solid debris. Kote on Bytying the Lead Trap The unit Is shut down and allowed to cool a little, and some diphenyl etc. my run back into the lead trap and may take fire when the trap Is opened. Each cyclone has a lifting hook welded on, opposite the s Inleti this Is for the yard crane; Inside the shed a pot Is lifted by a stirrup bolted to the oentral outlet flange. Tls cyclone Is lowered to the ground floor, picked vp by the yard crane, and taken to a melt ing table In the scrap yard, and put on the table with the vapour inlet downwards, (vessel on Its side), and flumes of natural gas played on the pot until the lead DSW 462152 WATER PCB-SD0000050745 (To follow page 92.) table, etc. USED FOR CLEANUP THE LEAD TRAPS Sco4ft V* l-O' g^.O:t.x.?c. 3ide view Plaon .1 gqulpmsnt List, contd. 4ft Load Trap, contd runs out on to the table and down Into a lead Molting pot, from which It can bo run off at convenience into Molds UJce those used to enpty the pots. The Melting pot has a brick setting, and a proper gas burner, but the cyolono is heated sloply by open flames, confined some what by Means of sheets of corrugated iron. The natural gas has a high calorific value, so the Melting is quick. 93 The molds are the same as those used for emptying the lead pots, see page 39 above. Koto on Repairing Lead Trap k Column These units ore exposed to the blast of hot vapours opposite the gas inlets (especially the lead trap since the introduction of crifice plates in the vapour pipes fro* the convertor), and they wear thin. The unit being shut down, and purged with COg* external patches can be welded on. A useful trick first is to weld a pipe nipple on the center of the patch and to leave It open as a vent until the patch is welded to the vessel, then to put In a screw plug. Such repairs are done with the vessels In their working positions, but filled with CO2* The trap has to stand up to fairly heavy handling when being cleaned, apart from the high tonperature during ser vice, and it should be of heavy plating. Stainless is not necessary, and thin stainless would be unsuitable. #8 Traotlonatlng Columns on Conversion Units framing 9C-8131-*. One to each conversion unit. Vapour in at the bottom from lead trap #7 Benzene reflux in at the top from reflux box #12. Vapour out at the top to condenser #11. underflow out at bottom via carbon trap #9 to sump tanks #10. Thermo meter well at point of entry of reflux line to column, and one In the exit from the carbon trap. , The column has 12 plates, but the caps have been removed from the lowest 2 plates because they became choked with entrained lead, etc. The lead would be above its melting point as it entered the colum. The need for so many plates, for a comparatively rough separation of benzene from diphenyl, has been questioned, but plant experience Indicates some difficulty In maintaining a reasonably DSW462154 WATER PCB-SD0000050747 Equipment List, COOtd #8 Tractionst.tag Colums on Conversion Utalts, eantd low content of diphenyl in the distillate and of benzene in the under flow of the colum. Presumably the plate efficiency la low because of the bydilution hydrogen and possibly there Is fouling of the plates by carbon oarrled forward In the rapor stream. The plates have the usual drain holes to prevent block age by freezing. Design data were unavailable. May l*/*8. The crude diphenyl tanks are heat Insulated, and any ateaa Jacketed lines outdoors* The ooluans have no heat insulation. Packed columns were used on the earlier units, but they became plugged with solid natter carried forward from x the pots. #9 Carbon Trap Drawing 90-81*3-3* Underflow in from colum. Side outlet at top to suup tank #10, with thermo well in the pipe bend. The outlet pipe goes down newly to the ground level before rising again to the susf> tanks, so forming a seal bend between the trap and the strop tank. This seal bend and all the piping are steam Jacketed, inner tube 2" diameter. These traps are removed from the unit for cleaning when ever necessary, about every third month. An opportunity may occur when the unit Is shut down for some other purpose. The temperature of the column trap is below the melting point of lead so that the lead that settles here Is usually in finely divided solid particles. However, due to the nature of the infusible, highly carbonaceous organic materials that settle also, the lead Is Intermingled and agglomerated^' loto a solid mass* Thus It Is necessary to burn and melt out the trap to recover the load. The proportion of carbon to lead Is greater In the "carbon* trap than In the "lead* trap. Anniston report 16*2 of Oct.l5/*l states that this carbon trap needs cleaning every 3-6 months, and that, although lit has a bottom outlet plug. It has to be taken out and cleaned by burning. As seen at Anniston In Hoveaiber 1950, these pots were simple vertical steel cylinders about 3r-0n deep by about 1*-6* diameter with about a 9" flanged neck DSW 462155 WATER PCB-SD0000050748 Equipment List. oocrtd #9 Carbon Trap, oontd to connect to the bottom of the colum. The cover also had an Inspection plug and two lifting eyes to facilitate handling for Maintenance. 95 #10 Suag) Tanks for Crude Diphenyl Horizontal, steel. Insulated tanks, underflow in, by gravity, from column #8 via carbon trap #9* Out by sub merged pusg> #101 to Diphenyl Still #13. All liquor lines thoroughly steam Jacketed. Vent to off-gas system. Steam coll connections. Three liquid depth Indicator. Vent with explosion disc. Mr. Ellenburg, Hay 14A8, stated that these tanks are operated at about 125C to prevent crystallization. When seen In Hovember 1950 the vent pipes were quite hot to the touch. - #1QA Submerged Pm For transferring crude diphenyl from stag) tank to diphenyl stills. One to each susg> tank. Explosion proof motor. #11 Condensers on Conversion units drawing 9C-9055. One to each conversion wilt. Tube and shell type, water through the tubes. Vapour In from column #8 and conden sate out to Reflux Split Box #12, through a strainer with by-pass. Balancing line from top of Reflux Box #12 to condenser, with connection to hydrogen off-gas system. There Is also an emergency vent to the atmosphere. 'Anniston advise against the use of copper tubes, because of the sulphur Ispurltiea in the benzene. They advise making this condenser of asqple capacity. #12 Reflux Split Box. r~ Drawing 9C-9055* DSVV 462156 Standard Rotaweir, with sight glasses and levelling screws. One to each conversion unit. Forward flow returns via sight glass to recycle feed tank #3. Sight glass In reflux line. Both lines have deep seal bends, and there must be sufficient head between the reflux box and the return Inlet to tbs colwm. WATER PCB-SD0000050749 Equipment List, coatd #13 Diphenyl Still Drawing 9C-825*. Two are Installed. Horizontal, Insulated, steel tank. Submerged ping) with water cooled bearings, and outlet to external gas-heated coll with branch to crude Santowax receiver #20. Inlet from #10 sump tank and "Heads and Tails" receiver #18. On one dished end Is return line from bottom of colon #14. Return line from heating coil. Vfcpour outlet to bottom of column #lk. k" rent pipe, from a branch on the vapour pipe to the colon, carried up through the roof. This vent pipe has a bursting disc between two flanges located about 6*-0" above the vapour pipe. In one case, the 6' length of pipe Is Jacketed (Jacket empty), and In the other case it is bare. immediately below the flange there Is a therao well, with a thermo cotple giving a continuous record on the Multipoint temperature recorder In the instrument room. It Is found that the discs slowly become perforated, probably by sulphur compounds, but even a small perforation Is detected by the tempera ture rise. The pipe above the flange Is supported on rods with a turn-buckle fitting by which the pipe can be lifted for easier replacement of the disc. The bursting discs are B.3.&B. dikes, type Al-Pb (lead on the concave surface) "Vinyl service . Bursting pressure 39 (In another case, 13) p.s.l.g., carried be tween two B.S.&B. type flanges. 96 #13B Heating Poll Drawings 90-8211, 90-8212. One for each still. Supported In a steel easing with firebrick lining. Gas fired. Liquid circulated by pump #13A from and to still #13. Plow is countercurrent to flue gases. Half the casing hinges away on a steel trade for aooess to the coll*. The circulating lines are steam Jacketed. Hote, however, that If water Is trapped in the Jackets and the hot diphenyl circulated, a very high pressure would develop In the Jackets. DSW 462157 WATER PCB-SD0000050750 (To follow page 96) il m --WiM-l----JIj'TnrfcfKCTlOn owni-ng ^ Front View FT I To distillate rxeervers DSW-46215S SUjVJdj BOX ITEM 17 WATER PCB-SD0000050751 Equipment List, contd #14 Coltgp Crawing 9C-7870. One for each still* 20 plate eoluan. Crain holes In plates. Vapour inlet at botto* fro* still #13* Be- flux return fron split box #16. underflow from botto* back to stHl #13. vapour out to condenser #15 Tin goose neck. Anniston foraarly used a packed eoluan, with a direct fired still, but the results were not so good as with the present bubble cap eoluan. On May 12/^9 Anniston expressed the opinion that the 20-plate eoluan, proposed by London Central Engineering Dept, would be adequate. Special care is needed If the reflux line to the still Is not to freese. 