Document JJvoywywMq6Xq0op4M0Zj3NNv

DownloadSearchable PDFRandom document
SOUTH CHARLESTON PLANT UNION CARBIDE CORPORATION CHEMICALS AND PLASTICS P.O. BOX 00, SOUTH CHARLESTON, W. VA, 25303 January 7, 1977 VCt' IjJj fa ffQ. RECEIVED JAN 11 1977 J.F.E. Ms. Susan Wyatt Emission Standards and Engineering Office of Air Quality Planning and Standards United States Environmental Protection Agency Research Triangle Park North Carolina 27711 Subject: Additional Data Union Carbide Corporation's Request for Approval of Equivalency 861. 65 (b) (4) Leakage from Relief Valves Dear Ms. Wyatt: In answer to your telephone inquiry regarding Union Carbide Corporation's request for approval of equivalency 861. 65 (b) (4) Leakage from Relief Valv s dated November 19, 1976 we offer the following additional information: Union Carbide Corporation's proposal to use "O" ring seat pressure seal safety relief valves in vinyl chloride service is based upon the use of "O" ring gaskets in vinyl chloride service since 1952 and the use of the "O" ring seat pressure seal safety relief valve since 1969. The only innovation involved is the use of the "O" ring seat pressure seal safety relief valve in vinyl chloride service. (See Attachments I and 11 for data on conventional relief valves and the "O" ring seat pressure seal relief valves). Union Carbide Corporation has been engaged in the manufacture of organic chemicals and resins since before 1930. During this period it has established a base of safety technology and engineering standards which is widely recognized throughout the industry. We would not propose a safety relief valve system which did not meet our own and governmental agency standards for performance, longevity and safety. Union Carbide Corporation's relief valve comments on the EPA National Emission Standard for Vinyl Chloride dealt with its concern for process safety. You wished to know what documentation we had that "O" rings can be made of materials which are essentially impervious to chemical attack by vinyl chloride monomer. You wanted published information as opposed to a statement by Union Carbide Corporation. occ 052382 Letter to Ms. Susan Wyatt January 7, 1977 Page Two Union Carbide Corporation has been unable to obtain any published data on vinyl chloride resistance by synthetic elastomers used for the manufacture of "O" ring gaskets. Experience has shown that vendor's information and published reports, if available, are not satisfactory material selection criteria unless they encompass laboratory testing, trial installations, followup and actual operating experience for the specific application. To meet its needs Union Carbide Corporation has a materials selection group as an integral part of its Engineering Division whose sole function is to study materials resistance to chemical and corrosion attack and to recommend specific materials for various applications. This group screens materials by laboratory testing and then tests promising ones under plant conditions. This program gives good reliability in selection of materials for process equipment. Attached is a tabulation of Union Carbide Corporation's laboratory test data on gasket materials (Attachment III) exposed to liquid vinyl chloride for 111 days at 33#F, Tests show effect on hardness, volume and weight change as well as visual inspection. These items are summarized in a resistance rating. "R" means good resistance for any application. "F" means fair resistance, use depends on the specific service. "X" means unsuitable. For "O" ring gaskets the material must be resilient, not cold flow and resist solvation; thus, the potential "O" ring gaskets are Viton, Neoprene, Buna N, Hypalon and EPR. Plant tests and experience with Buna N, Viton and EPR have shown that "O" ring gaskets made from these materials perform well for long periods. The tabulation of test results shown is the accumulation of years of testing. As new materials are marketed they are tested in the laboratory and plant tests scheduled if they show merit. One such material is Dupont's Kalrez, a perfluoroelastomer (See Attachment IV). Kalrez is currently under laboratory test in vinyl chloride with completion scheduled March 1, 1977. This could conceivably be superior to the presently used Buna N, Viton and EPR in which case it would be used in vinyl chloride service. (Hie resistance to ethylene dichloride and vinyl acetate by Kalrez is excellent. ) The ultimate test for any "O" ring material is its performance in plant manu facturing equipment in vinyl chloride service. "O" ring gaskets in vinyl chloride service at Union Carbide Corporation plants are as follows: 1. Eighteen dispersion resin polymerizers at South Charleston, West Virginia equipped with double mechanical seals on the agitators using Buna N "O" rings for shaft packing have been in service since 1952 without any seal failures due to "O" ring problems. New "O" ring gaskets are installed each time the seals are overhauled which occurs between six months and a year. (Most mechanical seals contain one or more "O" rings for shaft packing. See Attachment V. ) ucc 052383 Letter to Ms. Susan Wyatt January 7, 1977 Page Three 2. Twenty-six suspension resin polymerizers at Texas City, Texas equipped with double mechanical seals on the agitators using Buna N "O" rings for shaft packing have been in service since I960 without any seal failures due to "O" ring problems. New "O" ring gaskets are installed during overhauls at one-to two-year intervals. 