Document M4k8EzwKZm6kjqm57O1k3ZNO7

iANSACTIONS 1-9 10 or' Copies More SET OF 28 VOLUMES: $23io $1.50 $1.19 1.50 1.19 and & Forging 1.50 1.19 es .88 .80 svelopment 1.50 1.19 fouth, Farm 1.50 1.19 1.50 1.19r 1.98 1.74, .88 .80- ndustries; 1.50 1.19 s 1.50 1.19?' tions 1.98 nir 1.50 1.19 > 1.50 1.19* .88 .80s 1.50 1.19 1.50 1.19. tals .88 .80 1.50 1.19' 1.50 1.19k * 1.50 U90 .88 ^80 1.98 1.74. 1.50 1.19. a 1.50 1.19.3 i .88 .80 !, 1.50 1.19 1.50 1.19 \gency Trade Discount; 1-30 days. Government ' orders of $5.00 net or iLUME 19 PR GAS 64th NATIONAL SAFETY CONGRESS Gas detectic Papers Delivered in the from rather n ^modern electr Petroleum Sessions Principles and Limitations of Gas Detection * |of detecting [[combustible a m.m The basic u Intent today if Recent Research Activity on Fire and Explosions by Factory Mutual Research...................................... Myron J. Miller [area for comb [Yet, modern one of the le, Underwater Safety--........................................................ C. H. Klunick 1! Offshore Drilling and Its Safety Problems ..................... Bill Goolsby [ under-estimate Gas detectio chosen, opei Hazards of Hydrogen Sulfide in Refining.......................... Wes Wallace ll Industrial Hygiene Considerations in Refinery Operations.............................................................. Mt 'iton M. Garcia 23ss,- carefully in lin ['limitations inf i equipment mu i times. Service Station Safety--What Can You Do? ;J Think About It!.................................................................. C. W. Haven 2tiU Principles l ^detection inclu BUSTION; T Pipeline Explosion--Back in the "Not-So" Good Old Days................. ............... .........................Wayne Kinison 30' i High Level Tank Alarms on Pipeline Tanks................... R. L Johnson 34ffr| TIVITY; ELI GAS CHR DETECTOR Aflame ion: Explosion Suppression of Large Turbulent Areas......... W. A. Crosley 37 | PHYSICS. Eac I specific applica Officers of the Petroleum Section 1976-77 ............................................. 4lHp i tion situations. Other Volumes in 1976 Congress Transactions..................... Back Covedfcl Before a cho [|made, the profc uf ; the expectatic jr usually lead to if tors used in de n delude: AREA r* [TYPE OF C a* - ?? ' SPECIFICIT' I.BIENT CONI TIME; POWI and PERSON!4 ...Common lim -detection progr; 2' J Petroleum Sessions ETY CONGRES I in the sions ...................Warren J. Riley sions ...................Myron J. Miller 'J ......................C. H. Klunick ....................... Bill Goolsby ....................... Wes Wallace ^ ry .......... Mt 'iton M. Garcia ......................C. W. Haven ^ j| ..................Wayne Kinison 3G ................. R. L. Johnson ^reas......... W. A. Crosley 3| ........................................................... 41 ns..................... Back Covq| PRINCIPLES AND LIMITATIONS OF GAS DETECTION INSTRUMENTATION By WARREN J. RILEY Warren Industries, Cupertino, CA Gas detection techniques have evolved rom rather primitive methods to i5dern electronic instruments capable f detecting minute concentrations of ombustible and toxic gases. The basic use of gas detection equipent today is to check or monitor an area for combustible or toxic conditions, et, modern gas detection remains as iue of the least understood and most inder-estimated safety tools available, as detection equipment must be faosen, operated, and maintained refully in line with its importance. The ipjitations inherent in the use of such uipment must be kept in mind at all es. Principles used in modern-day gas ection include: CATALYTIC COM- STION; THERMAL CONDUC TIVITY; ELECTRO CHEMISTRY S CHROMATOGRAPHY TECTOR TUBES; OPTICS ;AME IONIZATION; and CHEMYSICS. Each of these principles finds cific application in one or more deteci|n situations. ^Before a choice of detection method is |de, the problem must be defined and g expectations set forth. This will ilitally lead to an obvious choice. Fac ets used in defining an application inlude: AREA TO BE MONITORED; IYPE OF OUTPUT NEEDED; SPECIFICITY OF SENSOR; AMl$pNT CONDITIONS; RESPONSE IfiVlE; POWER REQUIREMENTS; PERSONNEL. ommon limitations of the typical gas Section program include: too little time allocated to the program; little knowledge of basic gas detection definitions (LEL, TLV, etc.) on the part of operating personnel; lack of con sideration for factors effecting gas detec tion (type of gas, sources of leaks, temperature, ventilation, gas movement, etc.; misunderstanding of accuracy vs repeatability; insufficient calibration and check-out procedures; insufficient record keeping; improper choice of operating personnel; operator belief that, through some magic, the detector can "see" a hazard; and insufficient training of engineering, operating, and maintenance personnel. While this is not a complete list, it does include some of the more common problems found in many operating gas detection programs. Any gas detection program requires a minimum allocation of time for each of its four major phases: planning; training; equipment selection; and maintenance. Example of time allocation: Function Planning Equipment Selection Training Maintenance Time Spent Typical* Optimum* 10% 20% 75% 30% Your _ _ 5% 20% _ 10% 30% _ Fill in the time that "Your" organization spends or has planned to spend on these four phases. Compare "Your" times with the "Optimum" values. You may find that some reconsideration is in order. Potential limitations exist, whether you are using basic detection methods or the most sophisticated electronic systems. A lack of understanding about these limitations can lead to incorrect readings. For example, can the equip ment: Operate under ambient con ditions? Give needed stability and ac curacy? Be easily calibrated? With few exceptions, the information gained from a gas detector must be inter preted and acted upon by an individual. Each person responsible for installing, using, or maintaining gas detection equipment must be knowledgeable of the limitations connected with their equip ment and take proper precautions to avert trouble. Many of the personnel problems can be solved through training of the right people at the right time. Training should include: objectives of the program; prin ciples of operation; program limitations; operation of the specific hardware; maintenance requirements; and record^ keeping. Training begins (with thfg engineering personnel) before the equip^l ment is purchased and continue^ through commissioning. RoutinJ retraining should then be conducted. Modern gas detection equipment can .i and will provide an important source of information on the existence of poten tially hazardous situations. A clear understanding of the prin ciples and limitations of the equipments necessary to realize the full value of any gas detection program. This understand ing could also prevent a false sense^pf; security leading to disaster. ^ In the Final analysis, the interpretation of readings and decisions for corrective action must be left in the hands of.jjje responsible individual. Values are based on the author's evaluation of percentages stated in conversations with numerous clients. Petroleum Sessions itenance requirements; and recorijt&sd >ing. Training begins (with the^l leering personnel) before the equirl;$ RECENT RESEARCH ACTIVITY ON FIRE AND EXPLOSIONS t is purchased and continue! >ugh commissioning. Routing BY FACTORY MUTUAL RESEARCH ining should then be conducted. odern gas detection equipment cap.-J will provide an important source of. -1 By MYRON J. MILLER Factory Mutual Research Corp., Norwood, MA mation on the existence of potep. 1 f hazardous situations. if clear understanding of the pri]L?,4 In 20 years in the loss prevention s and limitations of the equip mefafs.vi msiness, I have been troubled, not only isary to realize the full value ofa|y J' [by the questionable effectiveness of letection program. This understaqjj- .'Jf tany protective systems, but even more ;ould also prevent a false sens^f^ io by the inadequacies in our ability to ity leading to disaster. j/* [measure that effectiveness. The question the final analysis, the interpretatjdppi How are we doing?" may elicit a adings and decisions for corrective-** [response that is satisfactory to the un- n must be left in the hands of. the jderwriter but hardly satisfying to the rnsible individual. [research engineers and scientists seeking Preprints of these revised standards are now available from the NFPA,3 and public comment must be received before December. I encourage your review. Flame Extinguishment With Gaseous Agents Figure 1 shows a comparison of recommended design concentrations now in the 1973 Standard with those to be voted on at the spring 1977 meeting of . arc based on the author's evaluation of percentiles [a goal for systems improvement. NFPA. As you can see, there are some n conversations with numerous clients. yjr Consequently, in 1974, I presented a significant changes. These changes have paper on. the "Reliability of Fire Protec resulted from an evaluation of many tion Systems"1 at the AIChE Loss large- and small-scale tests. Some com ^Prevention Seminar. The paper parisons of results at different scales are [presented some comparative statistics on shown in the following two figures. ipparent system performance but, more Figure 2 compares dynamic ex (importantly, dwelt on the needs for periments. Figure 3 compares total letter information. Since then, a number flooding experiments. Jof'efforts have been started. Some of These comparisons resulted in the these have been cited in my Karlsruhe development of a laboratory system with iaper.2 New studies are presently in which extinguishing concentration data ireparation by Factory Mutual and could be obtained both economically ithers for release early next year. Inter- and with a high degree of consistency. lational cooperation in this area was Provisions for heating both the burner tlso discussed in Karlsruhe. and the fuel made it possible to compen [t This interest in systems reliability, sate for differences in heat generation [maintainability, and performance and heat transfer at large- and small- capability (the components of the term scale. The apparatus and techniques "effectiveness") led naturally to have not yet been completely standard lliestions regarding the assumptions ized, but a comparison of the results pised in the design of specific systems, from three laboratories shows lince I have been involved in the reasonable agreement, as shown in levelopment of halon systems and serve Figure 4. Recommended design values [bn NFPA 12A and B, this area became have been obtained by adding a 20 per the first target of opportunity. Some of cent safety factor to the average of values Sthe changes that are being proposed in from a minimum of three laboratories. [NFPA 12A/B involve pipe flow To date, the Factory Mutual Research ^Calculations and the design concen Corporation (FMRC), Du Pont, ICI, trations for flame extinguishment and and the Ansul Company have par Inerting. I will only dwell on the latter ticipated in this effort. Others are en ItWo here. couraged to participate. Acetone Benzene Ethanol Ethylene Methane Heptane Propane Figure 1 Comparison of Old and New Design Values for Flame Extinguishment 1211 Old New 4.2 5.0 3.9 5.0 5.0 5.0 6.5 8.6 4.3 5.0 4.3 5.0 4.8 5.8 1301 Old New -- 5.0 ~~ -- 5.0 4.0 5.0 7.2 8.2 2.0 5.0 3.7 5.0 3.2 5.2 Methane Heptane Hydrogen Burner Size Enclosure Size 6 3.3 214 in. ID 314 in. ID .1.2 x 12 in. 180 ft3 Figure 2 1 1.5 .47.7 5.4 mm ID 50 mm ID Dynamic Experiments 2 34 4.5 1.3 4.0 4.5 3.8 27.0 18.0 38.5 mm ID 28 mm ID 5 in. ID (Liq) VA in. ID (Gas) 105 mm ID 105 mm ID 2 x 2 ft m _______5_ __ rJ 1.7 18.0 8 mm ID \fl ;c 105 mm ID !c -------------------- U Figure 3 Total Flooding Tests 78 9 3.5 2.5 3.4 214 in. ID 314 in. ID 12 x 12 in. 214 in. ID 314 in. ID 38 x 38 in. 893 ft3 1000 ft3 --9242 ft3 10 (1211) 3.4 5 ft.3 Circular 2500 ft3 m 11 a A 0.9 j! ''I 38 x 38 in. Of .id 9856 ft* J) *1? A fivsidered t since the to the fu five per < variety c The d tor a po complies jjfuel, sprs rave not jut in all boncentra Flame fof halons lependen; 'he des jresume " sion withi Jess) disch |"rapid" de Inature of grange fron Several sec pill genera Jot only iresulting fr ftities of )osition/r jrobability lay deer temperature balance con Work is 5 Idardization tuch of the nd New xtinguishment 1301 5 Old New - _ 5.0 . A -- 5.0 if 4.0 5.0 ii\~ 7.2 8.2 1 2.0 5.0 3.7 5.0 -m 3.2 5.2 f Fuel g Methane | Propane L Ethylene I Hydrogen I Acetone 1 Benzene [ Ethanol L Heptane 1973 NFPA 1.8 2.9 6.5 18.0 -- -- 3.3 3.4 Figure 4 Halon 1301 1976 AB 3.0 3.0 4.3 4.2 6.4 6.7 -- 22.2+ 2.9 3.6 3.2 3.5 3.3 4.3 3.7 4.5 C 4.0 5.5 8.1 27.5 3.3 3.3 4.2 4.5 Petroleum Sessions Halon 1211 1973 NFPA A 1976 B C 3.9 3.4 3.2 4.5 4.4 4.7 4.8 5.8 5.9 6.8 7.0 8.5 20.4 25.9+ 27.3+ 34.0 -- 3.2 3.8 3.9 -- 2.9 2.9 3.6 4.5 3.9 4.5 4.8 3.9 4.0 4.4 4.6 ; 5 5 .0 -n ID m ID ests 38 in. 2 ft* 4 1.3 3.8 18.0 5 in. ID (Liq) VA in. ID (Gas) 2 x 2 ft 10 (1211) IT 5 ft.2 Circular 