97 #15 Condenser Crewing 90-9872. One per still. Horizontal tube and shell. Vapour In froa top of ooltam #14. Condensate out to split box #16. Cooled by water through tubes, circulated in a closed system by puag> #23 through heat exchanger#?2 with make-up tank #24. #16 Reflux Split Box Drawings 90-8223 and 90-8224. Standard Rotawelr, with sight glasses, and levelling screws. One for each still. Condensate In from condenser #15. Reflux out to top of eoluan via sight glass. Forward flow to benzene receiver #21 or through box #17 to re ceivers #18 and #19* Vent pipe, steam Jacketed for occasional anltlng out, (should hare flaae arrestor). The reflux box should hare stoaa tracing for use while the diphenyl fraction Is running. #17 Swivel Boxes for Switching Diphenyl Still Distillates Crowing 90-7886 0ns for each still. Condensate In frost Reflux Spilt Bax #16, and out, by annual adjustment, to either "Heads and Tailsf receiver #18, or Diphenyl receiver #19. DSW 462159 WATER PCB-SD0000050752 (To follow page 97) r. <r IV. CL Equipment List, oont*d #18 "Beads and Tails Receiver. Drawing 9C-812*. One Tor each still. Vertical tank with subaerged ping) and aanhole. "Heeds" or "Tails" fraction of distillate in froa Swivel Split Flow Box #17* Ste&a coll Inlet end outlet. Submerged pusqp delivery line back to still. Vent line with lower portion ete&a Jacketed and carrying burstSqj disc. . Inspection hole with loose hinged cover. Hotton outlet. #18A Subaerged Puas> One per Heads and Tails receiver. For transferring Heads and Tails fraction to Stills #13 #19 Diphenyl Receiver Vertical tank fitted with 6ubaerged pump #19A. Diphenyl fraction in frow Swivel Split Flow Box #17* Finished product out via puE$> delivery line to Aroclor Plant, Store tank or Flaker. Vent with lower portion steam Jacketed and with bursting disc. Kai hole and cover. Handhole with loose hinged cover. Stean coll inlet and outlet. #19A Submerged Puag> One for each Diphenyl receiver #19* For transferr ing diphenyl to Aroclor Plant, Store Tank or to Flaker. Explosion proof actor. #20 Chrude Santowax Receiver Drawing 90-7877- Vertical tank with submerged punp #20A. Inlet froa subaerged punp #13A of still. Crude Santowax froa outlet of submerged pur%> #20 to #3*B of Santowax Dis tillation Unit. Handhole and loose hinged cover. Banhole and eov*r. Steam coil inlet end outlet. ' Stoaa Jacketed vent and burs^tag disc. Crude Santowax to drums through bottom outlet. Second stpaa Jacketed vent with hand valve. 98 DSW 462161 WATER_PCB-SD0000050754 Equipment List, eontd #20k Submerged Pump One far each crude Santowax receiver. Far transferring crude Santovax to #34B Santowax still charging tank. 99 #21 Buffer Storage Tank far recovered benzene. Vertical tank with nanhole and rent. Recovered benzene in fro* reflux split box #lb and out to ptap #21A. Inlet and outlet for ste&a coil. Gauge glees. #21A Pup Receiving x*ecovered benzene from buffer storage tank, #21, and delivering to recycle feed tank #3. Explosion proof aotor. #22 Beat Exchanger . One for each stl3L Receives hot circulating rater froa condenser #35 and delivers cooled rater to puqp #23. Cooling water through tubes. The circulating water is Maintained at 30-40C during the benzene forerun, and about 80C during the diphenyl - fraction, (by hand control on the fresh cooling water through the tubes In each ease). Compare page 47. #23 *JSL . One far each still. Receives circulating water froa heat exchanger #22 and delivers to condensers #15* Explosionproof Motor. #24 Make-up Mater Tank One far each still. Receives make-up pond water as necessary to replenish water lost by evaporation on cir culating through condenser #15* Bottom outlet to clrculatlng line. Overflow. Constant level device for con trolling receipt of Make-up water. Condensed water would, of course, be better for this make-XQ>. DSW 462162 WATER PCB-SD0000050755 Equipment List, oontd 100 #25 Cooler for Hydrogen Drawings 9C-80JK), Griscom Russel Co. NXB-14836. E&ohcf the four parallel units of the original cooler has been surrounded by a water Jacket to Increase the cooling surface. Receives hydrogen, for cooling, from a main serring all the convertor units. There Is a vent with hand controlled valve connected to the hydrogen Inlet via a benzene trap. The latter is a email pot containing baffles and drain ing to the surge tank #26. #26 Surge Tank for gydrogea framing 9C-8188 Receives cooled hydrogen and condensed benzene from cooler #25 and benzene from the vent to the hydrogen cooler. Hydrogen delivered to Inlet of compressor #29* Condensed benzene flows by gravity through bottom outlet to either of the two tanks #27. Pitted with 55" water seal*to provide pressure release If necessary. Water seal has a vent with flame arrestor. Line to Instrument controlling pressure within limits of 11-16 oz. Pressure gauge. Balancing line to benzene receivers #27. When the pressure drops to 11 oz., the compressor low stage becomes inoperative and only the high stage operates. When the pressure reaches 16 oz both stages operate until the pressure falls again to 11 oz. (per sq.ln.) * through the roof, #27 Benzene Receivers Drawing 9C-8188. One far each compressor. Receives condensed benzene from surge tank #26, from low pressure tank #31, and high pressure tank #33 through sight glasses. Benzene out via bottom outlet to punp #28. Balancing line to surge tank #26. Level gauges #28 Puap One far each compressor. Receiving condensed benzene fro* benzene receiver #27 and delivering to recycle feed tank #3- DSW 462163 WATER PCB-SD0000050756 Equipment List. oontd 101 #29 Two Stage gydrogen Compressors Brewings 9C-8188, 9C-7366, wiring and oontrols, revised Torsion sent to IIp. Buis, jan.l8/5o. Two are 1nstailed. The first stage compresses the hydro gen to 60#, and the second to 300# pressure. Provision for Introducing OOg for purging purposes. The compressors themselves are water cooled and there Is a swell verti cal tubular condenser after each stage, draining to the corresponding pressure tank*, Items #31 ft #33. Ir. Bstrdy stated, Boveaber Ij/kB, that the compressors have: 316 stainless shafts and valve discs, Bi-resist cylinder liners and valve cages, Ooetr gaskets, stainless steel or aluminum type, and that copper and its alloys are tndes treble. Bo copper or brass gaskets should be used, and generally copper and Its alloys are avoided. Hr. Clegoro, Hov/50, mentioned gaskets of Babbitt metal for the compressors Anniston were asked If un-lubricated cylinders with car bon rings would not be better, but they replied that finely divided oarbon, which might come over in the hydro gen stream, would abrade the carbon rings. Drawing 9C-7366 showing the wiring and control system for a single compres sor, was sent to London, May 2k/h8. The compressors are fed rather liberally with lubricating oil to cope with the possible condensation of benzene in the cylinders, but Anniston have not noticed any trouble from lubricating oil getting into the process stream. Mr. A. K. KLlenburg, June 30/48, gave a general discus sion of the operation and control of the compressors. Incidentally giving 150 and 300 p.s.l.g. as the two pressures, but stated that these valuss were not critical and that the choice turned on considerations of cylinder size and quantity of gas to be handled. Progressive unloading would be satisfactory as an alterna tive to the "trigger, on ft off" unloading control used at Anniston. By Hov./50 the solenoid unloading controls had been replaced by pneumatically operated controls, partly for reasons of safety. Electrical switches would ta any case be outside the compressor room. The room Is lllwrt gated by flood lamps placed a few yards outside the windows. DSW 462164 WATER PCB-SD0000050757 Equipment List, contd #30 Interstage Pooler One Tor each congress or, 81^>ly oonslsts of a vertloalj water -Jacketed pipe between outlet fro* first stage of compressor and Inlet to law pressure tank #31. Gas and vapor In at the top Aron the congressor. Condensate and gas out at the bottom to the low pressure tank. Item #31. 102 #31 Low Pressure Tank Erasing 9C-8188. One for each compressor. Receivea hydrogen at 60#, via Interstage cooler #30 from first stage of cospreaaar, and delivers to second stage of oospressor. Condensed benzene blown to benzene receiver #27. Bottom drain oock. Pressure gauge. The tank has a pop valve to open at 100 p.s.l.g. and a bursting dlsc fB.S.