3. Forty-five double mechanical seals have been installed since 1974 on solvent resin polymerizers at South Charleston, West Virginia and Texas City, Texas using EPR "O" rings for shaft packing rings. No failures due to "O" ring problems have been experienced. This installation is too new to have data on seal overhaul schedules. 4. Fifteen pumps in vinyl chloride service at the Texas City, Texas plant hav double mechanical seals with EPR "O" ring for shaft packing. These were in stalled in 1975; thus, no statement of "O" ring life can be made except that the rings are over one year old in liquid VCM service. Liquid vinyl chloride service is much more demanding than gaseous vinyl chloride service. 5. Forty-four quick opening Lenape manways equipped with EPR "O" ring gaskets are in service on suspension and dispersion resin reactors at South Charleston, West Virginia and Texas City, Texas since 1965. These gaskets are replaced after one-to two^xobks due to mechanical damage to the gasket from opening and closing the manway. There have been no gasket failures, however. 6. Union Carbide Corporation uses a fleet of eighty tank cars for vinyl chloride transport. Each car is equipped with "O" ring seat relief valves such as Midland Manufacturing Corporation type A-1150 equipped with Viton "O" rings on the seat and the stem. These valves are approved by the Association of American Rail roads for tank cars in vinyl chloride service. The safety valves are overhauled every five years and new "O" rings installed. Detectable leakage through these valves is unusual. In the year 1976 only one or two of these safety valves leaked and had to be replaced. These safety valves have only a single resilient gasket seat while the proposed safety valves have a lapped metal-to-metal seat and a resilient "O" ring gasket seat; thus, resistance to leakage is much greater in the proposed safety valves. (See Attached Midland Manufacturing, Inc. tank car safety valve drawing. Attachment VI) Union Carbide Corporation is using "O" ring gaskets in many other vinyl chloride process equipment items including ball and plug valve stem packing. It is felt, however, that the preceding examples show that "O" ring gaskets are reliable in vinyl chloride service particularly with our program of annual overhaul of relief valves. ucc 052384 Letter to Ms. Susan Wyatt January 7, 1977 Page Four The second question you asked was what failure frequency rate would be ex pected for the "O" rings in the proposed safety valves. The preceding examples of "O" rings installed in vinyl chloride service indi cate long life. Based on the performance of "O" rings in the tank car relief valves and the fact that all plant relief valves are overhauled annually and new gaskets installed, the failure frequency rate for "O" rings in the proposed safety valves would be nil. The third question you asked was do we know of other companies who have used "O" ring seat pressure seal safety relief valves in VCM service. The safety valve vendor, Farris Teledyne, states that a major producer of vinyl chloride/polyvinyl chloride is purchasing the valves for vinyl chloride service. Despite telephone inquiries to many companies, we have not been able to confirm this statement by the vendor. In our request for equivalency we pointed out that the Chlorine Institute, Inc. accepted "O" ring seat pressure seal safety valves equipped with Viton "O" rings as equivalent to a rupture disc in series with a conventional safety valve. In terms of resistance to corrosion and to chemical attack as well as freedom from leakage and good valve reliability, this application is much more demand ing than the proposed use of the "O" ring seat pressure sealed safety valves in vinyl chloride service. As noted earlier, Union Carbide Corporation would not propose a device which we were not confident would perform as expected. Union Carbide Corporation has fifty "O" ring seat pressure sealed safety valves equipped with low temper ature Buna N "O" rings in ethylene service at its plants in Texas City, Texas; Victoria, Texas; Taft, Louisiana and Ponce, Puerto Rico. These have been performing as expected since 1969. More of this type of safety valve are being installed in our plants where freedom from leakage and positive reseating after relieving are critical. I hope this letter answers your specific questions on "O" ring gaskets and shows the sound technical base for Union Carbide Corporation's request that "O" ring seat pressure seal safety relief valves be approved as equivalent to 61. 