2500 ft* ______ i.7. - 8 mm ID-i^l /frfrr 105 mm IDjjjfl 11 0.9 38 x 38 in- H# 9856 ft*i >a /ite A five per cent concentration is con sidered the minimum that should be used since there is always some uncertainty as |o the fuel, that will be present in a fire. A 3ve per cent concentration covers a wide Variety of common fuels. The design values recommended are jra pool fire configuration only. The iplications introduced by running tel', sprays, vapor jets, or hot surfaces jaye not yet been completely defined, it in all probability will require higher pncentrations. lame extinguishment concentrations lialons for all fuels are temperature |pendent, many to a significant degree, ie design values recommended isume "rapid" detection and suppresoji within the required 10-second (or jpj discharge period. The definition of lipid" detection will depend upon the pre of the fire anticipated and may ge from milliseconds to as much as eral seconds. Longer detection times 5 generally be an invitation to trouble. only will the increased fire size ilting from delay increase the quanies of combustion/decom|ition/reaction products, but the bability of successful suppression gy decrease due to increased nperatures and generation of energy pee conditions favoring the fire, fork is still progressing on the staniization of laboratory systems, with :h of the attention to be focused upon the effects of temperature. Many of the changes in threshold concentrations, seemingly dependent upon flow con ditions, can probably be ascribed- to flow-induced changes in the temperature of the air/agent mixture and resultant changes in heat transfer and stability at the flame/burner interface. Flow and mixing phenomena and flame stabilization effects are important, however, and must also be considered in both experiment design and translation of the results to real fire situations. Some of these questions have been addressed in a report just published by S. N. Bajpai and M. A. Delichatsios of FMRC.4 Copies of this report should be available shortly on request. A related experimental system was developed by Tewarson of FMRC to study the burning of plastics.5 Inerting Gaseous Mixtures All that we have said about flame ex tinguishment can also be said about inert ing, except that we have not as yet had the advantage of an interlaboratory comparison of data. A report by John Coll of Fenwal Inc.6 has just been issued and apparatus is being assembled by FMRC and ICI to conduct similar ex periments. Round robin experiments with several other systems have preceded the choice of this apparatus. This ap paratus, with a one psi pressure criterion, appears to indicate higher con- 7 1976 National Safety Congress centration requirements for inerting than The multiple pressure peaks often pres--m previously reported using the U.S. ent in vented explosions also need no Bureau of Mines type of tubular explo better theoretical treatment. ; sion burette. The papers of Reynst11 describing the-,rt Not only should existing inerting development of pulsating combustion >g values for halons be viewed with suspi show that under very special conditions- ;^ cion, but values for traditional agents a self-sustaining, self amplifying series of:, a such as CO2 and nitrogen should also be explosions can be generated in a vented.Vr,; questioned. In addition, modest changes chamber containing fuel. in temperature and ignition energy can Explosion suspression still remains aianr result in significantly higher agent con subject with very little publicly availabledid centration requirements. A careful in information. The recent Coast Guardrto-i vestigation of this dependency is needed. tests 9 demonstrated the possibilities for i The future of halon systems depends using water and dry chemicals as well as -. 5? largely on the record of reliability and halons as suppressants. Optimized com- effectiveness yet to be established. There binations may also be appropriate. .. is, as yet, no basis for complacency, and In March 1976 I attended a meeting mnsiM a careful analysis of all design assump Stockholm of the International Stan-i.fa\ tions and data is in order. dards Organization (ISO) Committee onfbq Explosion Venting and Suppression Extinguishing Agents (TC21/SCG). oH * There was no eagerness on the part ofnetj members of this committee to address ; Just a few brief comments on explo explosion suppressants, particularly sion venting and suppression. The Fac since there are already many questions it tory Mutual System has just issued Loss associated with the testing of fire ex- Prevention Data Sheets 7-76 dealing tinguishants. There is as yet no standard- C with combustible dusts.7 Much informa ized laboratory system for evaluating! ; tion on explosions-and venting has been explosion suppression agents. This subr-. -; considered in the preparation of this ject is being discussed by the ISO Com-' ). guide. Copies are available and I would mittee on Explosion Suppression ',,-fe appreciate any comments you might Systems (TC21/SC5/ WG3). have on the utility of this document. In September I attended a meeting in As Dr. Zalosh of FMRC pointed out Karlsruhe of this committee at which the1 M last year in a review of explosion venting need for expanded research was dis literature,8 there may be order-of- cussed. As a result of work done by Dr.. magnitude differences in pressures W. Bartknecht of CIBA-Geigy12 a . predicted by various computational program using a one-cubic-meter test methods. Much of the difference is due chamber is being proposed. Several in to the assumptions used. dependent laboratories hope to par For instance, in a recent series of ex ticipate so that results can be compared, periments conducted by the Coast Guard first on explosion behavior alone and in the pump room of the tanker Rhode then with standardized explosion sup Island,9 pressures increased from three pression systems. Factors to be in psig to twelve psig when obstructions vestigated include type and nature of the were added to generate turbulence. The fuel, degree of turbulence, and type of rate of-pressure increase also changes ignition. Water, halons, and ammonium significantly. phosphate will ultimately be compared Reproduction of the pressure curves using several discharge configurations. could be obtained by changing the tur Hopefully, this work will also investigate bulence factor in the FMRC computer the effects of temperature and be coor model.10 However, predicting these fac dinated with our studies of inerting using tors in advance still remains a problem. gaseous agents. The question of financial Support for |>pen. In conclus Bew results Eddition tc gapers, sumr Antonio in 1 all. I have ai fibliography gent referer Bbliography arted in var | If any of y Research put Crite me anc vailable. Th action of F^ bibliogr; lblication |owever, I c pmments on |L. Miller, M. J. Systems." Ch (April 1974.) Miller, M. J.; f Suppression." national Fire . Germany, 22 Report No. R Ir-National Fire Bl TCR-77-A An iNFPA Pub. N jf.Bajpai, S. N.; 'by Halons--Pi -'No. 22545 for I *July 1976, ir 1>1- 8 Petroleum Sessions multiple pressure peaks often pre$+J i vented explosions also need 3 theoretical treatment. n papers of Reynst 11 describing theif pment of pulsating combustiomfi, hat under very special conditions^ ustaining, self amplifying series ofr! ions can be generated in a ventedifi er containing fuel. ;$ osion suspression still remains' with very little publicly available ation. The recent Coast Guard demonstrated the possibilities fod /ater and dry chemicals as well as as suppressants. Optimized comff ns may also be appropriate, arch 1976 I attended a meeting ift| olm of the International Stan4 )rganization (ISO) Committee on .uishing Agents (TC21/SCG)<4 was no eagerness on the part of rs of this committee to addressl on suppressants, particularly^ iere are already many questions^ ted with the testing of fire ex-' lants. There is as yet no standard-); boratory system for evaluating^ on suppression agents. This subn >eing discussed by the ISO Comj)| on Explosion Suppressior s (TC21/SC5/WG3). .`ptember I attended a meeting ir he of this committee at which thell 3r expanded research was dis-1^ As a result of work done by Dr.sjjj rtknecht of CIBA-Geigy12 a> n using a one-cubic-meter test'd r is being proposed. Several in- ;nt laboratories hope to par- so that results can be compared, - explosion behavior alone and ith standardized explosion sup- n systems. Factors to be in-i|j ed include type and nature of the' gree of turbulence, and type of i . Water, halons, and ammonium^ ite will ultimately be compared > averal discharge configurations.^ lly, this work will also investigate^ cts of temperature and be coor-;< with our studies of inerting using agents. The question of financial ipport for this program still remains ||en. |In conclusion, we have tried to publish results as work progressed, and in |dition to individual reports and Jpers, summaries were presented at San pitonio in 197513 and at Karlsruhe2 this jjl. I have also compiled a computerized |bliography of halon literature14 to augent references in these papers. The Sbliography includes over 600 citations jrted in various ways. Jf.any of you are interested in FMRC esearch publications mentioned, please rite' me and I will try to make copies gailable. This represents only a small action of FMRC's total research effort. [ bibliography of other FMRC lblications is also available.13 lowever, I do solicit your review and 1>mments on any material you receive. REFERENCES jjMiller, M. J.; "Reliability of Fire Protection s Systems." Chem. Eng. Prog. Vol. 70, No. 4 '(April 1974.) ^Miller, M. J.; "The Use of Halons for Flame 'Suppression." Proceedings of the 5th Inter national Fire Protection Seminar, Karlsruhe, ^Germany, 22-24 September 1976, FMRC Tfcport No. RC76-TP-41 (May 1976). National Fire Protection Association (NFPA) FCR-77-A Annual Meeting Technical Reports )6F-PA Pub. No. 13M-I0-76 SM (Oct. 1976). iBajpai, S. N.; "Extinction of Diffusion Flames Shy Halons--Progress Report 2." FMRC Serial vo. 22545 for E. 1. du Pont de Nemours & Co., JfJtily 1976, includes Appendix by M. A. Delichatsios entitled, "Extinction of Diffusion Flames by Halons in Co-flowing Streams." 5. Tewarson, A., and F. Tamanini; "Research and Development for a Laboratory-Scale Flam mability Test Method for Cellular Plastics." FMRC Serial No. 22524. 6. Coll, John P.; "Inerting Characteristics of Halon 1301 and 1211 Using Various Com bustibles." Report No. PSR-661 of Fenwal, In corporated, Ashland, Mass., August 16, 1976. 7. Factory Mutual System; "Combustible Dusts." Loss Prevention Data Sheets 7-76 (Sept. 1976). 8. Zalosh, R.; "Review of Recent (Post-1969) Ex plosion Venting Literature." FMRC Serial No. 22535, June 19, 1975. 9. Richards, Robert C,, and Daniel Sheehan, (U.S. Coast Guard); "Explosion Suppression Systems for Marine Applications." Paper No. OTC2561 at the Offshore Technology Conference, Houston, Texas, May 3-6, 1976. 10. Yao, C., J. deRis, S. N. Bajpai, and J. L. Buckley; "Evaluation of Protection from Explo sion Overpressure in AEC Gloveboxes." FMRC Serial No. 16215.1, December 1969 for the U.S. Atomic Energy Commission (Contract AT(11D-1393). 11. Thring, M. W. (ed.); "Pulsating Combustion-- The Collected Works of F. H. Reynst." Pergamon Press, 1961. 12. Bartknecht, W.; "Explosionen von Gemischen aus Brennbaren Gasen, Dampfen Oder Stauben und Luft-Teil 1: Explosionsablauf." Report TS 3, Zentraler Sicherheitsdienst, Ciba-Geigy AG, Basel Switzerland, May 1975. 13. Gann, R. G. (ed.); "Halogenated Fire Sup pressants." American Chemical Society Sym posium, San Antonio, Texas, April 23-24, 1975. 14. Miller, M. J., and J. M. Kenney; "Bibliography of Halon Literature." FMRC Report RC76-T74, September, 1976. 15. Factory Mutual Research Activities; Sponsored and In-HouseJtesearch Summaries, 1976, Fac tory Mutual Research Corporation, Norwood, Mass. m 4?. 