&B. aluminium, with lead facing am the pressure slde)>to open at 150 p.s.l.g. Both these have hooded vents well above the roof of the com pressor room. The vents should be well away Aram the roof louvres of the conversion rooms, and the still furnace room. A connection, originally installed be tween item #31 and the gas Inlet to the oospressor, was removed. #32 Cooler Similar to #30*and cools the gas passing from outlet of second stage of the compressor to the high pressure tank #33. #33 High Pressure Tank Drawing 90-8188. One for each otqpressor. Receives hydrogen at 300# via cooler #33 from second stage of compressor and delivers to BB-to department. Excess vented to atmosphere. Con densed benzene blown to condensed benzene receiver #27. Bottom drain oock. Pressure gauge. A connection, origi nally Installed between item #33 and Item #31. with a reducing valve, was removed. There Is a spring-opera ted release valve to atmosphere, opening at 300 lb gauge pressure, also a pop valve to open DSW 462165 WATER PCB-SD0000050758 Equipment List, contd #33 High Pressure Tank, contd at 380 lb and a rupture disc to open at 580 lb. Anala ton Installed float type traps to discharge conden sed benzene from the pressure tanks. Items #31 & 33, but discarded them because of leakage past the Valves. It is believed that traps made, for Instance of 316 stain less steel, with hardened seatings, sdght serve. At present, the benzene is discharged by means of hand- operated valves, as described on page above. 103 #3$B Charging Tank for Santowax Still Vertical tank vlth submerged pusg> #3^D. #3^B receives crude Santowax from crude Santowax receiver #20 via the submerged pva%> #2Q&. Return line from preheater #3kC. Delivery of submerged pung> Is to preheater or, for the preheated material, to the still #33 There is an inter connection, low down on the side of the tanks, between #3*A and #3%B, also a connection, about 8" from the top for use in oase the bottom connection should be frozen at the start of the heating process. Inlet and outlet of heating coll. Manhole with hinged Ud. #3Aa Charging Tank Similar to #3$B, but only' connections are the inter connecting line between the two. Vessel 3tA singly adds to the oapaolty provided by 34b, and it has no . ` put*) of its own. #31W Preheater A gas fired coil within a brick-lined steel casing. On side of firewall remote from the still. Entry of crude Santowax from the pun?) #3^D is at the teg) of the coll, and the flow Is countercurrent to the flue gases. There are four gas burners. #3*D Submerged Pump for circulation Tor circulation of crude Santowax through the Preheater #34c. Explosion proof motor. DSW 462166 WATER PCB-SD0000050759 8guipment List, eontd #35 Santouax Still Drawings 9C-8550 9C-8551 9C-8552 General Layouts, Horixontal tank with submerged puap to external gas heated eoll #35A Preheated charge in frost submerged punp #3tD of eharge tank #34b, Return line from the till heater #35A. underflow In from column #36. Delirery line from submerged ptap #35B to the still heater, #35A, or, for still bottoms, to the HB-40 de- partarat or to the pond. VHpour line to oolusn #36, with bursting disc* Inlet and outlet far steam coil. The tespereture spread In and out of the Santowax still eoll la normally 15-1&C on a Santowax R distillation. The pop has a capacity of 100 gpa and a 2n extra hoary pipe Is used for the eoll. 104 #35& Still Heater Similar to Preheater #34c, but Is longer and has eight gas burners. Gas supply controlled to giro correct tesperature spread between inlet and outlet of the eoll. #35B submerged Pmp For circulating still charge through heater colls and discharging bottoms to 'pond*. Explosion proof motor. #35 Ooltmn 12-plate column. 3 x 1/3" drain holes In each plate to prerent blockage by freezing. It Is thought 1 x " hole would be a better safeguard. Inlet for rapour at bottom from still #35, mist from reflux split box #38. Underflow from bottom back to still #35, vapour out to condenser #37. A much simpler column will serre If only Santowax R Is to be smde. #37 Condenser dsw 462167 Single pipe la form of slg-xag and Jacketed for water or steam, Reoelres rapour from top of colum #36 and dellrers condensate te reflux spilt box #33. It Is Ijpartaat to ensure that water cannot be trapped in the Jacket e^l raised to a rery high tespereture by the Santowax rapour. WATER PCB-SD0000050760 i Equipment List, oontd 105 #37 Condenser, oontd Actually, only a dribble of water Is used. In parallel flow with the vapour, and the water stress Is so adjusted that the reflux split box teaperature la around 200*0, (point #10 on the Multipoint reoordar} for Santowax R. An earlier condenser, of shell and tune design, using Aroclor as a cooling nedlua, was discarded in favour of tM sinplar condenser, because of maintenance troubles. There are no drawings, hut Mr. Ellenburg, Aug.23/48, wrote the following desorlptlom *Fhe inner tube is a standard 2" blaok iron pipe, Jacketed throughout its 451 of length by a 3" black iron pipe. --------by extending a well-Insulated 6" riser 8* above the top of the ooltom, ve were able to get 45* of condenser to connect to the original looatlon of the split box as follows: Pour 10' lengths of pipe were doubled back on eaoh other to give a hair-pin shape condenser (a descending slg-eag. E.&.), with a pitch of about 1" per foot. Vertical sections about 9* in length were used to conneot the long sections. A final vertical section of 2&' long connects the split flow box with the rig-zag part of the condenser. The inner pipe is Jacketed continuously to prevent any cold spots en the inner tube.-----------performance under the --- heaviest load 2000 lb total coadensate/hr---------- -----------For the ortho and Beta terphenyl fractions, where much lower condensate tesperatures are per missible ---------the water rate oan be Increased.----------It would be worth while to provide a blanked off connection for saapllng the condensate even if a complete (sajpllng; arrangement is not Installed." #3* Reflux Split Box Standard type, but no reflux is used in the purification of crude Santowax to Santowax R. #39 Santowax R Receiver A vertical tank operating under a pressure of 50 saa of mercury absolute. Receives Santowax R froa condenser #37 via #38. Inlet and outlet for steam coil. Connec tion to vacuum pop #40. Bottom outlet to the buffer vessel, item #39B. Valve for breaking vacuum. 33W 462168 WATER PCB-SD0000050761 Equipment List, contd #39& Submerged Ptnw? In Item #39B Delivers to the Aroclar plant, the Santowax R flalcer, or the HEMK) plant as required. 106 #39B Santowax Distillate Reoelrer Tertloal cylindrical vessel 5' x 5* Distillate In from. the receiver, Item #39, ssd out by submerged ptsq?, item #391. (forks under atmospheric pressure. Baa steam coll.) #40 Tacutni Pump Manufactured by Do Tine. Between Item #39 and the vacuum puaqp, Anniston have a sublimate trap and a vacuum tank in the line,to allow settling of particles. Vo filter is esployed because traces of terphenyl reaching the ptUM> are either dissolved in the oil or discharged to atmosphere, and do no harm. DSW 462169 WATER PCB-SD0000050762 Equipment List, eontd 107 instrumentation Sm Erasing 9C-9060. A. Conversion Chits Each convertor unit hast (a) A L.&H. ^cromax* recording-controlling Instrument for the tesperature of the convertor pot, which operates by controlling the gas supply to the pot. This same Instrument, still responding to the thermo couple on the convertor, also controls the gas supply to #2 preheater. If autoaatlo control should lead to towards a drop in temperature of the #2 preheater, than th* Is corrected hy turning down burners on the convertor. (b) A Foxboro "On-Off" type potentiometer temperature controller far the #2 preheater to prevent the tem perature of this pot from getting too high. All pilot lines are taken from the gas line to the #1 preheater, which only has manual control of the supply. The Foxboro controller Is a safety device which will work If the convertor control does not operate. The convertor control should cut the gas off on both Ho. 2 pot and the convertor pot If the teaperature gets too high In the convertor, but If the eontrol falls the Foxboro outs off the gas In Ho.2 pot. There Is no definite limit on this control, but It operates at approximate 76oC. (c) A L.&X. 