65 (b) (4) Leakage from Relief Valves. Very truly yours, R. in. vvneeier, jr RNWJr. /pm Attachments ucc 052385 Farris Safely-Relief Valves All AV^ruvj.JJlN 1 [ ' Farris "O" Ring Seat Pressure Seal | for Conventional or BalanSeal Eliminates leakage and costly product loss as well as costly down time and maintenance on troublesome applications such as: Operation too close to set pressure. Light, hard to hold lluitis. Minute foreign particles. Vibratory applications, Coirosive fluids. Nozzle icing conditions. Discharge piping strains i if!'! FLAT HO MACH SCREW oisc-^, ' j'/ftt, l DISC HOLDER----- "Cf'RING SEAT SEAL ____l-.r -v- -w-7 ' ^ blow down ring----_ 1 v{ '~t ! :NOZZLE- 'I ! I *4*r> NIB i . : i Li O THRU K ORIFICE Recognizing the ncerl for a resilient se.il in n safetyrelief valve for extreme tightness, I .mi'. f npineenng Corporation first made avail.ilile an "O" Itinj; '*,.il tl in early I960. Although limited in piev.pie, Ihr "0" Ring design received phenomi'nnl m ' ept.mco since it made possible complete tightness at pressures much closer to the valve set pressure than was ever possible with the standard metal to metal seats. flie present "0" Ring Sent Pressure Seal design is an improvement which allows the use of the "0" Ring Seat Seal to higher pressures and. equally important, the spring load is solely carried by the metal to metal por tion of the seat with the "0" Ring becoming a pressure seal within its recessed chamber and the outer edge of the nozzle, assuring the ultimate in tightness. The "0" Ring Seat Pressure Seal is available in our 2600 Series line of flanged safety-relief valves - con ventional or BalanSeal construction - for pressures up to a maximum of I'iOO f'MO fb'li'riuig In llip M'l'Tlimi tallies on pages 2? thru 49, s'lb'dilntc .i "P" f<-r llm loiulli digit "A" m ilie type number fur Hie feoveuliiiMel valve and a "D" for the fourth digit "D" m llm type number fur the BalanSeal construction when an "0" Ring Seat Seal is required: 26FA10 becomes 26FC10 (Conventional) 26FD JO becomes 26FD10 (BalanSeal) The pressure limit of the Conventional or BalanSeal valves covered in the selection tnhles is the same for the O Ring design in all type numbers and orifices with the 150 lb., 300 lb. and 600 lb. inlet flanges. On the "0" Ring design in all type numbers and orifices 1500 lbs. is the limit for the "0" Ring design, not the conventional valve limit shown in the selection tables A charts. Refer to "0" Ring Material Selection Chart on page 21 for Temperature and Pressure ratings of the various "0" Ring materials available. ucc 052386 Steel / Mangeu Why an *0" Ring Seat Pressure T>enl? In the normal operation of a safety-relief valve the disc must lift off the nozzle very slightly to "simmer," allowing pressure buildup within the secondary police (huddling chamber), causing the valve to "pop" fully open, Simmering occurs many times rri the process industries whore, as a result of process Changes, minor upsets, etc., operating pmssure fluctuates higher than normal, causing safety-relief valves In "simmer," blit not fully open. Dus will cause serious misalignment in the valve and after the pressure drops the valve will very niton Continue to leak below the normal operating pressure. While this can be overcome by actually popping the valve, this is not always possible. The use of the Tarns "0" Ring Seat Pressure Seal will always overcome this difficulty. As occurs very frequently operating pressures are too close to valve set pressures. As the operating pressure approaches the set pressure, seat loading is diminished reducing the force which affects tightness. With the use of the Farris "0" Ring Seat Pressure Seal, tightness can be obtained at relatively higher operating pressures than with metal to metal or other soft seat safety-relief valves. On light, hard to hold fluids such as hydrogen, helium, light hydrocarbon, anhydrous ammonia, etc., metal to metal seats are often penetrated, causing leakage problems. The Farris "0" Ring Seat Pressure Seal will overcome leakage on these hard to hold fluids. On applications where heavy vibrations occur - i.e, barges, tankers, pumps, compressors, etc., leakage of metal to metal seats develop as a result for as the set pressure is approached the spring force is minimized and the vibration tends to reduce the effect of spring loading, causing leakage. The Farris "0" Ring Seat Pressure Seal will maintain tightness as the spring force is not a factor m the tightness of the "0" Ring design. Where occasional minute foreign particles are carried in the flowing medium, metal to metal seals are usually marred or scratched when the valve is blowing. This oeates tmuhlesome leakage problems after the valve closes. The Tarns "0" ring seat pressure seal will absorb without damage the impact of these particles and will eliminate particle deformation of the mating metal seating surface on the nozzle as the valve closes, thus reducing the incidence of leakage in valves on most process units. When necessary, simply replace the "0" ring to maintain tightness. Due to corrosion, metal to metal seats may eventually leek. With the proper selection of the Farris "0" Ring Seat Pressure Seal tightness can more satisfactorily be maintained. Nozzle icing conditions result from the refrigerant effect of the flowing media when a valve relieves. Ice actually forms on the seat, thus causing leakage. The Farris "0" Ring Seat Pressure Seal again overcomes this troublesome type of leakage. O" Ring Material Selection Chart,"; *:vit'A'.) ; . FIGURES IN PARENTHESES INDICATE .jkt 7`. . SHORE A DUROMETER HARONESS AT 70* FIV; y,y IV,' -; . ' . 