9 IWT 1976 National Safety Congress UNDERWATER SAFETY By C. H. KLUNICK Dir. of Safety, Continental Oil Co., Houston, TX u For the past ten years, subsea American National Standards Institute operations technology has been in a (ANSI) for review and adoption. The dynamic state of development with an ADC "Manual of Safe Practices in Coup almost continuous gap existing in "what mercial Diving Operations" provides tiig has seemed possible to do" versus "what basis for a universal code of diving saftftijf was proven safe to do." Commercial div with jurisdiction over diving operatic^ ing and submersible service companies in U.S. territorial waters which will alsoj serving offshore oil exploration and most probably have substantial impact production operations around the world on the spirit, if not the letter, of diviftg| have been pressed on by subsea work codes and regulations berlifj' requirements in deeper and deeper promulgated around the world'%^ water--with a minimum of time to ade various countries with offshore oil frfcVjj quately and safely prepare for the ever- terests. iSfef- increasing demands. Review of the ADC Manual of Some efforts have been made to Practices and ABS Guidelines reveal$ig4 provide uniform safety practices for the tendency towards "broad-minimUiBKIr subsea work industry. In 1968, the rather than "specific-tough" rules American Bureau of Shipping prepared operate by, which should not be too ^ and issued a "Guide for the Classifica prising from an inter-industry consensfi|| tion of Manned Submersibles" as a to accommodate the needs of the broa supplement to the "ABS Rules for range of service companies involved: * Building and Classing Steel Vessels." The availability of comprehensh The ABS Manned Submersible Guide rules and regulations does not insili provides a general basis for standard safety. Given the time, effort, anj izing the design, construction, test, cer money, we can identify and regulator tification, and, to a limited degree, of the risks inherent in the offshore Si operations of small submarines. subsea operational environment, but t! The U.S. diving industry has also most important ingredients in accidel recognized the desirability of having for prevention are trained, qualified, caref mal, industry-wide regulations to assure people on the job. safer working practices. However, the job of pulling together a "universally acceptable" set of equipment design, per Review of Critical Subsea Work Safety Areas sonnel qualification, and procedural As an introduction and rationale standards has been frustrated with suggested amendments to the AI numerous delays, largely because of the Manual and ABS Guidelines, our st differences in technical and business re ing point is to review and discuss thd quirements which exist across the diving areas where subsea work safety is mo industry and the competitiveness which critical. In instances where ffi has tended to restrain the free exchange recommended safe practices of information, including safety related guidelines seem to have a need data. .strengthening or better definition, After nearly seven years of effort, the ditions or amendments have bb| Association of U.S. Diving Contractors proposed in a subsequent section It (ADC) is about to submit a set of com overall incorporation into the Cone prehensive diving standards to the Subsea Safety Manual. ^ 10 The mo safety appe 1. Pressi T mannei iressure a integrity ar iperations quipment ssting dut ion, fre taintenanc procedures, [ualificatio !/ 2. Main \osphere. I taintenanc tbsea wor if the com lerwater a ised vesse te careful tpmperati _ tinimizatio iuch as CG jpospheric q like of env |&it.h adeq pressure ve; procedures, jjj. Decc clthough se ifastic pres: Passion pre &d result i trying fron |prt to perm \re.ssion loroughly t ..rould be dt the desii tl time u trsonnel sh jb'the prop ijbles and sc field dive rc fr'decompn j?le modi tnges shot Commi tjpjftoteness i&ir assist lnications -ft- Petroleum Sessions SAFETY NICK il Co., Houston, TX 'It erican National Standards InstifQjj ISI) for review and adoption. C "Manual of Safe Practices in G cial Diving Operations" provide; s for a universal code of diving s; 1 jurisdiction over diving operati J.S. territorial waters which will .t probably have substantial im :he spirit, if not the letter, of di^ ies and regulations beii mulgated around the world ous countries with offshore oilsts. eview of the ADC Manual of :tices and ABS Guidelines reveal lency towards "broad-minimlfi er than "specific-tough" rules rate by, which should not be too$ ing from an inter-industry consei' ccommodate the needs of the broi >e of service companies involved^ he availability of comprehensf s and regulations does not inSi ty. Given the time, effort, a ley, we can identify and regulat le risks inherent in the offshore 4] .ea operational environment, but ti t important ingredients in accid ention are trained, qualified, care; lie on the job. iew of Critical Subsea k Safety Areas s an introduction and rationale jested amendments to the AE| lual and ABS Guidelines, our st point is to review and discuss thd s where subsea work safety is nig ical. In instances where. )mmended safe practices'"WT lelines seem to have a need lgthening or better definition, a )ns or amendments have be iosed in a subsequent section rfi all incorporation into the Cone ;ea Safety Manual. The most critical areas to subsea and visual communication is often [safety appear to be as follows: desirable. This requirement affects 1. Pressure Control. In every aspect equipment standards, testing, and $f manned subsea work, control of maintenance operations. It also demands ressure and maintenance of pressure functional helium voice unscrambling Integrity are essential. Assurance of safe equipment, routine and emergency com Operations involves strict adherence to munication apparatus, and good view juipment design standards, adequate port visibility for submarines, diving sting during the following construc bells, and deck chambers. tion, frequent inspection and 5. Personnel Physical Condition. To maintenance, proper execution of diving function safely in the water, the subsea rocedures, and adequate personnel worker must be in excellent physical jalification and training. condition; well rested, and free of undue f.2. Maintenance of Breathing At emotional stress. Thorough, regular mosphere. In addition to proper pressure physical examinations are mandatory to Maintenance, it is essential to provide the have assurance that no physical condi Ibsea worker a breathing atmosphere tion may exist which would be |f the correct quality whether he is un- aggravated by pressure exposure, perwater at ambient pressure or in a precipitate decompression sickness, or [losed vessel. This requirement dictates embolism. Good pre-dive planning and tie careful selection of gases, control of scheduling, coupled with adequate div temperature and humidity, and ing crew size, is required to assure minimization of atmospheric impurities reasonable crew rest. pch as CG2 and odor. Appropriate at 6. Personnel Comfort. A subsea mospheric quality is assured through the worker who becomes physiologically of environmental control systems overstressed and is consequently uncom jfith adequate backup, insulated fortable tends to be distracted and less ssure vessels, and good operational efficient on the job. Depending upon the procedures. type of stress and its severity (work rate, 13'. Decompression Procedures. breathing effort, respiratory and body Although seldom fatal, except in cases of surface heat loss), personnel safety may astic pressure loss, improper decom- vary from impairment of individual ession procedures can be debilitating safety consciousness to dangerous, un iigjU result, in physiological impairment controllable physiological reactions. Miying from minor short-term discom- Adequate personnel comfort is primarily jrt to permanent disability. All decom- assured by proper selection, use, and re,ssion procedures should be maintenance of equipment and un Joroughly tested prior to field use and derwater tools, employment of proper Mould be designed to permit execution procedures, and personnel training. fc the desired task while minimizing 7-. Medical Treatment. Regardless of ||al time under pressure. Responsible all safety precautions which are taken, it Personnel should be thoroughly trained must be anticipated that the need will m>;the proper selection of appropriate exist for treatment of decompression ibles and schedules and their execution. problems and traumatic injury. All per |jeld dive reports should be monitored sonnel engaged in or supporting ambient Mr decompression-related incidents, and pressure diving operations must be tble modifications or procedural properly trained to recognize the signs jnges should be made accordingly. and symptoms of decompression $4i'. Communications. Because'~of the sickness and embolism. Adequate treat pjnoteness of support personnel and ment procedures, equipment, and ie.ir assistance, clear audio com- therapeutic gases must be available and anications in all situations are required personnel must be trained in their use. 11 1976 National Safety Congress Aboard rigs or vessels without extensive chores are as frequent or perhaps mofe [Guidelines to assure medical facilities and corpsman per frequent than those happening to divefs ithe elimination or sonnel, the diving team must have its and others under subsea work con (potential hazards. A' own adequate supply of first aid equip ditions. General deck safety procedures fj revisions, a survey 0 ment and be knowledgeable in handling must be rigorously followed by subsea [summary reports of traumatic injuries. work team personnel. Specializes ind submarine accic 8. Minimization of Exposure Time. procedures for securing, handling, and four years has been r Thoughtful minimization of the job time maintaining equipment unique to tile been placed on underwater reduces risks. This goal is diving operation must also be developed l.ations to prevent a best achieved by thorough pre-dive plan and implemented. tifiable sources of th ning, coordination of subsea functions 12. Emergency Procedures Prepara- The Conoco "Subs for maximum single dive effectiveness, tion. The harmful effects of accidental |s to be employed in c and adequate post-dive recovery situations can often be reduced^ ipplicable national, s capabilities. eliminated by properly preparing and ftandards, and regula 9. Handling Safety. Launch and practicing emergency procedures to-be lesign, construction retrieval of submarines and diving bells followed in the event of an incident. -An tnd operation of div and the handling of other equipment and effective accident prevention progrSfti Submersible systems, supplies in the diving area must be ex includes a thorough review of all po&f- (reduction system' ecuted smoothly and safely to avoid the ble mechanical and human failures,' Fnited States and its potential for major accidents. Lifting analysis of their causes and effedfe; [his would include E equipment must be built with adequate elimination of the potential hazard'dSty (revisions ofthecurre safety factors and be maintained in good design, if possible, and adequate lSME pressure vessel ' working condition. Personnel must be emergency training and practieife,- iuard marine engine *' well trained in handling procedures and Emergency procedures in the form'ASf' md the American Be j associated-safety considerations. Alter- specific plans, equipment, and trainifi^' l pGuide for the Classifi ! nate lifting equipment should also be are established to provide immediajf; Submersibles (1968).' , available and procedures for its use es support in those instances where <i >reign codes and regi tablished in advance of need. Sea state is acceptable but finite risk remains a pi cecedence over the Cc often the limiting criteria for execution of a given operation. ri [ices if their require ! of a dive due to associated handling 13. Documentation. It is absolute ftringent or explicit. : dangers. The decision to execute or not essential for good written eb Although some prov i execute a dive is the responsibility of the munications to be available to suppi Sidelines are appli, subsea work supervisor. hazardous work situations. This inclu< ianual of Safe Pract 10. Fire Safety. Both support vessel proper documentation and maintena te- general basis for and diving equipment fires constitute of standard operational procedui [ruction, test, certifica significant dangers to personnel while safety practices, and other speciali: r?. . f equipment use confined in a closed environment. The information, such as decompressil (ient pressure divii potential for fire can be significantly schedules and emergency therapy, Jonnel transfer capsi reduced by proper equipment design, information supporting even ftticular, all pressure - rigorous procedures excluding flam- simplest manned subsea, operation is Ibmpression chambe mables and ignition sources from inside voluminous and too critical to be pas$ (ells (SDCs), persor pressure chambers, and the use of ap on by word of mouth or to rely'4 s (PTC's), and s propriate fire resistant clothing by per memory. -WC s) will have press sonnel. The possibility of drilling rig fires resulting from blowouts and other causes must also be considered in view of Provisions to Assure the Elimination or Control of Known Potential Haza\ iu certified according '"the ASME Unfired ide and Subchapter dangers to pressurized personnel, and The recommended additions $&st Guard Marin evacuation under pressure procedures amendments to the ADC Manual ^ ||ulations and Materia must be established and understood. ABS Guidelines are the result of a revi| toted in the ADC m 11. Deck Safety. Accidents occurring and correlation of the preceding sectioj Similarly, the ABS to personnel aboard ship involved in list of critical subsea work safety af ?.vide the general basis basic operations and ship handling with the ADC Manual and A`E instruction, test, cer 12 Petroleum Sessions res are as frequent or perhaps nV<$e |uent than those happening to I others under subsea work chit ons. General deck safety procedu^js ;t be rigorously followed by subrifc k team personnel. Specialist cedures for securing, handling, tfjjd ntaining equipment unique to:J ng operation must also be devel implemented. Emergency Procedures Prep( guidelines to assure provisions for either Ithe elimination or control of known * iptential hazards. As a primary basis for revisions, a survey of available accident jummary reports of commercial diving jnd submarine accidents over the past |pur years has been made, and emphasis jps been placed on making recommenjations to prevent a repetition of idenBfiable sources of these accidents. ^ The Conoco "Subsea Safety Manual" operation of manned, free swimming, or power-cable tethered submersibles. The submersible hull will be designed, con structed, and certified according to Sec tion 2 of the ABS Guidelines. Submer sibles equipped for ambient diver lockout operations will utilize pressure hulls for the diver lockout capsule which have been certified according to the aforementioned ASME code and U.S. Coast Guard regulations. . The harmful effects of accideh^l . to be employed in conjunction with all ations can often be reducedS||#.,.