'Hicromax' 12-point recording Instrument far temperatures at the following points: Vhpour from outer vapouriser 0011 to Vhpour from Inner vapouriser coll to Vhpour from #1 preheater. Vapour from #2 preheater. Vhpour from convertor. #2 preheater lead, top of column condenser water outflow flue gas to vapouriser 10 flue gas from vapouriser. #1 preheater lead, susp line from col: preheater. preheater. DSW 462170 The points for units #4 and jj have not been detailed here -- they are two-pot units of a different an! less satisfactory design. WATER PCB-SD0000050763 Equipment List, contd 108 A. Conversion. Otalta, oontd (d) A Poaebaro flow controller for the benzene supply to each xmit, activated by an orifice In the feed line, Fcibcro Pneumatic Pressure Transmitters have been Installed. In jjore&ber 1950 these were installed, with one spare, in a separate room which contained also the meters for the fuel gas, and which had a communicating door to the control room where most of the other instruments were located. These Indicate the flow and also transmit the ori fice pressures to the control Instrument. There arc also looated, within the Control Room, hand valves pressure gauges for each benzol line. See also page 31 above. B. Off-gas Treatment (a) A tespereture indicating Instrument responding to thermocouples at the following points on the Rydro- ccn cooler I) I21 3) Hydrogen In (2 points). ^drogea out (4 points). Cooling water out (4 points). (b) A controller responding to pressure changes at the hydrogen surge tank #26 and controlling solenoid (moire recently pneumatic) operated valves on the com pressors so as to maintain this pressure within the limits of 10-16 os. per sq.ln. Rubber fittings most not be useS/^erc^^e^will be exposed to the process gas,because of the swelling which would be caused by the benzene In the gas. C. Diphenyl Stills Bach still has: (a) A L.& X. 1 Micromax' controller similar to that used on the converter pot, but controlling, via the gas supply, the spread of tenperature between Inlet and outlet to the heating coll. (b) A L.& V. 'Hlcroanax1 multipoint recording instrument for temperatures of #1 Still Body, #2 Coll Inlet, this will be about the same as point #1. #3 Outlet circulating line. #4 Top of ooliaan. #5 Reflux to column. nsv\M62i7i u #6 (jcudenser Water. If the circulating puss should stop, the gets supply to the burners ie automatically etoppea- WATER PCB-SD0000050764 Equipment List, contd 109 D. Santouax 3tlll (a) A L.&H. tMlcrooBkX> Controller similar to that used on the convertor pot, bat controlling, via the gas apply the * spreadr of te^>ereture between Inlet and outlet to the heating coll. (b; A !-*? If. 'Mleroaax' multipoint recording Instrument for temperatures of: Still body. Heater Inlet, Heater outlet. Vapour leaving eolum. Top Inside of colum. Sight glass at reoelver. Bottom Inside of colum. Condensate In flow box. Preheater Inlet, Preheater outlet. Specimen charts from most of these Instruments were sent to Mr.V.M.Cooper, London, In October 19*17. & Anniston suggest uslng^glass wool trap, about 1 pint else, to protest Instrument lines from blockage by diphenyl etc., and using glycerin traps to protect instruments operating on benzene serrloes. DSW 462172 WATER PCB-SD0000050765 DOST q? BQPlPMKgr no Amistoo ertlmte for the Installation of 3 diphenyl units (fll, 12, fc 13), June 26/47, was given on page 53 of the re port hy Bavercroft & Mather, Septenber 19^7, and Is not repeated here, nor are the Installed cost figures for other parts of the plant vhleh were given on page 53a of that report. DSW 462173 WATER PCB-SD0000050766 OOMKE8T3 Off THE PROCESS 111 Sm "Earlier Reports" listed on page 2 above, and "ffote on Choice of Process" on page 6 shore. Effect of Tferiatlon in Benzene Quality See also notes on benzene under "Specifications & Test Methods, page 5* shore. Anniston report 1642, 0ct,15Al. states that, at that tine, ben zene with a distillation range of 79*4 to 8lC was used, but that the freezing point had to be 5.0C Mininun. Traces of toluene caused the fCroatian of styrene etc. which accumulated in the "heads" fraction of the diphenyl distillation. The amount of this was only about O.O653C of the diphenyl aade. Aliphatic Impurities were United in amount by the 5C specification. The naterlal contained 0.06 to 0.085^ total sulphur, but this wes not thought to cause any operating difficulty, and sulphur was found to delay the formation of carbon on the metal sur faces of the pots. Experimental work indicated that water in the benzene favoured the formation of carbon. (Excessive amounts, if ptsqped Into the vapourizor, may oause a violent development of steam, endangering the Joints of the equipment.) The report refers to work by F.E.Hubbard & R.A.Ewing, Sept.l7/4l, showing that fresh benzene gives 1.5 to 1.8 more conversion per pass than does recycled benzene, and other reports mention even greater differences. See for exanples pages 112 & 113 below. This effect might be explained In one of the following ways: (1) The fresh benzene may contain some material which pro motes the reaction. * (2) The recycled benzene may contain some material which fchlblts the reaction. (3) The recycled benzene may contain some material which destroys an inhibitor which is present In the fresh benzene. Explanation (3) seems to be ruled out by the observation that in ordinary routine working, with about 8o recycled benzene, and about 20j freah, the conversion goes quite well. As between (l) A (2) it is argued that some grades of freah benzene show even higher conversion rates and it has been shown by experiment that the addition of certain ehemloals does actually speed ig> the conversion. DSW 462174 WATER PCB-SD0000050767 Co--enta on the Process, oontd Bffeots of "Pro--tore" in Benzene 112 The report lay Conover cited on page 3 above gives the follow ing results In a laboratory unit. Proaotor Added 2% by weight Sons (recycled benzene) None (fresh benzene) Aoetle Acid Ethyl Aloohol Ethyl Ether n. Butyl Aloohol sec* Butyl Aloohol lso Butyl Aloohol Acetone % Conversion at 800C (dphenyl in product) 10 14.5 23.0 24.0 24.5 26.0 26.5 27.5 30.8 Effect of Size of Dose 2.0 la% Propyl Aloohol 1.0 0.5 0.2 0.15 0.1 0.0 % Conversion @ 750C 17.6 17.0 17.1 16.0 12.7 10.8 6.3 DSW 462175 WATER PCB-SD0000050768 Qcmaent* on the Process, oontd Effect of "Proaotora" In Benzene, oontd 113 Promoters have not been used at Anniston In manufacturing work. Conover's report reeoanends that if the use of prono tars in routine cenufaeture Is planned, a trial batch ought first to be carried right through to Arodors, to --sure the product does not contain any detrimental by-products. If the plant Is charged with fresh benzene the conversion figure for the standard conditions will be about but this figure will fall as the unchanged bensene Is recycled, and at the end of the day nay be as low as 10. The incoming benzene, though of very high purity, appears to oontaln enough of an unidentified promoter to hare a narked effect cm the rate of conversion. The falling off in the conversion rate would naturally be very inconvenient in routine a&nufaeture, so the feed of fresh benzene Is made practically continuously into the re cycled bensene tank, so that there Is always about 20 of fTeah benzene in the feed to the Tapourlzers. The reports Just quoted discuss the conversion reaction in orach greater detail. Impurities likely to occur in the fresh benzene are toluene, paraffins, cycloparaffins, and similar products, boiling around oOC. Pyrolysis of benzene which is contaminated with these will give diphenyl which Is likely to be contaminated with a variety of products, and unless these contaminants are removed before the diphenyl Is chlorinated to give Aroclors, the Aroelors made from the diphenyl are liable to be contaminated with chlor compounds in which the chlorine is more labile than in true Aroclors, Consequently the contaminated Aroolar may be zqjected on the very stringent tests lzposed by sons electrical users of Aroclors, The Diphenyl made by HCLis to be used to make Aroclors, and Mr. T, 0. Crane, Oct.17/47j estimated that about half the Aro- clcr production would be as 1254 for electrical uses, so the question of stability of the Aroclors Is of Importance to MCL. Apparently, however, the trouble has been limited to "Pyranols", mixtures of Aroolora with trichlorbenzene, and It la the trlehlorbenzene which has been mainly at fault. Mr, EUeriburg suggested that sure Intensive fractionation of the diphenyl might c<mq>ensate far the use of inferior benzene. Dr. Iferdy, Dec.23/47, stated that benzene contaminated with paraffins, would yield styrene and apparently vinyl biphenyls, and that these products, on chlorination, would yield materials DSW 462176 WATER PCB-SD0000050769 Consents on the Process, oontd 11* Effect of "Premotor* " In Benzene, oontd liable to oauae skin troubles. The chlorinated products night be stabilised by being held at high temperatures, but It was not certain that the stabilization would go far enough. Toluene on pyrolysis is said to give some styrene, and It was believed at one tine at Anniston that chlorine derivatives of styrene were oauslng skin troubles among the workers in the Aroelor plant. See page 16 above. In a letter dated April 2l/*9, Mr. Kllenburg listed alkylated benzenes and alkylated biphenyls as possible by-products from the used impure benzene, and he stated that styrene and stllbene have been identified In the "heeds" fraction of the diphenyl atill- Benzene hoaologues on pyrolysis give also dlbenzyl or deriva tives (M.Szwarc, "Bature", Sept.20/*7). Effect of 'Variation In Tlme-Teaperature of Exposure The process Is normally run with the lead in the 3rd pot at not above 8*0C and with a benzene flow rate of 170 U.S.Gal/tar., these conditions