450 . * >* 400 350 t ollw 300 1af 250 3 < 200 CC. W 150 ' 2',; Ul i100 1 I-- ' w H Ui ' -zJ ; 0 . I -so -too ii,. 150 7: ~ 11; 7 ' i' ; i! VITON (70) . SILICONE (70) i ' t ' i;; i :; i 1:1 iTiTlTl m! I I ' ' in t 4- i! i i viton (90) 11 11:;1 111 l:;!( ;siliccin (eni j ) j 111 m hii i (ro> !.i`: M i|: ! 7 . i;! : ` i! i 11 illi \\ VITON 170) a . SILICONE (70) -BUNA N (70) (VITON (no) .MLIC'INC (no) -BUNA N (UOI M-j H .i: l' . ... JT-.viioii (70) i. ;:i; ;n III! : i; i 11 i: ;viroN (eoi niuiA it (r*u) ill: i;ii M(l v) SlLICOtIE (70) j j I *3 --BUMAN (70) L- I l., I , z... -.--.- uo-----.-'1 j 3;!; 11111; 1 i I :;; 1 i 1I _J;1 Li1'1 I ' VITON (70) SILICONE (70) 'ij'j: ,:. i . -.I- i Ii ; ii : ii'i: i CONE (70) ]i [IIIiil M ilii iil i!! i: Ml! tjthrrT11!1 ii:. j [silicone (eo) ilii i!!i !m! f 7 vV'ji* 100 200 300 400.500 *00 700 *00 *00 1000i Ilo5o0o0'k.i,-'.,.VIJV.vMV .!>, . v I ^7 SET PRESSURE r press-) ,;a;IKV-v : f. f./lf WKMEVER wrr^rv^V, Jt - * j Applltt toNtbMoHthMc*oonfv"0nt"lofhintr vntpvabttinioowbauvbabllvoaar t A93t %ept f pnreetspsurvretstur6*0. i j* y'J PSIG ohO botow, tost Abl) bo moot ot 9 P8IG botow tot proaturo - a Wltf 1 4kl3^mkk*i aor^ieo rommnootlonii tor O* Rlnp tnstorlaia BUNA N Sir Ammtnla 6utn BtityIrM C.nnott hlaiHJf dimh oh rrofi r*htr mutMa Ore*) rrffltts ait 6 12 r.iri mi f amllnf Mrlltint MyrtFttfft Sulpbl## Sfttrrnt Hthr Oil Njlurjl Gat Niiteffd 9rcn Prftp inf irS Suf'fi'tit 'nrort Scrtyirtit Air Amyl Mcaftol Argchlar #1241 Sfftrinf SutJtJitfi* luUnv iuWflt Butyltnf CarSan Dtiulphlp* C*rit* trtracbiortis Chlffttn {Orf fat) Cyclahstano Ofotstno "A1' VITON (con't.) Ctftrl CMtrltfa CthyitMc llhytmt Glfcfl fthyl Alcohol Gnifllnt llltiM HyKroeAlftrlt Ac16 INtfriUrti SalfAlSt Isbliulyl SlcaHSl II* 4 '*! ip 5 ritpi Ktffti'ftf tutor 0(1$ MPtc snlin HritiyiPfit Dkhlarltt NRliital Ga$ Nitric Acl NUrnypn Oranltr 1300 A ISIS P?O0*n PropylRttf Propr' AlcrPfl Stnrfft-rp SfNvot Sulphur Oifiltff Sulphuric Adi tatuenc lilChlfrpfttopIpM Turptnllftf Vinefif Water krl SILICON# Air MfllPhi Nltrtr*1* OiKM Thftt recammfltfitfon$ trp (uidt only ani icrvict lit* of tho "0" I*, o# eot/ryt, deptnirn! on itn* prraluro, conechtrauont, cati'ytti that mr N titfri ni oth*r conilKoftf which art brponi tor control. ucc 052387 r ` Hesiw And Analysis f%THaos-9,2 Page 67 of 74 January, 1974_________ ATTACHMENT III ELASTWCR AM) GASKET HATE RIALS FOR CHEMICAL SERVICE MATERIALS chemicals/plastics engineering The date In this section were obtained to provide personnel with useable information for the rapid selection of a gasket material for a proposed service. Selection of the test chemicals used was bated on representative homoglogs from all types of organic chemicals. By analogy, the data should be applicable for the proper selection of a gasket material for nearly any chemical mixture. It is Important to understand that the presence of a very small quantity of an aggressive chemical may be sufficient to attack the elastomer through selective absorption of that chemical from the solution. Consequently, the testing of mixtures is always advisable unless the entire composition Is stable and innocous. The materials chosen for testing represent typical comatrclal products available in that category of material (e.g. butyl). In no manner is it ixplied that the material listed for that category is either the beat or the poeroM.ef materials available in that group. Initial data were obtained by exposure of the gasket materials for 105 days At 7#*Mntbe chemical. Those materials showing good resistance to a chemical In the initial exposure at 78*F were evaluated in the same environment at a higher temperature. The criteria used for the recoaendat1on were as follows: x unsuitable Volume change of X percent or more Hardnets change (Duro. A) of 15 points or more Weight change of 50 Mg/Cm2 or more r F Fair Voliane changes of 10 to X per cent Hardness changes (Dura. . " IS points Height changes of 10 to R Recomnended Essentially no changes in the physical properties The rating of "F" (fair resistance) in the Table indicates that a significant amount of attack occurred on the elastomer during the test. These data should be interpreted to indicate the elastomer would probably provide satisfactory service when confessed between two flange faces in a piping system, but could not be used as an exposed diaphragm In a value. The "R" rating would Indicate that the elastomer could be used in any service while *X" shows Inadequate resistance of the elastomer under any circiaastancet. hole: Teflon gaskets are resistant to all chemicals currently processed in the Division. Texqwratures up to 320*C are allowable with the possible exception of amine exposures above 175*C. When using Teflon gaskets, do not use