^ Applicable national, state, or local laws, inated by properly preparing Uypsfl mdards, and regulations governing the ticing emergency procedures to&jj* |esign, construction, test, certification, >wed in the event of an incident.'^' md operation of diving and man-rated ;tive accident prevention progfiftr submersible systems, including subsea ides a thorough review of all po$&. production system habitats. In the mechanical and human failures;*^ Inited States and its territorial waters, iysis of their causes and effetj^ this would include adherence to the ination of the potential hazard^ irovisions of the current revisions of the gn, if possible, and adequate: iSME pressure vessel codes, U.S. Coast rgency training and practifcif iuard marine engineering regulations, rgency procedures in the fornttljgif- jnd the American Bureau of Shipping ific plans, equipment, and traittii4'' ruide for the Classification of Manned established to provide immedii jgubmersibles (1968)." Supplementary ort in those instances where T >reign codes and regulations will have ptable but finite risk remains a pi cecedence over the Conoco safety prac given operation. tices if their requirements are more . Documentation. It is absolui ringent or explicit. ntial for good written cot' ^Although some provisions of the ABS ications to be available to sup guidelines are applicable, the ADC rdous work situations. This inclui (anual of Safe Practices will provide er documentation and maintena' ! general basis for the design, con Other than the specific demarcation in the certification requirements for pressure vessels as noted above, the sub sea equipment procedures, and per sonnel requirements for Conoco's manned subsea operations will integrate applicable portions of the referenced sec tions in the ADC Manual, applicable portions of the referenced sections in the ABS Guidelines, and the Conoco Ad ditions and Amendments. In the event of a conflict occurring between the provisions of the ADC Manual, the ABS Guidelines, and the Conoco Amendments, the most stringent or ex plicit provision will be followed. Manual of Safe Practices In Commercial Diving Operations All known sources of information have been considered in preparing this manual. These sources include practices of the U.S. and foreign navies, foreign itandard operational procedui y practices, and other speciali; mation, such as decompress lules and emergency therapy, rmation supporting even lest manned subsea operation is ninous and too critical to be paSl y word of mouth or to rely ory. traction, test, certification, and operaAioh of equipment used in support of >ient pressure diving and tethered sonnel transfer capsule operations. In irticular, all pressure vessels, including jpfcompression chambers (DDC's), divjjpjells (SDCs), personnel transfer capjjiles (PTC's), and subsea habitats pVHC's) will have pressure vessels rated id certified according to Section VIII and domestic diving regulations, recommendations from industry members, U.S. Coast Guard and In dustrial equipment codes, and scientific data from university, governmental and industrial sources. Diving technology is a rapidly ex panding science, and the variations in techniques and task complexity ex perienced throughout the industry are isions to Assure the Elimination fcthe ASME Unfired Pressure Vessel vast. It has taken several years to assess mtrol ofKnown Potential Hazai^ Shde and Subchapter F of the U.S. regional differences in practices, per e recommended additions dments to the ADC Manual Guidelines are the result of a revM orrelation of the preceding secti'd| f critical subsea work safety art Jti&st Guard Marine Engineering emulations and Material Specifications, Igphoted in the ADC manual. 'J rSimilarly, the ABS Guidelines will H&hvide the general basis for the design, sonnel requirements, and equipment considerations to determine a basis of safe practice. Consequently, the manual is considered to be a basic guide in the conduct of diving and may be com the ADC Manual and ABS phstruction, test, certification, and plemented by additional procedures and 13 i y/o nauonai dajeiy congress requirements as dictated by specific com checks, and maintenance. (He also serves pany operations or government rules as diver on a regular work roster.) and regulations. Life Support Technicians. Life sup The manual is intended to serve as a port technicians control and maintain guide to members of the diving industry life support equipment, including gas in developing those safe operating prac analysis and mixing, and control of the tices and procedures compatible with the environment of the deck chamber com prevailing circumstances inherent in plex. Technicians are rostered on 12- each individual location and type of hour shifts. These men are required dur operation. It is not intended to provide, ing deep dives requiring decompression and in fact cannot provide, inflexible of 15 hours or more and for all satura practices or procedures to be followed tion diving. blindly by every operator in every ac Backup Divers. In deep dives or in tivity with eveiy type of equipment. two-shift saturation diving, whenever Drilling Rig Diving Personnel "back-to-back" diving is required, two additional divers are required. In deep A description of each position or job diving, the first pair of divers is usually classification and duties is as follows: under decompression when the second Supervisor. The supervisor is in dive is required, thus, the need for overall charge of the diving operation, "backup" divers. decompression, and crew functions. He is generally stationed in the control area Emergency Procedures which is the central communications and The following is a discussion of control point for operation dives. When situations which could occur during div two supervisors are assigned, each man ing operations, requiring the implemen is responsible for a 12-hour shift, with tation of emergency procedures and one supervisor designated as the lead quite possibly the help of non-diving rig man. personnel. Although every emergency SDC Divers. The divers use the SDC has a.unique set of problems, familiarity to get to the jobsite, and one man locks with the "standard" emergency out to perform whatever tasks are re procedures listed below will increase the quired. The second man remains in the effectiveness of personnel helping to deal, SDC as the safety man and tender. He is with any given situation and help to in prepared to leave the SDC, if assistance sure the safety of those involved in an is needed by the lockout diver. incident. Winch Operator (Diver). The winch Rig Power Failure. The diving operator operates the SDC lift winch system's electrical component^ are wire<j through directions from the supervisor. so that in the event of a rig power supply (He also acts as a diver on a regular work failure they will automatically be run- roster.) from a standby generator. Both of these, Umbilical Tender (Diver). The power supplies must be operable before primary function of the umbilical tender a dive is undertaken. In the event of rig- is to maintain the SDC umbilical during power failure, the dive will be terminated the raising or lowering of the SDC by immediately and the bell hoisted to the use of the chain stoppers. (Additionally, surface using the emergency powfer he is required to act as diver on a regular supply. Should both supplies fail, the work roster.) bell can still be hoisted to the rig using Tender (Diver). The tender is an assis the auxiliary air winch. Battery-powered! tant to the supervisor in making pre-dive communications are available to maim? checks, tending deck equipment dutfhg tain constant communications. 'A; the dive, assisting umbilical tender, con backup battery supply is available for* trolling decompression and post-dive the CO2 scrubbers to insure a safe; 14 Petroleum Sessions cks, and maintenance. (He also ser$|s liver on a regular work roster.) q.^ .ife Support Technicians. Life sup- t technicians control and maintkjh support equipment, including lysis and mixing, and control ofitftg sronment of the deck chamber cqhl- .. Technicians are rostered on r shifts. These men are required deep dives requiring decompress^ 5 hours or more and for all sattt\$- diving. ?4{ ackup Divers. In deep dives oriip shift saturation diving, whenej^ k-to-back" diving is required, t%j. tional divers are required. In degj^ lg, the first pair of divers is usuajj^ ;r decompression when the second is required, thus, the need folr. kup" divers. rgency P,,roced,ures * j,decompression without relying on rig [ ^power. Communications Failure. The bell is ^provided with two sources of hardwire ^.communications. One set of wires is constained in the umbilical, and one set is in jjthe core of the strength wire or "Electrorporteur" cable. In the event com[jnunications are lost, the dive will be ter minated immediately, using the ^following procedures:. 1. Normally, communications are ex changed between surface and diver/bell &every five minutes. Topside and bell /tender must keep note of the time of the . last communications check. 2. When communications are not es tablished at a five-minute interval, the time of last communications should be i noted and a period of 20 minutes will be . allowed before the bell is hoisted. In the event of communications failure, the probably place for such an occurrence), it is likely that the umbilical will also part, and the bell will descend to the bot tom. The bell will still be attached to the dual guide wire system, which is capable of returning the bell to the surface, where a crane or tugger line can lift the bell back to the deck. The bell will be halted at a depth attainable by a standby diver to check the occupants' condition, if there is no communication. The on board supply of gas and scrubbing capacity will enable the bell occupants to continue normal decompression in the event they are already pressurized when ' the accident occurs. The bell is always supplied with the necessary decompres sion tables for the appropriate depths at which they are working, as well as an emergency procedures list. Main Winch Failure. As noted above, the guide wire winches will act as a he following is a discussion*^ ^tender must immediately bring the diver .tions which could occur during divjf H Kjto the bell and prepare for ascent. A operations, requiring the implemeife.; ^prearranged signal via the diver's hose n of emergency procedures an; Liyill be given by the tender for the diver : possibly the help of non-diving ng to return to the bell and prepare for as- onnel. Although every emergen i unique set of problems, familiarii! i the "standard" emergent edures listed below will increase ti] tiveness of personnel helping to de| any given situation and help to1: the safety of those involved in a| ent. if! ig Power Failure. The divit m's electrical components are witi at in the event of a rig power supp e they will automatically be. a standby generator. Both of thtl :r supplies must be operable befo e is undertaken. In the event of-'i r failure, the dive will be terminate ;diately and the bell hoisted to tlj tee using the emergency po\ .y. Should both supplies fail, :an still be hoisted to the rig usiijfj uxiliary air winch. Battery-powetfl aunications are available to maip constant communications.5 rp battery supply is available 202 scrubbers to insure a safiPI. b9f?ntjjk 3. At the end of the 20-minute inter nal, the bell will be raised by the surface distance of 10 feet from its bottom position as a signal to divers that ascent Os. imminent. After 5 minutes at this [depth, the bell will complete the ascent Sto the surface. If sea conditions permit, ghhe bell will ascend to a depth attainable lliy standby diver and stop, allowing the ^S^andby diver to ascertain the condition j$n the bell which will allow com munications between the surface and the 1 via the standby diver's comImunications