giving about l8$6 conversion of the benzene Into diphenyl, terpbenyls etc. The retention time correspon ding to the given flow rate, naturally depends on the pressures at which the pots are operated. As the process Is carried on day after day, the vapour passages become more and more restric ted by the deposition of carbon in them, and this raises the pressure in the pots,nd therefore the retention time Increases. Also, orifice plate restrictions are being tried In the exits from the 3rd pots. When the pressures Increase, from either cause, the practice Is to reduce the tesperatures enough to maintain the same conversion figure of about 18. Increasing pressures also give less entrainment of lead because of the lower linear velocity of the vapours, and the lower tempera tures should reduoe the corrosion of the pots, but Anniston experience is showing that these advantages may be dearly bought, at the oost of Increased maintenance of the pot covwjs, gaskets, etc. See also under "Process in Detail", page 3* above. DSW 462177 WATER PCBnSD0000050770 Cm--cnta on the Process, contd 115 Effect of Variation In Tiae-Teiiperature of Exposure, oontd Anniston report 1642 of Oot .15/41 sires the following state ment of the effect of variation In pot teaperatures i Temp, of lead In 1st pre heater Te^> of lead In 2nd preheater Teap of lead In convertor % Conversion to crude diphenyl #9 mil 58o 700 580 50 580 580 725 725 720 775 800 800 840 840 840 825 825 800 Carbon formation becomes excessive above 84oC. 18 20 18.0 to 19.0 17.7 to 17.9 21.1 to 21.9 17.9 to 19.3 The effect of variation In pressure in the lead pots was sttidied by Conover & Huff of Plant ^A*. see report Oct.5/35. At 800C tinder certain conditions they obtained 6.2% conversion at 5 paig* and 14.4% at 3$ palg* It Is stated that ratio of terphenyla to benzene in the reac tion product Is about the sane as the ratio of diphenyl to benzene. At 18% conversion, the ratio of diphenyl to ter- phenyls, etc., (high boilers) Is about 4.5 to 1, and the amount of carbon formation Is not too serious, but if the % conversion Is pushed beyond 18%. the ratio of hlghboilers to diphenyl In creases ^also the ratio of carbon to diphenyl, so 18% appears to be about the economic balance for American conditions, even though Anniston cannot sell or use the whole output of high boilers. See "Design Data for 4th Unit, " Aug.9/39# for a curve of costs v. % conversion. Raising the pot temperature gives an Increase In output of di phenyl with relatively little Increase In the consumption of fuel g&s and at times small variations in conversion rate have been made at Anniston to get a greater output of diphenyl, at the cost of a higher proportion of wastage as hlghboilers. Anniston report 1642 cltee a report by Durgln, Eeald A Perry. Nay 15/31# Indicating that If hlghboilers are recycled In the benzene, the amount of HB in the crude diphenyl is not increased, but that it has a higher boiling range. Anniston experience up to Hovnber/^0 Indicates that recycling of diphenyl In the benzene Increases the amount of oarbon formation. DSVV 462178 WATER PCB-SD0000050771 Comments on the Process, contd 116 The benzene distillate fro the diphenyl still is examined be fore being recycled to the conversion stage. It usually contains very little diphenyl, btit if It contained more than about 3% It would be fed bade gradually, or in an extreme case, returned to the heads h tails tank for redistillation. It was stated in September 1947 that conversion was reaching 13$6 in the second preheater and ld in the convertor. The extent to which the reaction proceeds in each of the 3 pots, naturally depends on the conditions of teog>erature and pressure in each pot. Anniston report 1642 gives the follow ing figures for vapour leaving the second preheater: 2nd preheater lead temperature jo conversion 64oC 720-725C 7750 1.0 4.0 to 5.0 10.0 Distillation of the Crude Diphenyl . By-products, which boil between benzene and diphenyl, build up, only very slowly, in the continuous recycling of the "heads" fraction, and it is not difficult to run the fractionation so as to give diphenyl of the necessary quality. Mr.Ellenburg, May 12, 1949, stated that about 0.1# to 0.2^ of the diphenyl yield was side -tracked In the form of the aH from the "beads" fraction. Anniston report 1642 gives 0.065# on the diphenyl as the amount of styrene, etc. formed. Variation in the diphenyl distillation conditions might also go some way to offset the effects of lapuritlos in the benzene used, see pages 55 & 113 above. benzene recovery froa the gydrogon Amis ton reported moderately good results from scrubbing the hydrogen with mineral oil. An activated alumina absorber was in stalled at Anniston, to take up benzene from the hydrogen, but It gave poor results and was discarded in favour of the present compression system. It is thought that the alumina was not pro perly re-activated. See Anniston reports 214, Sept.22/33, & 9*8, Aug.21/35. Since there is a use for the compressed hydrogen, the compres sion system appears to be the best for Anniston conditions. Originally the hydrogen at Anniston was piped to a gas main and burned, with natural gas, under the unit*, but difficulties in keeping the correct air-gas ratio in the burners soon caused a discontinuation of this use of hydrogen. DSW 462179 WATER PCB SD0000050772 -{ COST DATA Labour Requirement At Anniston in Hoveriber 1950, one operator per shift could look after up to 5 conversion units and one diphenyl still. In addition, there was one chief operator per shift, (but his responsibilities Included Aroclor & HB-tO,) and there was a general day foreman, (Curry), covering a number of depart ments, Including Aroclors. Hr, EUenburg wrote Aug.l8/48jabout the breakdown of the various service charges among the different products: Diphenyl Crude Diphenyl Distilled Steam, lbs City water, eu.ft. Recovered water, ou.ft. Gas, eu.ft. Electricity KWH 0.10 0.037 0.49 13.5 0.050 0.91 -- 4.4o 2.2 0.008 The units are given per pound of product -- for June/48 and he comments tiat there are seasonal variations in such figures 117 Repair Work The practice sheets, pages 120 ft 121 below, give a record of the frequency of various repair Jobs. osw 462180 WATER PCB-SD0000050773 UOMCAMTO CHCMICAi. OQMfAMY Q+i.9 1 1 118 O ffQ ANIC DIVISION M nc .~7 lx -*3 M 77 i MO DSW 462181 NfT VWM. Nrva *T*AM 1X WO n < Ml a 7* <* ri / *<" *v ' f*i r? 5* 4**0 K in 70 4t*i 74 < f-jjt --*Ti. >o* ^ 1 -*l 7f I* *w *>j Tf a** 47 >*13 070 - /ru *57/ >**j; *7* <o 4*7 V*-4<? 01V o~it *>/ W7 7/ *7f 7'>v / *. 4~* 47 47 ro 7u 775 tO 4S *< Om 4 t (X >* -SJ X <5T7 Itl Ml ' M/ t e if A0"7 *74 1 n e*r 4C X 7' < *44 r m r u> XJ fr* M X ~ 7xf w - 443 7* S IX 4n 'w -##r 0/M *0-77 -*r `**7 o<v -- H *o O <%* v 1 *04 f *0/4 '*+7 tif *01.7 oot OOf *0#*M *xr* XI -o>m -ot>v * 00/ 5 X*/ * Id* */4 017 *005 *04 4 /4 MM | 4< V ^ to *|t 4 *7 4* ( 1*4 -* /I n* ># X 01 1 t* 44 44% J Ml t/; -- 1 J Tf V,4 ei - X-* 3 4 /**/ , - Vt *4A . uf i >-* ' - *404 7tf "V i-X.f -oj i* X -<7 J'f <T> I ni| t%4 0 0% n -n !* 4# * */ . flf) * 4*70* ^ r^t p o -4 * So* 4M 1 , 4 * 4 c- >io f * 0/7 . * f m- 4 WATER PCB-SD0000050774 4 > ( d~i.lt, oKXUj > ^7- ) lI'fclMiflhl ^n.*cW/- Si 4tr TOT* , I DSW 462182 -1*. <j-o -7 vi v-7 T >4 Oa ru Sx 7 ** ot *< 733 >7 3t ^+ >7 / ST >7 16 1 oib o 002 JJI -o>t ;w 9 Ow V 'OS . 0 7* oW ' o/ -OM J*71 vf #/ V* - < b 4 / *S~ *-/ r ~i *6 4/ 71 V '1 n /f *o f7 i >7 A. a *7 e *7 *9* o / 0/*r o^s OOi 00| O r J fll 0*6 Of) *o,r i 1 i> > '-H > ^ > >7) r.>| ,<T 0*1 3a/ 067 -j l7 oif on V-M Of 5" v /J1 * {0* . v . W* . . **o **a-_ vif V- 7f - v*f 7 < *f **) - a * ^ <9 bt or - * 3~C e-r -OJ . T17 a*f To f 07 031 -.7 f . *7. v* * -- ta &i . **a .71 . - r`* >7! < W* . OTf - 1 V'l . i *rf . oo l 4i? l>+4 wf /*a v'* 6* .7t i a oo f f* WATER PCB-SD0000050775 i 12 j --1--l Jo-V" 1rv a."i f m*jj3mS iau ------- MO.Mfc l*rM r >sr WUaoi f a irw 171,3B1 i iifti *71 i.m i Sr 14. iOia 11.M6 teoalitr ' i.*9o 990,000 . flam 20.000 I 3.047 3,000,000 * 19.446 10.330.000 i 0.036 33,300 < ________:_______a Um. k.ri. 650 1.02 iTua 7 *.552 11.7*0 305 54 - * .. *T . -*3. : 7. .>* t c-r.- .:V PMLvtK* '* SWflBll J fcUl . * Poundj 7*641 ,.3*7 ^ Ik9. Qrit Vvr - 3.70* Salami 0 * 1.739.130 JW*oa * 137.409 `- fea# kstcrlsl* tteed >otol - JMrtwl^ 0 CMkmtlwl k Low - OoIIoob 1.369,036 1493,936 * - Jtwomr ffcrrlow Btwa - Um. Wtr-Cltj*-Cu. rt. toentltr 7,976,000 143,000 WUr-Beoor'd-Cu. rt. 30,403,000 ow - cu. rt. OMmoltr -HK 34,991,000 413,000 S3? OprtlrH Malntwonaot Pairing 11 fryi Wi ^Tr??rr 7'2 ab Shire1m 109 y~r Tto Date ... , < 01.99 10.71 . fvr Lb. 0.1793 0.1693 0.9446 ^1.044 .019 3.026 - 17.666 .034 T-h M. tT. 666 996 5.394 24.140 --i 100.0 0.65 99.37 0,70* Onlt-Eouro 6,696 ------------- ^42 Z&k 52 6.602 Oparatiiw Tine Bom CtlPl^r Bom - but Boro ajura-Chxw Orarated Bom - Mat Bcwre-Bet Operated Polar* * of Qromm Opoj Iwfcjt, knn | ' **^Ef*"I* isrfe" 3.464 49,024 508 4fl,3l6 6.60 93.40 92.43 1.04 Dol*r* 1 Clean Cocaanenlni (Tjm b) 11 2 Clean Cooaane>fra ( fwl i \ 7 Cleon Matrlbutr*- P 1 dun Cartx-n Trap 9 1 Cl nn Tjufcl Twip a Clean Coli 1 Clem Tsxnrlzor Coll1 ? 