solid sheet material. The following are suitable gaskets where Teflon is required. r n Johns-Manville 91 or 92 Garlocfc B406 R/M Fluorabestos Flexon 500 Taskline (Durlron Co.) Teflon sponge Teflon felt Teflon envelopes ucc 052389 r 13.2.2.1 "" l' `""" * CHEMICAL RESISTANCE OF GASKET MATERIALS Test Chemical: Vinyl chloride Exposure: 111 days at 33F '"r n u* 8 o 00 ^ Gasket Material Viton A Neoprene 7797 Butyl K->53 Thiokol 3060 Black rubber 3773 Hardness (c) Start End (a) 48 hrs. (b) 86 75 82 5f 66 49 71 58 68 55 83 * 63 75 72 Volume Change, % End (a) 48 hrs. (b) +18.8 + 0.6 +35.7 +29.3 +40.3 nil - 0.3 - 1.0 Wt. Change, %(b) - 3.7 - 0.8 - 1.6 - 6.4 Color Solution (d) Yellow -- Yellow Yellow Yellow Comments on Specimen Slightly swollen SvMllen Swollen & tacky Swol 1 en Rating F P X X X Hycar, D-24 63 47 Buna N (OR-25) 83 71 Hypalon 72 SI Silicon No. 65 53 46 Buna S EPX 46 40 S3 yy Natural rubber, gum 47 35 Red rubber No. 107 81 63 Adiprene L-100 45* 37* Saraloy 300 65 Garlock asbestos 7735 65* 68* U. S. asbestos No. 899 71* 67* 78 +37.5 87 +25.5 -- *" *SS1 51 +39.6 56 +33.2 -- 46 +56.4 89 +35.9 43* +30.0 68* + 1.0 68* + 1.2 -17.6 - 9.9 - 3.8 -26.5 ---10.8 - 8.2 + 0.3 + 0.3 + 0.3 - 0.8 - 9.1 - 4.2 * ' -23.3 r -12.1 - 4.9 - 0.1 - 1.3 - 1.9 Yellow Yellow Yellow Yellow -- Yellow Yellow Yellow Yellow Yellow Swollen Slightly swollen S/sfMtf Sam#** Swollen Swollen Swollen Swollen Swollen No visible change No visible change X F P X X F X X X R R ..... ' (a) Information obtained on the specimens immediately after removal from the solution. (b) Information obtained on the specimens after drying 48 hours at 78F. ,. (c) Hardness expressed as Durometer "A" except where * appears. Asterisk indicates Durometer MD" hardness. (d) Color of solution ot start of test was 5 Pt-Co. .. < Resistance Rating: R - Resistance good; recommended material. F - Resistance fair; use of the material depends on specific service. X - Unsuitable. .This is true laboratory test data by Union Carbide Corporation - l-k-77 -ff J. C."Can ter burr"y / Given under my band this 4th day of January i977 COmm. ^ t VJ k Lufi y 1977. My conmussdon expires July 7, 1979, NOTARY PUBLIC ucc 052390 (JINJION CARBIDE INTERNAL CORRESPONDENCE VC" ATTACHMENT IV CHEMICALS AND PLASTICS P. O. BOX 8361, SOUTH CHARLESTON, WEST VIRGINIA 25303 To (Nwno) Civilian location Canterbury Bldg. 770-346 Copy to MECCO o.t May 25, 1973 originating Dopt. Engineering folftC*- St^tKt Chemical Resistance of DuPont's ECD-006- /CAL Perfluoroelastomer - ..... Black Vulcanizate i - In a continuing program to evaluate new elastomeric materials for possible use in chemical service, ECD-006 perfluoroelastomer, a synthetic rubber manufactured by E. I. DuPont de Nemours and Company has been tested. The material has far greater resistance to organic chemicals than any other fluoroelastomer evaluated. The ECD-006 is produced in both gum and black vulcanizates , the primary difference being the addition of carbon black to the black vulcanizate to increase the mechanical strength. Only the black vulcanizate was evaluated in these tests. Results of the laboratory tests (Table I) indicate an exceptionally broad range of chemical resistance for ECD-006, although failure occurred in organic acid, aldehyde, latex and certain amine test media. Compared with "Viton A", another widely used fluoroelastomer, ECD-006 offers a much greater potential for use in organic chemical process streams. In addition to the superiority of ECD-006 over "Viton A" in chemical resistance, DuPont reports greatly improved continuous heat resist ance for ECD-006. The ECD-006 remains usefully elastic after 200 days at 260C (500F) compared to +40 days at 260C for "Viton A" according to their data. The physical properties of the two elastomers are reported by DuPont to be similar. The recommendation of ECD-006 over "Viton A" and possibly other elastomers would be obvious except that the price of ECD-006 is approximately 20 times that of "Viton A" at the present time. Hope fully, the price of the new ECD-006 will be reduced with time. Th re may be special situations however, where the high cost of ECD-006 can be justified at the present time. &*+*'* T.tott , MaAU /W 77* -34*4 HTP/ds Attachment: 1 Table Not book Refer nc : 29JCC-62 3 #,jr 12043 Index: H. T. Pritt UCC 052391 G, 23, 50, 119, 155, 157, 162, 166, 168, 181, 184, 187, 190, 194, 195, 197, 199, 207, 208, 218, 221, 220, 22.1. 229,242 "4 5 .26^ TABLE I CHEMICAL RESISTANCE OF ECD-006 PERFLUOROELASTOMER Test Conditions: Specimens (3.0" x 0.5") of the elastomer were immersed in the following chemicals for 100 days at 45C. i Chemical Hardness Acetic Acid, Glacial Acetic Acid, 10% Acetic Anhydride 75 63 79 Acetone 75 Acetonitrile Ammonium Hydroxide 75 75 Aviation Gas 73 Benzene 75 Butanol Butyl Acetate 74 75 Butyl Carbitol Butyl Cellosolve Butyraldehyde Carbon Tetrachloride 75 75 71 69 Ethanol, SD-2B Ethyl Acetoacetate Ethylene Chlorohydrin Ethylenediamine Ethylene Dichloride Ethylene Glycol 2-Ethyl Hexanoic Acid FLEXOL Plasticizer DOP Formaldehyde 72 73 66 64 72 73 74 75 64 Mgs./CM2 Gross Thick. Change Change. Inches +40.95 +56.74 + 4.45 +0.012 +0.016 +0.002 + 3.01 +0.003 + 1.42 + 6.06 None +0.003 + 3.63 +0.001 + 3.88 Hone - 0.73 + 5.05 None +0.003 + 0.81 + 1.56 +39.1 +11.2 +0.001 +0.001 +0.019 +0.003 + 1.6 + 2.6 + 9.0 +20. 