line. The bell will then be fisted to the surface and mated with the |eck decompression chamber and nor|mal decompression carried out. -|2 Main Strength Wire Failure. In the g\milikely event the main strength wire ptarts, the bell is capable of being rought to the surface by means of the ell guide wire system. If thd"Strength yire breaks while the bell is between the feck and the wave zone (the most backup for lifting the bell to the surface. If both of these winches fail, the strength wire can be stopped off to a crane or air tugger line and hauled to the surface. Bell Entanglement on Bottom. As many bells will be working in water depths of 500 feet or greater and in cold and often severe sea conditions, the ability to send down a standby diver to assess the situation if the bell becomes entangled on the bottom is dangerously impractical, if not impossible. The same procedures that apply for other related emergencies will naturally apply in this situation, e.g., possible communications loss, strength wire parting, etc. The bell is equipped with cutting tools for sever ing wire rope, and if bell-surface com munications exist, oxy-arc cutting gear can be lowered from the surface. The bell will never leave the surface without a complete complement of diving gear, even on an observation dive, and is always equipped with partial pressure and decompression tables, emergency procedures list, spare water, batteries, and lithium hydroxide CO2 absorbent. Using the oxygen supplies on board as the limiting constraint, the survival time for the occupants of the trapped bell 15 -`S3 would be several days. However, ex check their supplies against those listed posure to a cold, wet environment would and order additional supplies as needed. be a more critical situation, and would likely limit the survival time to only a few hours once the bell is pressurized. At least two members of each operating team should have an adequate background in first aid treatment. A* First Aid Major and minor injuries may be ex pected in the conduct of diving operations. In many instances, qualified medical personnel and treatment facilities are not readily available. In these situations, diving contract per sonnel must be prepared to cope with major and minor injuries until such time as professional treatment may be rendered. It is the intention of this sec tion to establish the minimum re quirements for first aid- equipment and instructions to be available on the scene of any Conoco diving operation. As stated, these are the minimum re quirements. Additional supplies and equipment may be dictated by the operations supervisor as a result of dis tance to complete medical facilities, weather, or other considerations unique to a given operation. If questions arise as to the need for additional supplies or treatment procedures to meet such re quirements, they should be directed to either^ the diving company's medical director or safety director. Those installations which already have first aid kits and instructions should knowledge of first aid procedures is of value to all personnel to provide a background for coping with medical emergencies both at work and at home, All diving employees are encouraged to take advantage of free evening first aid courses offered by the Red Cross, county and city hospitals, adult education, and other programs offering first aid training and certification available throughout the world. Field operations groups should be familiar and have copies of the new U.S. Navy Diving Manual. In this manual isarevised section on first aid treatment that should be read by all diving supervisors A prevei worki darner ' he ca curren sion. The getting ' have c | is to g '' tain a ' crew. ^ The and others responsible for the physical well-being of personnel while working; offshore and in the shop. Know this material well--you will not have time to study it if an emergency arises. In addi tion to these references, all operating^ groups should have a summary first aid; * ; ' > rsituati availab lull se ' causes: ' tools; machir : mainter chart for inclusion in their first aid kits.v This chart should provide simplified in-y structions for recognizing symptoms and; giving treatment by those personnel who! do not have extensive backgrounds infirst aid. | ?from es 'training i:crew rru ' value o! 1 commui What Human Offsh< time ha ! ' handling bar, pipt wooden cessive 1 poor pla cidents. in offshc 1. Ser tied up racked ' handling over peo 2. De chocked are remo 16 itm Petroleum Sessions i their supplies against those list Drder additional supplies as need'el t least two members of eac OFFSHORE DRILLING AND ITS SAFETY PROBLEMS iting team should have an adequ'ail ground in first aid treatment. By BILL GOOLSBY /ledge of first aid procedures is|| e to all personnel to provide Safety Dir., Delta Drilling Co., Tyler, TX ground for coping with medich A supervisor's sincerity in accident gencics both at work and at hoi ^prevention must be combined with a iving employees are encouraged! P working knowledge of the basic fun- advantage of free evening first 1 f damentals of accident prevention before .es offered by the Red Cross, coumKj-J^, he can successfully control the oc ;ity hospitals, adult education, an<L-- currence of accidents under his supervi programs offering first aid trainife sion. certification available throughoq " The supervisor was once charged with ;orld. getting out the production, but times Id operations groups should P'have changed. Today, his responsibility iar and have copies of the new U$*j ITis to get it out economically and main- Diving Manual. In this manual s tain a harmonious relationship with his :d section on first aid treatment tHS^.... !'Cfew. d be read by all diving supervisors The offshore rig makes for an ideal others responsible for the physical" ^situation, with supervision readily ?eing of personnel while workiii|^i Available to schedule the work. Yet we ore and in the shop. Know this||| ^Still see accidents from the following rial well--you will not have time fcSf| '^cpuses: improper use and maintenance of it if an emergency arises. In addi^M ^'tools; faulty practice in operating to these references, all operating! tvTnachinery; neglect in the care and >s should have a summary first aids>v{ ^'maintenance of equipment; deviation for inclusion in their first aid kit 5 from established work practices; lack of chart should provide simplified lfl ^training and know-how in the part of .ions for recognizing symptoms and j^&few members; little value placed on the l treatment by those personnel wH< E<Y>value of care and accuracy; and lack of ot have extensive backgrounds infill ^"communication. tid. What do the above have in common? Human element type weaknesses. Offshore crews spend considerable ' time handling pipe. Equipment for handling pipe consists of slings, spreader will depend on how pipe is loaded. 3. Sufficient help should be available. Crew should be instructed on load and handling methods. Which way do you want it to go on the deck or in the hold? Collars ahead? Communication can save you many headaches. 4. Be sure to rack and secure all material that is placed on drill vessel. Adverse weather can develop before you have time to secure the entire vessel. An offshore location has characteristics that are different from most other locations. All materials and equipment used for the operation of the rig and the drilling of the well must be handled a number of times before they are used. Here is a short list of things that must be lifted on board: people, food, laundry, chemicals, water, mud, pipe, special tools, sand blasting equip ment, and special equipment. The list can go on. It is nothing to see a $ 1,000 TV being lifted aboard an offshore rig by the crane. If you think a land rig has ajob nippl- ing up, just think how this can be com pounded by the placing of the rig 100 ft. or more above water. Safety belts and life jackets are a must for people who are working on the nipple up operation. bar, pipe hooks, pipe roller, pipe chocks, .^wooden stripping, and pipe racks. Ex cessive handling of pipe as a result of poor planning has resulted in many ac cidents. A few points that are important Points of importance in the nippling operation: be sure that all equipment is available, clean, and ready for use; handrails, belts, lifelines, and preservers must be available and used; signals and in offshore pipe handling: communication must be agreed on; print 1. Service barge of vessel should be of stack should be available for check, as tied up as near to where pipe is to be the operation proceeds; teamwork is a racked as possible. This eliminates must when communication is difficult. ' handling and the moving of material This information was presented to get over people. you to thinking of ways to prevent your 2. Determine if pipe is-properly material handling accidents. Discuss ,' chocked for handling before tie downs your work methods and equipment with are removed. Your method of handling your crews. 17 " 'O' - HAZARDS OF HYDROGEN SULFIDE process. Ther IN REFINING the effects of and may not hours: br< By WES WALLACE lymphocitosis Vallen Corp., Houston, TX turbances. \ dividuals exp I would like to begin this discussion of hazards of hydrogen sulfide in refining operations by quoting from one of the reference documents, "Hydrogen Sulfide Poisoning In Texas" by Dr. C. M. Aves of the then Humble Oil and Refining Company in an address before the Texas State Medical Association on May 9, 1928: "I have interviewed and examined every man possible, known to have been unconscious with the gas, and have yet sulfide include the fact that it is flam* mable, with a range of 4.3 to 45.5 pet cent in air. A mixture of two volume^ H2S and three volumes oxygen will ex* plode violently when ignited. Our pur* pose here, though, is to discuss the health hazards of H2S rather than the hazards of fire and explosions. We are concerned with very small levels of parts per million rather than the flammable limit of 43,000 parts per million. H2S -it- become more (than less; in c become accu hydrogen su j posures will [creasing syi j prolonged exp | low as 100 par ' chronic poison It is difficult Ithe physiologic Imot subject hur to find one who has any recollection of heavier than air, with a specific gravity land exposures having smelled the gas before going un of 1.192 and a molecular weight of 34.0% I"for precision. F der its influence. The sense of smell is as compared to 29 for air. This gives us | respiratory pai certainly of no assistance as a danger concern for low places, ditches, trenches^ \ ! range of 500 to warning." Hydrogen sulfide is usually referred to as a colorless, transparent gas with a characteristic rotten egg odor. The younger people of today have probably and similar areas. Hydrogen sulfide does: f not always hug the ground, however, as J in higher concentrations such as 2,000- < i parts per million, the specific gravity ot^uv ( density of the mixture of air and H2S is}|-; j more, dependii ;Death will occu : case such as thi I Removed immec [. ibreathing is sta never smelled a rotten egg, so perhaps we could change the description around and say that a rotten egg smells like hydrogen sulfide. The disagreeable odor of hydrogen sulfide occurs only at relatively low concentrations. only 1.003. So we can see that thej, i " $ion prior to ces hydrogen sulfide may be bouyed up by*? word aboi air currents and carried to elevations!^ t lydrogen sulfu above ground level, as it diffuses rapidlyk^, tine to a caustic in air and may contaminate a wide area?^^ I tjie alkali of the quickly or, depending upon the luck of ipodium sulfide. From Paul Halley's work we learn that those who have been rendered un conscious from high concentrations and survived reported that they did not notice the so-called rotten egg odor, but the wind velocity and direction, it may be dispersed rapidly. H2S is soluble ii water in the proportion of about 4.' volumes of gas to one volume of water al ambient temperatures. It will help t Jkhat hydrogen s\ aesthetic effect sins and sympt after exposure hi 14? to 96 hours ai I rather that it had a sickening-sweet odor. decontaminate an area of H2S gas fc should be immec Others reported that the odor was sweet spray the atmosphere and flush down thi jiLet's now t; and not unpleasant. Just exactly at what area with a water fog spray nozzle, '^g-' |pjiysical effects concentration it becomes sweet-smelling We used to believe that hydrogen sul$|- depending on th is not known, but it is believed to be in fide was a non-cumulative poison. Re|i ^ill deal here \ the 500 to 1000 ppm range. We should cent API toxicological data leads us tq|- ather than volu; keep in mind, also, that some few people state that recovery from acute H2.