1 dwnm Blatrfkirfcrrr COI r. 12 O'*TM Imd Pert* Charuj* IJ Tif^i 1 4 Omim IWaivmnpl* 13 r^AriM r^vwanwiV Tiling K Ohanna Prnaaura OugM Repair Ralaaaa Mn. -1 1 11 UMlIeiMna data 41 * Mat aaalr teteaalr (TIam nr imj Cl** Twi Unaa c*--aa je rt Wat. --- J- t---- a ^v v- 14 ? 1 1 2 4 1 1 * L,-------------------------- . 6-1300 Pa&y 1 UTVilKUTT. RKPCHT t peo sn, <62i83 WATER PCB-SD0000050776 120a Jtr. Ellenburg, January 20/*8, gp.ro the following break down of the fuel gas consumption: cu.ft./Lb. Aroclar Distillation 0.66 Biphenyl - Crude Chits 16*7 Biphenyl - Distillation 2.2 Sautowax R Distillation 1.5 Calorific value of the gas about 10*0 B.th.TT./cu.ft. at 60P and atnospherlo pressure. OSVV 462184 WATER PCB-SD0000050777 121 c m 1 1 w c DSW 462185 WATER_PCB-SD0000050778 122 A*2.![Zi2LI><tM5S. fhie report la mainly a compilation of information drawn from Anniston reports and from discussions and correspondence with many members of the Annis ton staff, whose willing and able help Is grate fully remembered. The reports xaade by Messrs Sivercroft & Mather, September 1947, hare been Incorporated, also that of Mr. H. L. Harden, August 1950, and ob servations made by the author during visits to Anniston. DSW 462186 WATER PCB-SD0000050779 Index page 1 INDEX (Underlined figures denote the sore Important entry under any one subject) ACKNOWLEDGMENT..........................................................................................122 Alumina absorber oa hydrogen (abandoned)........................................116 Aroclor, electrical standard required 55,113 BENZENE, catalyst in, see "Benzene, effect of catalyst" danger froa overpunping Into feed tank . ... 30 danger froa water in ...... ...........................29,83 discharge from Ho pressure vessels .... 44,103 effect of catalyst In...........................................5,29-,111- effect of mpurltles in . 3, lo, 55.111- effect of promoters in........... .11? effect of sulphur In ........................... ...... .111 effect of water in, see "Benzene, water in" feed arrangements. ............ 29-,83 see also under "instruments" feed rate. ..................................... 31 feed rate during starting up...............................................37 filters........................................................... 30,31,85 handling ....................................................................................... 29-,83 leakage Into flues, method of detection described. ......91 mentioned. ..... 32,36,76 losses in the hydrogen stream, Anniston report mentioned ..... 3 physical properties of . .................................... .10,11 ptnf>s. .................................................................................................83,84 quality, effect of on Aroclor workers. . . 16,114 quality of, suggestion to do better frac tionation of diphenyl to com pensate for poorer benzene . 55,113,116 receiver from diphenyl still .........99 recovery from, off gas, see under "Hydrogen" recycled, reacts more slowly, see "Benzene, effect of oatalyst" recycled froa diphenyl still, examined . . .47,76 recycle feed tank. ................................................ .29,83 recycling..............................................................................5,29-, 111 recycling should be fairly uniform............................30 risk of flooding convertor with. ...... .30 DSW 462187 WATER PCB-SD0000050780 Benzene, specifications end test Methods. stare tanks. ........... store tanks, fire arrangements . . store tanks, objections to internal coils in. ...... sulphur in............ test records ........... trap failure in Hj> system. .... trapping of water froa ...... water in. dangers froa ...... effects of ...... . sources of ...... . Buildings see also under 'Ventilation'' Bursting discs, on diphenyl stills......................., on Eg vessels ........ By-Pass flue, see under "Vapourizer" Index P* 2 ... 84 . .5*, 111 ... 5^ 29.85 .29,83 29,111 ... 29 ... 82 . .18,96 18,102,133 CAKBOH, build-up, bursting pressure of. ...... . . 89 build-up, favoured by presence of water . . . .111 build-tp, favoured by recycling diphenyl. . .US build-tp, in the conversion system. . . . 3*,114,V' burning out froa pots and pipes ...... . . *1 detection of blockages. .......... . . 41 dioxide, connection to oonvertor. . . . 19,31 dioxide, handling ............. 31,38 dioxide, not used in diphenyl still .... .18,48 dioxide, suggested use to resaove carbon . . . 41 . .dioxide, use in checking conversion system, 37 intentional addition to lead pots suggested 22 precautions in burning out (damage to equipment). . . . . 4l (risk of poisoning by lead vapour). . . ............................................ . . 41 suggested removal by C02................................. . . 4l tetrachloride in fire prevention. . . . . 17, trap, cleaning of............. trap, description of............................................ . 94 trap, operation of. ............................................................ 43 Catalyst in fresh benzene, see under "Benzene, catalyst in" Charts, see under "instruments" Chlorination of diphenyl, Mentioned ........ 7 see also tinder "Aroclor" Chlorination of Santowax ............. Chlorine, determination of, in Santowax ...... Choice of process .... ............................ ....... d 6 Column on oonvertor, arrangement of........................... . . ?oonvertor, choking of.......... 93, 5 convertor, efficiency of............................................ 93- DSW 462188 WATER PCB-SD0000050781 Index page 3 Colum on convertor, operation of. ...... convertor, packed colum tmsuitable. .................... convertor, repair of....................... > . 93 diphenyl atlll ......................................... , 97 Santow&x still ........... . . . .104 Cements on process. ............. . . . .Ill Compressors, see tinder "hydrogen, compressors" Condensers on conversion units, description of .... 95 operation of ' . . . . 43 on diphenyl stills, description of. operation of. . .... 97 .... 47 on Santowax still ......... . . *9,104- Control tests. ................ *3,*7,Z,U6 Conversion pots, haffles ............................ baffles, sagging of .... . causes of failure of. .... cover ............ .... 33 ........ 41 .*1,*2,8*61- covers, bolts of, cut,In. . repair work ...... 88 covers. Joints, see under "Gaskets" covers, warping of ....... cracking of well between pot & rlw depth of lead In ........ description of....................... design of dlp-plpes ....... emptying of .......... . equalizing lines on............................. flue arrangements............................... leakage of lead or vapour from, detection of ................... Maintenance of novlng without emptying................... / operation of overheating of upper walls. . . pressure gauges, see tinder "Equalizing lines" also . ........................ . *2 pressures in repairing.............................................. rims thickened........................... rim welds craoking. ............................ rotation of between different settings. ....... . 88 *1,88 . *2 setting of......................................... 33,88 see also, "Conversion Pots, flue arrangements" starting of................................ stopping of. 30 temperatures of......... 33,3*,51 temperature control methods ... . . .107 DSW 462189 WATER PCB-SD0000050782 Index page 4 Conversion pots, tesperature controllore.................................34-,107 to be free fro* casings. ....... .~8& vibration of, Seo 'Vibration" Conversion process, adjustment of teoperature to oospensate for pressure varia tions. See "Pressure increases, compensation for" degree of conversion In each of the 3 pots. ..... .116 description of ....................................... ....32 effect of variables In ...... . .111- tvo pot system mentioned .... 32,82,85 Conversion test, details of..........................................................................75 mentioned ............................................................................43 Cooling fans............................................................................. .19*20 Copper and alloys, failure of In Hj system. . . 24,44,101,103 not recoanended for condensers ....................... ......95 Corrosion, effect of lfqaurltles inlead feared. ..... 57 general note ....... ................................................. 22- of convertor pots. .......................................................................4l of convertor pots by lead blast (Anrfaton report). .......................................................3 try stainless steel for traps InBg system . .103 Cost of equipment ........................................................................ .110 Cost of manufacture..................................................................................................117- manufacture, practice sheets. ......... .120- manufacture. Teraus % conversion. .......................................115 Cost sheet, diphenyl................................ .118 Cost sheet, Santowax R. . . ..................................................... .119 Crude diphenyl, analysis method for ...... ,....76 ccoposltlon of, see "Plow Sheet" tanks, description of ................................ 95 Crude Santowax, see "Santowax, Crude" IESXGH DATA BOOKS (Anniston), listed..................................................3,4 Design data book, (London) discussedby Anniston staff. . .4 Diphenyl distillation, consents on. .......... .116 control tests. ...... .47,76,116 cut points................................. 45,46,52,116 out points, variation of suggested to compen sate for poor benzene . . .55*113*116 description of ......... . 45- fire hazard In .......... 17 heads fraction, amount eliminated . .28,48,116 reflux ratios 45>47*52 Diphenyl, NCL do not expect to sell........................... .7*48 physical properties of,........................................11,12 specifications and test methods............................................56 DSW 462190 WATER PCB-SD0000050783 Index page 5 Diphenyl still, benzene receiver of, described, , .99 bottoms, see "Crude Santowax" bursting disc on, . ............................ . , . 96 colum, see under "Columns" :condenser described condenser. Indirect, cooling of , 97 condenser, temp. control of . ................... condenser, heat exchanger on water of,. . condenser, crude Santowax receiver, . is? description of ........................ ...... diphenyl receiver .......................................... . 98 distillate switch box...................................... . .97 formerly direct fired heads & tails receiver, . ................... heat cut off If prop falls....................... , I 108 heat exchanger on condenser water , . 99 heating coll........................ . 96 precautions In charging discharging. ...... . 48 starting ........ 18, HQ safety disc, see "Diphenyl Still, bursting disc". tesperature control .......... temperature recorder. ...................................... test records. ................................................... 52,53 Distillation, see "Diphenyl distillation" and "Santowax distillation" Drawing list. ................ ................... . .77 EARLIER REPORTS, see "Previous reports" Earthing of vessels 17,84 Effect of variables, benzene quality. .......... Ill catalyst, see "Benzene, effect of catalyst In" pressure ............................................ 34,87,114,115 proootara.............. 112 - recycled benzene................................ 112 recycled diphenyl. ......... 115 recycled high boilers. ....... 115 tine-temperature .......... 114 - Efficiency figures, see "Cost of manufacture" Anniston report on losses of benzene In off gas. .......... 3 Electrical use of Aroolors, see "Aroclars, electrical standards" Esptying of lead pots, see "Conversion pots, emptying of" Equalizing lines on convertors, see "Conversion pots, equalizing lines on" Equipment Cost. .......... ............................ ..... 110 Equipment list...................................................................................................... 83- Equlpment list, general notes on. ............................................................82 Extinguishes^?, see "Fire Fighting" DSW 462191 WATER PCB-SD0000050784 Index page 6 FENCE, round the pot lid. ...................18,87 Tire equipment.................................................................................... 18- Tlre precautions, Anniston report olted ......... 16 Tire screens, mentioned . ..................................... .19,75 Flanges In pipe lines ............................................................. .... 82 Flooring ............................................................................................ 19,20,82 Float controllers in benzene recycle feed tank, see "Benzene feed arrangements " Flow controllers, see "Instruments" Flow sheet of materials ................ 25- Fraotionatlng columns, see "Colons" Fuel gas, burners....................... .85 calorific value of. ...........................................................120a calorific value, mentioned ...........93 connections ............. ................... 36 consumption of. ............. 137,120a consunption of, rebat>ecl to pot temperatures .35,115 control ...... ............................ ..... 36,107 meters. .....................................................................................31,3b GAS, see "Fuel gas" Gaskets .................................................................................................................................24 Gaskets, making of for pot covers............................................................87 German reports listed ............................ ........... 2,3,6 HB-40, mentioned ............................ ..... ....................... ...7,18 Beads & tails fraction, build-up of by-produot In, see "Diphenyl distillation, heads fraction amount eliminated" Heads & tails fraction, composition of, see under "Flow Sheet" Heads & tails receiver, described . ................................ ... 98 Heat balance. .............................................................................. ... 15 Heat exchanger on condenser water, diphenyl still . ... 99 Heating coll for diphenyl still, described. . . . ... 96 operation of, see "Diphenyl distilla tion, description" Heating coils for S&ntowax still ......... 103,104 Heat of reaction. ................. ... 35 ...Heat treatment of stainless steel recommended . . . 22 ...Health considerations ........ ............................ 16 See also "Welfare" skin troubles in Aroclor. ...... 16,96, High boilers, ratio of to diphenyl, produced. . . see required. Hydrogen, alumina absorber for benzene abandoned. . . . .IK benzene recovery from (Anniston reports), 3 benzene recovery from, consents on. . . . . .us carbon dioxide contamination of. .... .. 19,38 oespression, choice of pressures at two stages......... . .44,101 CCMD V" CCMO CO Q WATER PCB-SD0000050785 Index page 7 Hydrogen, compression, now sheet of* Materials, see under "Flow Sheet* compressor, described. ........ compressor, lubrication of ..... . ........ 101 101 compressor, operation of................................ . . .44,101 Coolers.................................................................................. . . 100,102 . ,cooling and compression. ................................. . . .44,101 dangers........................................................... 18 12,101 formerly burnt with fuel gas . . . ... 116 oxygen determination unnecessary . . . . . 19,44 pressure control .44 100,101,108 . . , ?receivers described. . . ................................ 100 10 ....surge tank ................................. 100 temperature recorder on compressor . . .... 108 traps on instrument lines................................. traps for benzene. 44, uses for. . ................................ . 7, vent connections from crude diphenyl etc. tanks. . . . 19 INTERMEDIATE FRACTIONS, composition of, see under "Flow Sheet" Instruments........................................................................... 107- benzene flow controllers ......... 31,108 charts sent to J5CL ............................................34,50,109 conversion pot temperature controller. . . 34,107 conversion unit temperature recorder . . . 34,107 diphenyl still ............................................45,108 diphenyl still, cooling water, no longer Instrumented. ..... 47 float controllers in benzene feed vessel, "see "Benzene feed arrangementsB Santowax (till . ............................................ .... 50,109 Insulating compound *66" for pot furnace Joint. ..... 88 Insulation..................................................................................................................90,94 LABOUR REQUIREMENT....................................................................................... . . .117 Lead, depth of. In pots ... 32 effects of impurities In. ................................................. . .22,57 emptying pots............................................ . . . 39 entrainment from pots............................................................ . .36,41 farced back to vapourlzer ................................................ . .38,91 see also, "Conversion pots, equalizing lines" . .handling (roping off the area). ....... 20,29 leakage Into fumaoes ............ 23,42,85 detection of .... * . . . . 42 making ip losses of............................ molds................................................................................................. . .32,40 22*'93 DSW 4621 S3 WATER PCB-SD0000050786 Index page 8 Lead, oxide, corrosion of pots by. ......... .22,23 poisoning, little risk of. ......19 pots, see "Conversion pots" quality of .................... 