4 + 4.1 + 0.16 + 0.9 + 0.2 +28.0 None None +0.005 +0.009 +0.001 None None None +0.007 Comments ^Moderately swollen; no wrinkles Badly swollen; blistered; wrinkled Very slightly swollen; no other visible change. Very slightly swollen; no other visible change. No visible change. Very slightly swollen and bleached; no other visible change. Very slightly swollen; no other visible change. Very slightly swollen; no other visible change. No visible change. Very slightly swollen; no other visible change. No visible change. No visible change. Moderately swollen; blistered Slightly swollen; no other visible change. No visible change. No visible change. Slightly swollen and softened. Swollen and blistered. No visible change. No visible change. No visible change. No visible change. Moderately swollen and softened. ucc 052392 TABLE I (continued) Chemical Mgs./CM2 Gross Thick. Hardness^ Change Change. Inches Formic Acid, 90% Glyoxal, 40% Hexanol Hydrochloric Acid, 10% 42 75 75 73 +131.9 + 2.9 + 1.5 + 7.3 +0.033 -0.010 None +0.002 Isophorone Kerosene Methanol Methyl Ethyl Ketone Monoethanolamine Nitric Acid, 10% Phenol Sodium Hydroxide , 10% Sulfuric Acid, 98% 74 73 74 71 71 69 76 77 75 + 1.8 + 1.4 + 1.6 + 3.1 + 2.2 + 12.0 + 1.8 + 3.7 + 5.6 None None None +0.001 +0.001 +0.003 None None None UCAR Latex WC-130 Vinyl Acetate Water at 90C 59 + 28.4 +0.010 74 + 3.0 +0.001 73 + 11.7 +0.004 Comments Badly swollen aid softened. No visible change. No visible change. Very slightly swollen; no other visible change. No visible change. No visible change. No visible change. No visible change. No visible change. Slightly swollen and softened. No visible change. No visible change. Very slightly swollen; some absorption. Moderately swollen; softened. No visible change. Very slightly swollen; no other visible change. (1) Original hardness - 75 Duroraeter (Shore) A * No good. Based on previous standards of admissible changes in the properties of an lastomer when tested in this manner, the material should not be used in the chemical exposure listed. ATTACHMENT V Pump /V'-r , Fluid prfssur s-, U: f \ Staling f::es 1 Shoulder against housing '"J --( Uv--- rr- `-5` -i ! r nry -' 8 i Rings tcikv wear here Static scot i rI "T \ V| t *------- 1`"__jj Stone seat * ' l-.J 3 Gaskets supply 'give' Starting from scraSeh, Set's design a fypises! p> Simplest roitM of mechanical seal has a shoulder on shaft pressing against machine housing. Fig. 1. This will work if housing and shoulder are both finished properly and shaft is loaded against housing at all times so fluid inside easing won't leak through.-Hut this isn't practical on most machines. For one thing, there is always some shaft end-play and run-out (g)ration). There are other icasons, also. Next step is to press a sealing ring into the housing and press another ring on the shaft. Fig, 2. Now sealing surfaces of rings are easier to machine and lap. Also, rings can he renewed when they wear without machining sealing area of shaft or housing. But keeping shaft against the scaling area is a problem. So is aha ft run-out or gyration because rings arc rigid and run-out would cause leakage. Fig. 3 shows next step: mounting of resilient gaskets in hack of rings to give some movement to rings. This also seals fluid leakage around hack of rings. But trouble is this sim ple design is very seldom practical because of the very limited shaft end-motion these static seals (gaskets) can absorb. Besides, you won't' be able to obtain and maintain the correct contact fmee to keep sealing faces together. Solution is to give either one of the scaling faces axial motion to absorb cud-play and shaft run-out, t(ius holding the rings together at all times. Fig. 4 shows one way to do this. Here we place a diaphragm in the housing. This flexible member gives the sealing ting more axial movement than the gasket we used in Fig. 3. Diaphragm should keep a steady pressure against the two rings so there's no leak age, even if shaft gyrates a little and it has some end-play. So the diaphragm serves three purposes: (1) acts as inner MANUAL ON MECHANICAL SEALS POWER MARCH Is.. 4 Stationary sect 5 Rotating seal $?fifing ring !----------------- ^ flouting.-- Mating ring Spring----------^ --r~WiwL > Sftevt nut - -] ii:. Shaft ^ ^.ou"tr-- ^* r;*.'.... - -Gland . S ^ Coskat 6 Here is comp) to seat +>/*/a modern mezhcnkal seal for our plant seal between seal ring and housing (2) serves as spring member to create contact lorce, and (3) prevents rotation between seal ring and housing. Types of seals. There arc two basic types of mechanical seals--stationary and rotating. In a stationary seal the sealing ring is in the machine housing and does not move. In a rotating seal, the sealing ling turns with the shaft. Seal shell. On a rotating seal. Fig. 5, the seal shell is press fitted onto the shaft. Du a stationary seal, the shell is press fitted into the housing. The seal shell is a member to which the diaphragm is fa.