|v Sipndred standar cannot detect the odor of H2S at all, and it would be prudent to determine this in evaluating whether to allow an employee to work around potential hydrogen sul fide exposures. Other physical properties of hydrogen poisoning is usually complete, but there!" will be occasional instances of lastinj injury following prolonged (can w||\:', define^prolonged as four minutes a-*4'1 longer?) oxygen deprivation of the brail due to the arrest of the breathiri^p|' ^tabular form, we : parts per mill: physical effects, t Ste have an ob\ dor. For many jggin at five pa 18 Petroleum Sessions GEN SULFIDE process. There will also be cases where others will claim they can smell it as low the effects of the exposure are delayed as .025 parts per million, and so forth. At MG and may not become evident for several 20 parts per million it is safe for eight hours: bronchial irritations, hours exposure, and that is the ceiling kCE on, TX lymphocitosis, and gastro-intestinal dis turbances. We now know that in dividuals exposed repeatedly to H2S concentration. The threshold limit value is 10 parts per million. At 100 parts per million it kills the smell in 3 to 15 become more sensitive to the gas, rather minutes and may sting the eyes and e include the fact that it is flam* [than less; in other words, one does not throat. At 200 parts per million it kills with a range of 4.3 to 45.5 p$fc (become accustomed to exposure to the smell rapidly and stings the eyes and n air. A mixture of two volume^ (hydrogen sulfide, and repeated ex throat. At 500 parts per million, diz ind three volumes oxygen will exy posures will produce gradually in ziness, lack of balance, judgment is im violently when ignited. Our puni creasing symptoms. Repeated or paired,. breathing ceases in a few here, though, is to discuss t% prolonged exposure to concentrations as minutes. Prompt artificial respiration is : hazards of H2S rather than th& Blow as 100 parts per million will result in indicated. At 700 parts per million the Js of fire and explosions. We are ['chronic poisoning. person becomes unconscious very rned with very small levels of parts- jjj It is difficult to be precise on some of quickly; death will result by asphixiation lillion rather than the flammabfe |:the physiological effects because we do if not rescued promptly. At 1000 parts Df 43,000 parts per million. HaS.jb.- j.not subject humans to the necessary tests per million the person becomes un :r than air, with a specific gravity I and exposures which would be necessary conscious at once; this is followed by ?2 and a molecular weight of 34.0% I for precision. However, we can state that death within minutes. npared to 29 for air. This gives usf [respiratory paralysis will occur in the With that as a background, where in rn for low places, ditches, trenche^ range of 500 to 700 parts per million or refining operations do we need to be milar areas. Hydrogen sulfide does; more, depending upon the individual. concerned about hydrogen sulfide? Well, ways hug the ground, however, a& - Death will occur from asphyxiation in a with the amount of sour crude processed her concentrations such as 2,OOOi.f. 1 [- case such as this unless the individual is today, I think we can begin at the dock, per million, the specific gravity 01$ \ [.removed immediately to fresh air and if product is received via barge or tanker, y of the mixture of air and H2Ssi||y' j Inbreathing is started by artificial respira- and at the gate valve if it is coming in by 1.003. So we can see that th^l- gition prior to cessation of the heart beat. pipeline. So, we have to begin by train gen sulfide may be bouyed up byl^ ttjV)A word about exposure to the eyes. ing our dock personnel, gauging and dis rrents and carried to elevatiopf^^, ^Hydrogen sulfide does affect the eyes patching personnel, and by providing ground level, as it diffuses rapidij|K- iue to a caustic action when H2S meets protection for them. and may contaminate a wide area#..': the alkali of the cells in the eye to form From here the product would y or, depending upon the luck of^V ijsodium sulfide. It has been theorized generally move to storage tanks, so we ,nd velocity and direction, it may . jy rthat hydrogen sulfide may have a slight have concern with them and with the persed rapidly. H2S is soluble ii puiesthetic effect on the eyes, as often the pump houses that serve the storage area. in the proportion of about 4.4J jjspains and symptoms do not occur until Then we go to the operators in the dis es of gas to one volume of water: ('after exposure has ceased, sometimes for tillation unit, the gas absorption unit, nt temperatures. It will help t< :48 to 96 hours after exposure. Eye baths and the sulphur plant as we take our first :aminate an area of H2S gas toi/*. should be immediately available. liquid stream off the distillation tower of .he atmosphere and flush down the|. jp-Let's now tabulate some of the light naptha or gasoline. You will find it dth a water fog spray nozzle, re ^physical effects of hydrogen sulfide, will carry H2S with it going to storage used to believe that hydrogen sul>| ?depending on the concentration, and I prior to going to a reformer, or it may go as a non-cumulative poison. Re^' (twill deal here with parts per million direct and the off gas can go to the vPI toxicological data leads us toi ather than volume in air or grains per desulphurizers before going to the that recovery from acute H.2.S8 ' ^Hundred standard cubic feet. Thus, in reformer. ing is usually complete, but there.;,. ^tabular form, we have under one column When we cut the kerosene/jet fuel we ; occasional instances of lasting ^ie parts per million and in the other the have to be concerned at the hydroformer following prolonged (can s i^iysical effects. At 10 parts per-jnillion and with the intermediate storage. The prolonged as four minutes oifjfr jcwe have an obvious and unpleasant heavy gas oil cut will have lesser ?) oxygen deprivation of the brainIp. |Vgdor. For many people the odor will amounts of H2S as it goes to the 0 the arrest of the breathing'll |;|>egin at five parts per million, while hydrocracker. As we proceed to the cat 19 iv/o isaiionai bajety Congress cracker feed stock with intermediate hydrogen sulfide which is scheduled tb" storage again, we will have a lesser be transmitted to OSHA in early 1977;' amount of hydrogen sulfide with less NIOSH recommends that continuous concern for toxicity; at the very bottom monitoring, with an automatic audibly of the crude column we will have residue, signal, be required in certain areas where bunker C, asphalt, and so on, and sudden exposure to hydrogen sulfide if primarily here we are dealing with odor. possible. Their first concern is in cori-s 1 think we can go through certain fined spaces such as entry into tanks? operations and catch the worst offenders pits, process vessels, and trenches. Here* or the most likely places of hydrogen they recommend that such entry be co'rt^ sulfide exposure: trolled by a permit system with tH^ 1. Sampling, gauging, drawing water authorized employer representative cef-P from, blinding, going on top of or work tifying that adequate preparation hff ing in pits near sour crude tank or sour been made for entry into the confined" product tank. space, that precautionary measures anitf* 2. Going on top of agitators in which personal protective equipment are ad< some stocks are being treated. quate, and that precautions have beeii' 3. When drawing off caustic soda, taken to insure that prescribed water, or acid from treating operations. procedures will be followed. ThejP If acids come in contact with hydrogen further recommend that confined spaced sulfide loaded caustic soda sent to which have contained hydrogen sulfid'd* sewers, a tremendous volume of H2S shall be inspected and tested for oxygen6 may be released at ground level. deficiency, H2S, and other con'*2 4. Working in pump rooms where sour taminants, and shall be thoroughly ven crude and stock pumps are located. tilated, cleaned neutralized, or washed? 5. When working around gas treating and then retested for hydrogen sulfk0 plants. These plants are set up to remove and oxygen prior to and during entry. IP H2S from certain products. buildup of hydrogen sulfide into the coii^ 6. Refinery flare lines almost always fined space when work is in progress is'tp carry gases in which there is some possibility, then buildup shall bf3 hydrogen sulfide. prevented by positive means such a'l2 7. Sewer catch basins are points of forced air ventilation, securing intake hydrogen sulfide release. valves, or disconnecting or blinding in-; 8. Any excavation near any hydrogen take lines. -: sulfide equipment or lines that may con Individuals entering confined spaces' 4 tain H2S. If the excavation is in excess of where they may possibly be exposed to* three feet deep, we recommend tests hydrogen sulfide shall wear either a comm it should be made for hydrogen sulfide. bination Type C supplied-air respirator? 9. H2S in the crude when released to operated in the positive pressure o^ < the atmosphere has the characteristic of pressure demand mode, and an auxiliarj&i ! expanding 40 times; that is, 80 ppm in self-contained breathing air supply, or ^ crude, as may be reported by the marine self-contained breathing apparatus^ chemist, becomes 3200 ppm in air when operated in the pressure demand mode* vaporized. This is especially true in complete with a full facepiece. Each ihF ` vapor spaces above the liquid level. dividual shall also wear a suitable One of the items on our reference harness with life lines attended b^- material..is the NIOSH work practice another employee outside the space wli?^ recommendations, and it should be shall also be supplied with the necessatjfi|> emphasized these are interim work prac protective equipment, including a selP| r tice recommendations. NIOSH is contained breathing apparatus of tH$- preparing a criteria for a recommended pressure demand type. In areas labeled**!? standard on occupational exposure to "suspect," job activities which couldb 20 Petroleum Sessions gen sulfide which is scheduled ftp nsmitted to OS HA in early 1977? H recommends that continuous1 oring, with an automatic audibly , be required in certain areas when? n exposure to hydrogen sulfide is? 'le. Their first concern is in conspaces such as entry into tanks irocess vessels, and trenches. Here* ecommend that such entry be edit# d by a permit system with fH# -ized employer representative cigpp . that adequate preparation -nade for entry into the confing: that precautionary measures a'iftfi lal protective equipment are adS# and that precautions have bee$ to insure that prescribe dures will be followed. Thgjjp r recommend that confined space# have contained hydrogen sulfid# e inspected and tested for oxygerP .ency, H2S, and other corf mts, and shall be thoroughly ven^ , cleaned neutralized, or washed^ len retested for hydrogen sulfide* cygen prior to and during entry.TP p of hydrogen sulfide into the cbff&:pace when work is in progress is#? dlity, then buildup shall b#' * ted by positive means such d^'T- air ventilation, securing intakl?;.or disconnecting or blinding in-' nes. <S`';; viduals entering confined spaces" ;*- they may possibly be exposed to^., ;en sulfide shall wear either a coirf^ >n Type C supplied-air respiratorled in the positive pressure dPp -e demand mode, and an auxiliary? ntained breathing air supply, or'iPf mtained breathing apparatus ;d in the pressure demand mod^' :te with a full facepiece. Each irP* al shall also wear a suitably ;s with life lines attended r employee outside the space wfrcP* Iso be supplied with the necessary. ive equipment, including a self^ led breathing apparatus of th|; e demand type. In areas labele3|:. :t," job activities which could [expected to potentially expose workers that these are the only kinds of breathing ] to high levels of hydrogen sulfide such as apparatus that we should be using in Lank gauging, maintenance operations, hydrogen sulfide service. [and line repair, shall be protected by I know that in many gauging [adequate monitoring, ventilation, and operations, in remote storage areas, in [personal protection. tank farms, and in production work in We know that olfactory nerve fatigue remote areas, it is a hardship for people [occurs with hydrogen sulfide and that to be asked to wear the self-contained jfhe sense of smell is rapidly dulled or breathing apparatus, and it is sometimes filled, and thus it is possible for difficult for the employer to readily workmen to enter an area where high provide it. For this reason, compromises [concentrations of H2S are present and be are made, and in certain gauging and overcome. In our opening remarks we pumping operations, canister type masks stated that the sense of smell was no are used, I do not believe this is the best [protection. Thus, under the work prac practice, but if the concentrations are tice recommendations from NIOSH, we known and it is not possible for the con [find the monitoring requirement: "Some centration to exceed the limitations of fteans of constant monitoring with an the canister, if the employer is providing (automatic audible warning device shall an approved canister mask for hydrogen je used in places where sudden release of sulfide and the limitations of the equip [hydrogen sulfide