7 resulting........................... 40 spectrographic analysis of, mentioned ...... 57 Lead trap, cleaning o described...................................... .... 92 mentioned...................................... .... 39 description of. .........92 central tube of, shortened.................................. 92 must be heavy steel ...93 need not be stainless ........................................... ... 93 operation of. ...........43 repairing of..................................................................................93 KAIHTEHAITCE 0F LEAD POTS ............. 39,41,89 Hake-tp water tank on diphenyl still condensing system..............................99 Monel badly corroded under pot conditions........ 22 Montar 9......................*........................................................................... 7,* Motors, type of. ........ ............................................. 1 HXCKEL, and alloys rich in nickel, badly corroded under pot conditions ...... 22 OFF GAS, see "Hydrogen" Operation Manual, (not yet written at Anniston). Operating records, diphenyl still. ....... ,. . . 4 45 52,53 conversion unit ..... ... 51 Orifice plates on units, details of. ........................... effect on pressure. . . see also under "Effect of Variables" roaring of gas through. . . . 35 Outlets for products, as cable box coapound .... . 7 electrical use as Arodor. . *$ as firelighters (not suitable) general................................ . . .7 in HB-40....................................................... . * . 7 hydrogen ........... . .7,116 roofing compounds............................ . . . 7 Outline of process. ........ ...................... ... . . . 5 PACKEHO, diphenyl melted into drums Is inconvenient . .48 Santowax H Is now flaked ................................................ . *50 Patent position, Anniston report cited. ...........................................9 on high chrome alloys for lead pots. . .22 DSW 462194 WATER PCB-SD0000050787 Index page 9 Pipe lines Jacketed. ...........................................................................82,83*95 See also under "Water In Jackets" Pipe lines traced.................................................................................................82 Preheater pot, see under "Conversion pots" Preheater, tubular, see under "Conversion process, two- pot system". Pressure Increase by carbon deposition ....... 34,114 Pressure, compensation for by altering teaperatures. 34,114 Pressures throughout the conversion unit..............................*34,35 checking with C02. ... 37 Previous reports listed. ................................................................................ 2 - Process In detail. .................................................................................................29 - Promotors in benzene, see under "Effect of variables" Properties of Materials. ............... . 10- Protective clothing. ................................................................19 RATIO of high boilers to diphenyl, see under "High boilers" Record charts, see under "instruments" Records, finished product tests, see under "Specifications & Test Methods" operating, see under "Operating Records" raw material tests, see under "Specifications & Test Methods" Recycle feed tank, see under "Benzene" Recycling diphenyl to convertors, see under "Effect of variables" high boilers to convertors, see under "Effect of variables" Reflux split box on conversion units described. .... 95 on diphenyl still described........................... 97 on Santowax still described. ..... 105 Repairing column and lead trap. ...... ............................ convertor pot Repair work figures .................. 117 Reports (earlier) listed, see under "Previous Reports" Roofing compound from chlorinated Santowax. ........................... 7 DSW 462195 SAFETY CONS HERATIONS ......................................................................................16- hydrogen .....................................................18,19,101 showers ............. 19 Sasples sent to MCL listed............................... 21 Santowax, crude, analysis of by dietillation. . . 76 charge tanks for still ..................................... 103 conposition of, see under "Flow Sheet" disposal of.......................... 7,48- diatillation of...................................................... 5,49- distIllation yield ................................................. 7 >m.rvn Ing Of. .....* 48-- hold points of..........* 13,50 physical properties of... ...................... 13 receiver described .......... 98 WATER PCB-SD0000050788 Index peg 10 Santowax ,crude, still..................................................................... . 104 column on* ........ . 104 condenser on ...... . 49,104 designed for separation of the Isoners ... . 49 earlier design Mentioned heater described . . . 1 operation of. ..... 49 preheater described. ... 103 receiver ...................................... . 105 tesperature controller . . 109 temperature recorder ... vacuum vessel. ...... Santowax Isomers, amounts sold......................................... 8 Anniston report on separation of cited....................... 2 outline of separation process .... 5 specifications and test Methods ... 70 Santowax O.M.&P. see "Santowax Isomers" Santowax R, flaking of.........................................................................5,49,87 handling of......................... packing of.......................... 50 50 records of test results ..................................... 69 specifications and test methods ...... 64 Scale of working. ............ ................................. 8 Skin troubles attributed to poor benzene. .... 16,56,114 Specifications and test methods ...............................................................54- Stalnless steel pots, oarburatlon of surface suspected. .... 22 corrosion, see under "Corrosion" effect of lxpurltles In lead, see under "Lead, effect of impurities In* effect of lead oxide on, see under "Lead, oxide" heat treatment recommended. ... 22 types of stainless steel 22-,86,88,89 Steel floors, advantages of open grids. ..... .19,28,82 Still bottoms, see "Santowax, crude" Styrene, effects of on workers, see under "Skin troubles" Sulphur In benzene, see under "Copper" Sulphur delays the attack of benzene etc. on pots ... Ill Sulphur, determination of. In Santowox........................... 68 Stop tanks, see "Crude diphenyl tanks" t TEMPERATURES, adjustment of to oospensate for variations in pressure ..... *34,114 Te^erature control, detection of overpimping of benzene to converters. .... .30 DSW 462190 WATER PCB-SD0000050789 Index page 11 Teiverature control In diphenyl still. ........ 45 TMg>er&ture controllers, see under "instruments" Temperature, effect of variation of, see under "Effect of Variables" Tesperatures at various points In the conversion ' ays tea 22,24,51 Teaperature of pot, varying with tbs benzene rate. . 84 lagging behind controller. .... 84 Texpbenyls, see "Santowax Isomers" Test method,, see "Control Tests = and "Specifications and Test Kethcds" Thermal balance, see "Heat Balance* Tube process, see under ^Berman Reports" and under "Choice of Process" Tubular preheater, see under "Conversion process, two-pot system" USES for products, see under "Outlets" VACUUM pimp for Santowax still described. ...... 106 Vhcuua still, see "Santowax crude still" vessel, see "Santowax crude still vacuum vessel" Valves. ........... ........................... ...... 82 Vapourlxor, blockage by lead, see "Equalizing Lines" and "Lead forced back to vapourlzer" blockage by solid, unlikely.................................. coll life ................................................................ 30 91 described ...................................................................... 91 flue arrangements.......................... 32,37,86,91- leakage of benzene Into, see "Benzene, leakage Into flues" operation of. .....................................................................*32,37 overpimping of benzene Into ....... 30 pressure drop through.................................................. 34 temperatures in.......................................... 2s " Ventilation ............. .............................. 16,19582 Venting of conversion unit...................................................................37,2^ Vibration of pots ............................ ........... .41,88 HATER In benzene, see "Benzene, water In" Jackets of pipe lines, danger from .... 96, 104 see also under "Pipe lines" Welfare ....... ,....................................... ........ 19 See also under "Health" Welding, fire hazards In. ............. .17,18 beat treatment recommended after ....... 22 on oolunm and lead trap.................................................. . 93 YIELD DATA, see "Flow Sheet" DSW 462197 WATER PCB-SD0000050790