-tened and which was not used in the previous four diagrams. Fig. 6 shows all the functional elements of a mechanical seal. Here the mating ring is wedged into the gland, with the O-ring preventing leakage around it, and giving it some resilience. Shaft sleeve is held against a shoulder on shaft by a nut. Sleeve also acts as backing ring. If shaft is not threaded as this otic, then a setscrew through sleeve is used to hold sleeve on the shaft. Sealing ring is held against mating ring with coil-type spring. Flat gasket between hous ing and gland prevents leakage. tMjmjs seal easing fluid along shaft, past sealing ring and unlock of mating ring. For bast results, fluid being sealed should be on the out* side of mechanical seal so it must travel inward to seep between seal faces. This is important--particles in fluid tend to travel towards a region of higher velocity. This effect is distinct from centrifugal lorce, which also exists. The effect is most pronounced with viscous liquids; it is less with gases. Solids have less tendency to work between seal faces with fluid on the outside of revolving ring. Stationary vs rotatinij medmnkal seals Now you might ask why we have two types of seals. Here are some answers to help you decide on how best to apply each type. Advantages. Most machines have a rotating shaft with stationary seals, sketch, left. Main advantage of this seal is centrifugal forces don't work on the flexible seal ring because it's stationary. Mating ring is the only rotating member, but it's comparatively small and is usually rigidly attached to shaft. This is done hy either a clamping action or over balanced hydraulic forces which hold it firmly in place. These seals arc used on shafts that often travel at 15,000 feet per minute (not rpm) or higher. Shaft diameter may be more than 3 in,- This means that the centrifugal forces developed can be extremely high. On other hand, rotating mechanical seal is subject to centrifugal forces, sketch left. This keeps flexible seal-ring from wurking freely against mating ring. So this design is ..............."u "** WANMAl ON MECHANICAl SEALS seldom used for high speeds. Main advantage is easy instal lation on straight shafts found in tno.-t machines. Stationary seal needs a back-up shoulder for positioning the mating ring. Shoulder'* relative truencss to shaft will directly effect the face run-out of mating ring. In rotating seal, mating ring is retained in housing. This housing is usually more easy to machine, therefore it holds the locating shoulder in a truer plane. Since rotating seal ring is the flexible member, it's posiiioned by the mating ring faces. The seal itself can easily be locked hy a set screw, friction fit or similar means on any smooth shaft. Another advonlugo for rotating seal is its use in chemical solutions where solids may precipitate and tend to foul the sliding-seal components. Because of centrifugal force, the rotating seal has a greater tendency to free itself from sludges and sediment. But stationary seals may have other design details that often better overcome this advantage. ucc 052395 ~) 1 1 DOUBLE MECHANICAL SEAL ASSEMBLY TOLERANCE MOT SHOWN .0*3 own* Sf*l PARTS 1 SC AMO KINS t INSIST (A THROAT SUSHI MS I- N FT FOR SEALING FLUID INLET 3i- l* N PI'S 113* APART FOR VENT G ORA:n 3 SEAL MINS SA SEAL RING 4 COMP NINO 5 COLLAR S V SINS A V RING t SLEEVE COLLAR r? 10 VftING 12 SET SCREW IS SET SCREW 14 ECCENTRIC WASHER IL i 15 SLEEVE C SPRINGS 0 GASKET R SHAFT PK4. F L C S SET SCREW THROTTLE RUSHING NOTE P WE SUGGEST FARTS J.2A,3,5*,S.tA "^-------I11r16 _J the Dimensions shown anc the NECOMHSmED minimum CP,S G G IE KEPT IN STOCK AS SPANES ALTENNATC GLAND ARRANGEMENT FOR VENT G DRAIN WITH THROTTLE RUSHING 100- 1 oc{ A B1 ] J C z do;,-coil MIX D E 1F 1G H i 1| iooslioo; J K L1 N ]| |i |1 | [ 11 _4-- |l TSO 12501' 3*5 2 V4 BIT l S'Sli *1* 473 1 3'! Ii 310 937 1 *37 |. 3421 tOCC 1 30011 421' * I0ST 1342- 64*i ' M 2 S l2Sli'Sc( J 11ST i 44'* 11 4 2' IZSO 1 75C)l 6*5' 131* 1 412 II 93*! UTS T*tT 2 OOCl 1437 1937 2C62I * 1SOO 200C 2'2 5 * 1341 2CS2j2-4*i 1 423 212 31223C1 > ia? 2 H' iZi'i : IT3Q 2 290123?3l * 2.23 2000 2 ** 2 97i2:82l * 2 2 JO . 231 2 312 2 Bt( - 2 372 22301 2*3* 2 5121 2 uozjril 2 2 562 2*37| 2 425 250ol * 2 4 S'* 2 5421 2*3025231 2 412'2CS?! 2 47312 *3ol 2 33*12 12 3 QGC* 2 9 7 3 J * * - * * 3 042 2 43 71 3 2313 :CO| 3 3 174 t V.2 1 347S sc< > S 60*11 701 *-* IT1I' S 1," * 734(. 42 G - . ( 1 797J1 690 '\s l591i5V V, [lhi 1 *22 12OH*8 9 9B4I 2 079 204712.40* 2 2101. 2 203 I > ft 2 I72|224S:2 / 2 iSJli i V. 2297 2353 i 2422 ' 2* 244* 2 3 74* 2 - 2 3*7 640 2 V* I Gl 2 2112 !lT) 473 2373'23001 1937 243*123421 2000 290012425: 204} 2 9621244*1 " 3 147131621 3 2301 3 23 I * 3 3*9i 3 2S0 1 * 1*T*IS1 2l * 2409 2672 2 7631 2t/9 2T34 ! ja-i 2 797 2)Hl 5 2439 2933< 3 -4 2 123 262S-2T3C1 2-47 2 46*! 24'il 2230 ATS? 267-. *3'I 24 F2 29371 *373 2473 3COC+ 2437 2937 3 0 42: 2300 3000 31231 2342 3042 31471 5 9CO: 5 3 * 9 [ 3342 5*37 3623'55GO 344715542 373013*23 3 4l21347l 3 67513 730I 393713 9'2 | * - 2922 3 0-3' * < 23.