might not be expected ment are expressly known to the wearer, land would not be recognized without then perhaps exceptions can be made. In [such continuous monitoring." refining operations, however, the use of !There is further work called for in pressure-demand self-contained or com [training and education, as the effec- bination equipment, NIOSH-certified, is Itiyeness of any work practice program indicated. [depends on proper training and educa Now for the instruments that will tion of the employees. This would in- monitor or detect hydrogen sulfide gas. fclude training not only in first aid and One of our best devices is the detector ^emergency procedures, but also in per- tube method, and I think of such trade pnal protective equipment, with special names as Draeger, MSA, Kitagawa, and |emphasis on breathing apparatus. Gastec, all of whom have NIOSH- fI would like to quote from the API certified tubes for measurement of |recommended practice No. 49, entitled hydrogen sulfide in air in the toxic range. |afe Drilling of Wells Containing Detector tubes and pumps can be used jHydrogen Sulfide," Section 8, by almost anyone, with little training re iParagraph 8.1, where we read: "Masks quired. They give a quick, almost instan [which are designed to merely neutralize taneous reading with low initial cost. Itoxic gas do not provide the necessary Standard combustible gas indicators will [protection and should never be used in not read in the very low ranges required; [drilling operations when a hydrogen sul- in.fact, some of the popular brands are tfjde environment may be encountered." actually poisoned by hydrogen sulfide. [This same section goes on to recommend There are a variety of so-called super [that pressure-demand fresh air breathing sensitive instruments which will read (equipment, complete with a low-level hydrogen sulfide in the parts-per-million fajarm signal, should be used either in range. [self-contained or in combination As we have drilled deeper, as we have [pressure-demand airline breathing depleted reservoirs, as we have gone into ^equipment, with an auxiliary self- secondary and tertiary recovery, and es |cqntained air supply for emergency pecially as we have imported larger and |egress. From this, and from the ANSI larger amounts of foreign crude, we are ^Standard Z88.2, I draw the conclusion encountering more and more exposures 21 V luff to hydrogen sulfide. Luckily, the scien table form for use on the body as a detec tific effort has kept pace, and we now tor, but it can be built into systems fot have portable and line-powered in use in monitoring a plant or production strumentation for the detection and area. monitoring of hydrogen sulfide which There has been a great deal of con are much more convenient to use and troversy and question revolving arouiid much more reliable than earlier the newer instrumentation. Admittedly, methods. in terms of years, it is rather new, and Certainly, no discussion of hydrogen there have been some false starts and sulfide detection would be complete incomplete tests. Fortunately, th6 without mention of lead acetate ampules manufacturers who have lent their eft and lead acetate tape, both of which forts in this area are exceedingly reliable' have been used for many, many years in and they are aware of the necessity fcf the petroleum industry. providing instrumentation.which will err We know that upon exposure of lead on the safe side and which can operate acetate paper to hydrogen sulfide the for long periods of time with little or rid tape will turn silvery black to black and maintenance. While more expensive that the degree and speed of blackening than the typical combustible gas ih- depends on the concentration of the dicator with which we are all familiar?!! hydrogen sulfide, moisture, exposure believe it is necessary to bear in mirid time, or volume. With the ampules, we that we are asking the instrument to r&! have to be careful not to use them in too spond to very minute quantities of gaS prolonged a test because they lose their and to provide some sort of output sensitivity as they dry out: - signal, so we are now dealing with a One of the more popular newer quite sophisticated solid state device* methods employs the use of a metal ox More important, the reliability is there? ide semi-conductor, generally operating I can end this discussion as I began it!> on the diffusion principle. Another by referring to that talk of long ago til method of detection employs an electro the Texas Medical Society where Dr? chemical cell which gives almost an im Aves made the statement: "I have inter* mediate readout on a display meter in viewed and examined every man possfe terms of parts per million. This unit has ble, known to have been unconscious the additional advantage of very fast with the gas, and have yet to find one recovery so that successive tests can be who has any recollection of having made very quickly. It lends itself to line smelled the gas before going under its power and sequential sampling, and thus influence. The sense of smell is certainty it is possible to have it not only in por of no assistance as a danger warning."! ib .s ) !\ lI INDUS Mgr. A refiner; . engineering sk. and pipelines ;tion of crude c variety of petr products which economy of t |, world are bein To the casua not trained in only possible h The refinery is ; the novice in . presence of occ j, with the excep obscure. The c . "toxic" agei -background is itcan be detecti [,,refinery worke refinery, and to . that depend on refinery, the ` referred to as " During nortr ..in a refinery, th -> to occupationa operating unit ..reason for this i. . are in the unit a |. a closed system for exposure oc : ..pf a unit or unit t * opened, catalyst r tanks, pressure cleaned and we |i stripped and rej for exposure cor back on stream. The workm .greatest expos ..maintenance per to perform thei ^complex. The p form for use on the body as a de&&' 5ut it can be built into systems f&f 1 monitoring a plant or producti6'A' *& ere has been a great deal of coll rsy and question revolving arouifij ewer instrumentation. Admittedly^ ms of years, it is rather new, aitf) have been some false starts am} nplete tests. Fortunately, fjjw facturers who have lent their*%j| in this area are exceedingly reliaBt^ hey are aware of the necessity?^ iing instrumentation which willifi^ e safe side and which can opercfli ng periods of time with little or'la tenance. While more expenstfljl the typical combustible gas -ffjit >r with which we are all familiar^! e it is necessary to bear in miitff /e are asking the instrument to ftfi to very minute quantities of 0 provide some sort of outptfl , so we are now dealing with a sophisticated solid state devices? important, the reliability is theft!.*' n end this discussion as I began-ilk. erring to that talk of long ago f# exas Medical Society where D^i made the statement: "I have intef^ 1 and examined every man possfc, nown to have been unconscious he gas, and have yet to find one has any recollection of having' d the gas before going under its*- ' ice. The sense of smell is certainlj? assistance as a danger warning.-il;- t.`,q ib -T ;U .11 INDUSTRIAL HYGIENE CONSIDERATIONS IN REFINERY OPERATIONS By MELITON M. GARCIA Mgr., Occupational Health, Tenneco, Inc., Houston, TX A refinery is a masterpiece of are probably the unit operators. ^.engineering skill--in this maze of vessels Automation and centralization of con L and pipelines a marvelous transforma trol instrumentation have contributed tion of crude oil is taking place; a great greatly to the reduction of operating per ?variety of petroleum and petrochemical sonnel exposures. Of course, during turn r products which have a great effect on the around.most refinery personnel become t economy of this country and on the involved and the exposure risk increases. [(. world are being produced. At this time, contract personnel are To the casual observer or to a person brought in to perform all types of work not trained in occupational health, the and they are also exposed to the poten only possible hazard they associate with tially harmful agents. the refinery is a fire or an explosion. To The major occupational health .the novice industrial hygienist, the hazards that can be identified in refinery L presence of occupational health hazards, operation include noise, dusts, I}.with the exception of the noise, is also chemicals, and sanitation. Lobscure; The only evidence that some " `toxic" agent is lurking in the Noise E background is the "refinery odor" that Noise is the most obvious and per Lean be detected at all areas. To the vasive harmful agent. It is generated by L.refinery worker, to the owner of the electric motors, steam turbines, cooling j-, refinery, and to the community members fans, forced draft fans, primary and . that depend on the jobs provided by the secondary air intakes for burners, steam .refinery, the "odor" is affectionately leaks, high pressure air and steam let referred to as "the smell of money." downs, pressure control values, power ... During normal operating conditions generators, compressors, sand blasting, in a refinery, the potential for exposure fluid transmission pipes, flares, and f> occupational health hazards in an mechanical equipment needing operating unit is at a minimum. The maintenance or replacement. Sources |..reason for this is that very few personnel like the high pressure steam and air let are in the unit and the products are all in down can generate 125 dBA at a distance a. closed system. The greatest potential of 50 feet from the outlet. In a typical for exposure occurs during turn around refinery, at least 75 per cent of the area [ ..of a unit or units. At this time vessels are bounded by the battery lines of the h opened, catalyst is replaced, surfaces on whole refinery complex may exceed 85 [.tanks, pressure vessels, pipes, etc., are dBA. cleaned and welded, and insulation is Past and present activity in noise con > stripped and replaced. A high potential trol by industry and the enforcement of for exposure continues until all units are the OS HA noise standard has resulted in back on stream. the development and installation of The workmen experiencing the various noise control devices in refinery 'greatest exposure potential are the operations. For example: flares have . maintenance personnel who ace-assigned been equipped with silencers; cooling to perform their labors in the refinery fans have been designed with more complex. The personnel least exposed blades and operate at slower speeds; 23 f vu/151 CJJ forced draft fans on boilers or stacks Exposure to silica occurs when the the pot< have been equipped with silencers; vessels and pipelines are sandblasted. attending pipelines have been lagged with Not only is the person doing the . periencec acoustical materials; electric motors of sandblasting exposed, but so are the loading t the totally enclosed fan cooled (TEFC) refinery workers who may be working in vapor ret variety have been equipped with an area downwind from the sandblasting 3. Lab mufflers; primary intake air openings on operation. Recent studies by contractors the man; burners have been equipped with mutes; for NIOSH have shown that hydrocar acoustic plenums have been constructed sandblasters are definitely exposed even are ill-eq around the secondary air intakes for though they are wearing the common capabiliti burners; turbines, pumps, and motors sandblasting hood and supply air. <have been enclosed in acoustic booths; To ameliorate exposure to silica, 4. Per; knock ac high pressure steam and air let-down many employers have changed from stantly b outlets have been equipped with river sand to fine fused slag known as posure to silencers; greater attention has been "Black Beauty" and "Stan-Blast." Better Proper ,pr given to the repair of steam leaks, to the respiratory protective hoods are also be by perso replacement or repair of bad bearings ing used. ventilath and to dampening of vibrating equip When a unit is shut down for turn working ment; control valves are being replaced around, catalyst beds are changed. Ex spaces, with "trim or drag" valves that posure occurs when the workmen enter kerosene drastically reduce the noise at the valve the confined spaces to remove the that decc body and eliminate transmission of the catalyst or when the vessel is being clothing i: noise along the pipeline. charged. Toxic substances that may be 5. Perse Adoption of ope or more of the noise found in the catalyst include: silica, tion unit, control techniques has resulted in signifi nickel, vanadium, iron, cobalt, and other is used as cant reduction of noise levels at the substances. Use of appropriate protec strict prec source and significant changes in the tive clothing will reduce the exposure to very corrc quality of the ambient noise at the unit catalyst dust. General and/or local ex Impervioq and refinery as a whole. haust ventilation must be used in con extent of Dusts The dusts to which a refinery worker fined spaces. Chemicals ( by the wo location ir. 6. Wher can be exposed, especially the The chemicals to which refinery per posure to i maintenance and crafts personnel (com sonnel can be exposed, either in the gas the crude ; pany or contractor), include asbestos, eous or liquid state, are legion. The$e. the closed silica, catalyst, coke, and residue or scale chemicals can be raw products, in the sweeter accumulated inside vessels or pipes. termediate products, or final saleable posures ca Exposure to asbestos occurs when the products. The following are examples of- leading to old insulation is stripped from the vessels some of the common potential exposure? 