- rr-.r--r*3047 3 1*0' 3 r* 3109 1203 J Vft 3172 : * 3234 332B1 3r/W 3297 39901 SVj 3359 3*33) >l-1t 2423 1129 3 230: 2447 3.87 33-21 2730 3253 3 372J * . 4 147 1*0621 4-97 1*142 *3i2l*97 * * 3469 37031 3V* 3331 3703! 3 r* 3394 3l i j 3312 3 437- 312 |4 B7 > 1636 34291 3 *4 2973 3373I330C- * *4571* 3 2 3719 33 } 7t 2G3T 343* 13 9621 * 3000 3300 342*1 44371* J.I I * *342 **37 3791 3993} 3^1 3944 4074| 4 3042 3342 3S9*. 3123 342"1 37301 3147 3687118; 2 1 . *3421**37 44971*542 *447.* 5421 - 3230 373015 B 73. 3 h* * 4-21* 5671 2 -Hfi 3^5 3312 39-21395* 33T3 347314 000' * *9 2 144671 . 4937149 2 | * 3903 40741 */4 3949 4203i4Vg 4031 I03|* W 4093 4324' * <* 4 134 4324i^ 4219 44931* H 3437 195*1*042: 3300 4ooci4<3: 3342 4042" 4 4* 3423 412-14250 1447 4 147 *3.2 3730 4 23C.4373 1412 43-21**17' 3973 43731*3001 3917 443 T 4 362 4000 4 30C *62*1 * . * * * 1 493*1 * a 2 1 * 30421 <3171 S 062'4 937T 3250:5 231 9230 3 231 * 3 3.2 3 971 . 5 3>2 in- r 34171*3 2 I * 3 S6215 * 371 3 342 3*37l --fl-- 1 * 4291 4 433. * 7"* 4343 *3741 4 n 4404 437414 4469 444Q 14931 44401* 04 ' U394 TOJi * V* . 14694 4 705* 4 . 14 719 14791 14044 4932, 3 | 1 I ----------1--------- ______ 1______ ______ 1______ I DOftAMETALLTC CORPORATION | KALAMAZOO* MICH. DOUGLE INSIDE RO , FT 0 CARTRIDGE lOiTt |. 14 . 44 ** MW_____ 1 Min # w DAR*4 PA 1 nuM |CNC4 " jL -51.327 ucc 052396 MIDLAND VALVES AND GAGES FITTINGS MOUNTINGS ACCESSORIES MIDLAND MANUFACTURING CORPORATION 7733 GROSS POINT ROAD - SKOKIE, ILLINOIS 60076 PHONE NUMBER: (312) 677-0333 ucc 052397 ( 4. SEE NOTE t AND 2 ( ATTACHMENT VI ucc 052398 s' SECTION "A-A" SIZE OF FLANGE HOLES: 7.'8" is standard on all valves, except 3/4" on A-103; and A-IC39. If a special size of hole; i.e., 7/8" on A-1039 or 3/4" on A-1075 through A-IS^S is desired, usesuff Any other Flange.Hole, use suffix J. ' ... ___ BOLT CIRCLE: i, 6i" is standard on all valves. If a special Bolt Circle cf 6;" is desired, use suffix B. Any other Bolt Circle use suffix C. MOUNTING: The standard mounting on all valves, except A-1035 and A-1039, is the AAR Standard tongue and groove with an 00 of 4.745"i .003 and an ID of 4.0G5".0C3. Or tne A-1035 and A-10 the standard mounting is a flat flange (no tongue and groove). If a special nounting 4.99b" x 4.255" is desired, usesuffix M. If a flat flange is desired for any valve lotner than A-1035 a A-1039) use suffix F. For any other mounting use suffix P. EXAMPLE OF ORDERING FCR SPECIAL REQUIREMENTS: A-1150-HcW - is A-II50 valve with 4-5/4" with Holes on a 6j" Bolt Circle, with a 4.995" x 4.255" tongue and groove mouncir:. APPROXIMATE WEIGHT OF VALVE - 25#, SAFETY VALVES A-1035 through A-1450 ITEM NO. NO. REQ'D. DESCRIPTION 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 <6 17 18 19 20 22 1 l l 1 1 1 1 1 1 t 1 1 i 4 4 2 1 * 4** 4* * 1** __ l 1 TOP GUIDE STEM RETAINER BODY SPRING FOLLOWER SPRING GUIDE' TOP LOCK NUT WASHER SEAT "0" RltyG STEM "0* RING BOTT.ADJ.NUt BOTT.LOCK NUT STUD STUD NUT SEAL 3" STD. PIPE 3/4" STUD 3/4". NyT GASKET MOUNTING 0 RING NAMEPLATE MATERIAL FOR PART HUMBERS FOR SS TRIM VALVES A-1035 Thru A-1375 STL W/SS INSERT STAINLESS STEEL. CAD. PLT. STEEL. CAD. PLT. STEEL, CAD. PLT. MALL.IRON.CAD.PLT. STEEL. CAD. PLT. STEEL. CAD. PLT. LAMIN.NEOPRENE* BUNA N-700* BUNA N-70D* STEEL. CAD. PLT. STEEL. CAD. PLT. ALLOY. CAD. PLT. STEEL. CAD. PLT. LEAD W/SS WIRE STEEL ALLOY STEEL STEEL LEAD BUNA N.70D* STA1NLESS 10-1-xs 10-2-SS 10-3 -CS 10-4-CS SEE TABLE BELOW 10-6-MI 10-7-CS 10-8.CS 10 - 9 -LN 10 10 - BN 10-11-BN 10-12-CS I0-I3-CS 10-14-AS I0-I5-CS 10-16-PB I0-I7-CS 10 -18-AS 10-19-CS 10-20-PB 10-21-BN SEE TABLE BELOW MATERIAL FOR STAINLESS VALVES STAINLESS STAINLESS STAINLESS STAINLESS STAINLESS STAINLESS STAINLESS STAINLESS LAMIN.NEOPRENE* BUNA N-70D* BUNA N-700* STAINLESS STAINLESS STAINLESS STAINLESS LEAD W/SS WIRE STAINLESS STAINLESS STAINLESS LEAD BUNA N-70D* STAINLESS PART HUMBERS FOR A-1039 Thru A-1379 10-l-SS 10-2-SS 10-3-SS 10-4-SS SEE TABLE BELOW 10-6-SS 10-7-SS 10-8-SS 10-9-LN I0-I0-BN 10-II-BN 10-12-SS 10-13-SS 10-14-SS I0-I5-SS 10-16-PB 10-17-SS 10-18-SS 10 -19 -SS 10-20-PB 10-21-BN SEE TABLE BELOW * Alternate material available. ** Not furnished unless requested. PLEASE NOTE: PLATING ON PARTS CAN BE CHANGED OR ELIMINATED ON WHICHEVER PARTS CUSTOMER SPECIFIES. THE PLATING ON THE ABOVE PARTS IS TYPE RECOMMENDED FOR GENERAL SERVICE. VALVE HQ. A-1035 A-1075 A-1150 A- 1225 A-1247 A-1300 A-1375 A-1450 STAINLESS TRIM VALVES PRESSURE CFM SPRINS SETT INS (Air) NO. NAMEPLATE NO. 35 661 10-35-AS 12-35 SS 75 1 109 10 - 75 - AS 12 - 75-SS 150 2004 10-150-AS 12 - 150 -SS 225 2884 10-225-AS 1 2 - 225-SS 247 3070 10-225-AS 1 2 - 247 -SS 300 3516 10-300-AS 12 - 300 -SS 37 5 4220 10-375-AS 12 - 375 -SS 450 4705 10-450-AS 12 - 450-SS VALVE NO. A.1039 A-1079 A-l154 A.1229 A-1304 A-1379 STAINLESS VALVES PRESSURE CFM 'SETTING (*r) SPRING NO. 35 661 tO-39-SS 75 1 113 10-79-SS 1 50 1940 10-154-SS 225 2670 10-229-SS 300 3695 10-304-SS 375 4615 10-379-SS NAMEPLATE NO. 12-39-SS 12-79-SS 12-154-SS 12-229-SS 12 -304-SS 12 - 379-SS t--4 t'.'J t. V L- ' w- I (.1 REVISED 1-4-67 ucc 052399 ))