7. Expos and pipes, when the new insulation is to be anticipated: % in ancillary trimmed to fit the vessel or pipes, and 1. Operating personnel who sampll* tal lead use during clean up of the waste insulation the product streams, check the product, in vessels, c and debris. Preventing exposures re storage tanks, and open drain pipes on welding or quires that the material to be stripped be various vessels or water knock-out tanks lead oxide - wetted down before stripping begins, are exposed to varying concentrations of outside sur that the scrap be disposed of in closed hydrocarbon vapors. ! elemental rr. and labeled containers, that dust- 2. Personnel at the product loading ing devices: producing operations such as trimming locations, tankers and barges may;!? that inclu( and grinding be done with the use of depending upon the prevailing at exhaust ventilation, and that personnel mospheric conditions, experience signifi- 7 ti caustic ma used asleve wear respiratory protective devices when cant exposures to hydrocarbon vapors:^ fumes frc conditions warrant it. A recent study conducted by Shell Oil operations. 24 icposure to silica occurs when the :1s and pipelines are sandblasted, only is the person doing the blasting exposed, but so are the ery workers who may be working in rea downwind from the sandblasting ation. Recent studies by contractors NIOSH have shown that blasters are definitely exposed even gh they are wearing the common blasting hood and supply air. > ameliorate exposure to silica, / employers have changed from sand to fine fused slag known as :k Beauty" and "Stan-Blast." Better ratory protective hoods are also beised. hen a unit is shut down for turn nd, catalyst beds are changed. Exre occurs when the workmen enter confined spaces to remove the yst or when the vessel is being ;ed. Toxic substances that may be j in the catalyst include: silica, 1, vanadium, iron, cobalt, and other ances. Use of appropriate proteclothing will reduce the exposure to /st dust. General and/or local ex ventilation must be used in con- rspaces. ncals e chemicals to which refinery per il can be exposed, either in the gasor liquid state, are legion. These deals can be raw products, indiate products, or final saleable icts. The following are examples of. of the common potential exposure^ anticipated: Operating personnel who sample roduct streams, check the product ;e tanks, and open drain pipes on is vessels or water knock-out tanks :posed to varying concentrations of carbon vapors. Personnel at the product loading ions, tankers and barges may, iding upon the prevailing atheric conditions, experience signifiixposures to hydrocarbon vapors. :nt study conducted by Shell Oil of Petroleum Sessions the potential exposure of personnel confined spaces or on contaminated sur attending to the loading operations ex faces. perienced their highest exposure when loading tankers with top loading and no Sanitation vapor recovery systems. Refineries are show places when it 3. Laboratory personnel who analyze comes to housekeeping and outward the many samples are exposed to the appearance in the units. Unfortunately, hydrocarbon vapors; most laboratories this zeal for cleanliness is lost when it are ill-equipped as far as local exhaust comes to lunch rooms, restrooms, water capabilities are concerned. fountains, change rooms, and places 4. Personnel who work with anti where food is prepared and served in the knock additives for gasoline must con unit. stantly be on the alert to prevent ex If occupational health hazards in posure to those extra-hazardous agents. petroleum refining operations are to be Proper protective clothing must be worn identified, quantified and controlled and by personnel, adequate local exhaust if prevention of exposures are the objec ventilation must be provided when tives of the management, a comprehen working with these agents in. confined sive industrial hygiene program must be spaces, and an adequate supply of instituted to accomplish the following: kerosene must be available in the event 1. Examine the work environment that decontamination of the skin or and environs: (a) study work operations, clothing is necessary. processes, and materials to assess the 5. Personnel working in the alkyla nature of the potential for exposure to tion unit, especially if hydrogen fluoride harmful conditions. This will require an is used as the catalyst, must observe very assessment of work habits, equipment strict precautions if contact with this use, materials used as raw and in very corrosive agent is to be avoided. termediate materials or products and by Impervious clothing must be worn; the products, number of persons exposed, extent of protection needed is dictated duration of exposure, and mode of ex- by the work to be performed and the posure; (b) make appropriate location in the unit. measurements to establish the concen 6. When sour crude is processed, ex tration of contaminant or magnitude of posure to hydrogen sulfide can occur at exposure. the crude storage tanks or anywhere on 2. Interpret results of field studies, ex the closed system prior to its removal at aminations, and measurements to deter the sweetening unit; from this point, ex mine the ability of the stresses to cause posures can occur in the closed system illness and the degree of hazard potential leading to and at the sulfur plant. created by these stresses. 7. Exposures to toxic substances used 3. Make specific recommendations as in ancillary operations include: elemen to the need for and type of control tal lead used to seal the connecting joints measures necessary to minimize or in vessels, or lead fumes generated while eliminate exposure. These measures may welding or burning on painted surfaces; include one or more of the following: lead oxide used as a primer paint on the isolation of the process or work opera outside surfaces of vessels and piping; tion to reduce the number of exposed elemental mercury used in pressure sens personnel; substitution of a hazardous ing devices; water treatment chemicals material with one of no known hazard or that include chromates, acids, and at least one of lesser hazard; alteration of caustic materials; radioactive sources the process or work cycle to minimize used as level gauges and density sensors; human exposure; appropriate applica fumes from welding or brazing tion of ventilation practices and equip operations, especially when working in ment to provide an atmosphere safe for 25 human occupancy; provide shielding or the education of managers and plant enclosures and limit the time of exposure employees in the prevention of oc- and distance between the stress agent cupational diseases. and the worker; good housekeeping 8. Conduct epidemiologic studies practices, including cleanliness of the among workers and similar industries to work place, proper waste disposal, ade discover the possibilities of the presence quate washing, toilet, and restroom of occupational diseases. Correlation of facilities, healthful domestic water plant studies, worker exposures, and in supply and lunch room facilities, and cidence of disease will help in guiding the control of insects and rodents; ap development of realistic standards by propriate personal protective devices governmental agencies. such as respirators, hearing protectors, 9. Coordinate with various company and other special clothing consistent organizations such as manufacturing, with the nature of the hazard. production, safety, fire protection, 4. Review plans for new or modified medical, employee relations, engineer- ` expansion of facilities or processes; ing, etc., on matters relating to the use of review engineering designs of facilities hazardous substances and control of per and health hazard control methods; and sonnel exposures. review and approve materials purchased for use in process or operation. This Conclusion practice will assure that occupational To the casual observer who tours an health hazards are recognized and oil refinery while it is in full operation,* eliminated while in the planning or the presence of occupational health design stages rather than retrofitting hazards and the potential for personnel after the fact. exposures are not obvious. However, to 5. Prepare realistic rules, regulations, the professional industrial hygienist, the standards, or operating procedures to subtle exposures that occur while the conduct work in a health-safe manner. refinery is operating smoothly and while 6. Prepare appropriate methods of the refinery units are on turn-around are alerting the worker to the hazards real and present cause for action. associated with the materials or process. Maintenance personnel have the highest This should include material safety data risk potential and turnaround operations sheets, warning labels for bulk and small pose the greatest opportunities for ex containers, placards at the site of poten posure. Noise and chemical dusts, mists, tial exposure, and hazard control and vapors are the hazardous agents of ^ bulletins. greatest concern. 7. Develop and conduct programs for t& 26 (: k Sc com boarc clima a fu: busin high of Sc furtht "accic tion o as sue short, and i: creatir This for me growtl ment i safety porate of this cident the cor ment nr his owr needs c the co motivat (i.e., fo selected for "Prc "Accide Facili sound referenct Alleir booklets service s employ books ar dated. AH acc nation of managers and plant es in the prevention of oc- f al diseases. "'t' nduct epidemiologic studies' orkers and similar industries to " the possibilities of the presence .*** ational diseases. Correlation of '} dies, worker exposures, and in- ' f disease will help in guiding the* lent of realistic standards by" :ntal agencies. rdinate with various company ions such as manufacturing, on, safety, fire protection,/' employee relations, engineer-* an matters relating to the use of** s substances and control of per-* posures. >n casual observer who tours an y while it is in full operation,* ' :nce of occupational health/, nd the potential for personnel / are not obvious. However, to sional industrial hygienist, the posures that occur while the operating smoothly and while y units are on turn-around are present cause for action, ice personnel have the highest tial and turnaround operations greatest opportunities for ex cuse and chemical dusts, mists, s are the hazardous agents of v >ncern. ?- Petroleum Sessions SERVICE STATION SAFETYWHAT CAN YOU DO? THINK ABOUT IT! By C. W. HAVEN Safety Coordinator, Marketing Department, The Standard Oil Co. (Ohio), Cleveland, OH Sohio's safety program starts with the commitment of the chairman of the board to "strive to create a working climate that establishes safe operation as a fundamental way of running the business." In order to insure that these high ideals become an operational part of Sohio's philosophy, the concept was further translated into the statement that "accident prevention is a control func tion of management. It must be accepted as such by all levels of management." In short, Sohio's management has accepted and is .committed to the challenge of creating a safe working climate. This commitment has been in effect for more than 25 years. It is not an out growth of OS HA nor any other govern ment imposed programs. Each year a safety program is developed by the cor porate safety department. The purpose of this program is to provide general ac cident prevention guidelines throughout the company. In addition, each depart ment manager is encouraged to develop his own program reflecting the specific needs of his department. During 1976 the corporate program emphasized motivation of the front line manager (i.e., foreman or supervisor). The logo selected was the international traffic sign for "Prohibit Movement" with the word "Accidents" superimposed on the face. Facilities are constructed according to sound engineering practices with reference to API, NFPA, NEC, etc. All employees are provided with safety booklets. These booklets are used by the service station managers as part of the employees' on-the-job training. The books are periodically reviewedTtnd up dated. All accidents--personal injury, vehicle accidents, fire, third party incidents--are promptly investigated and reported. The rules governing chargeability are those found in ANSI Z16.1, 215.1, OSHA, and API. Information developed from an analysis of the accident reports (all accidents regardless of severity) is used to pin-point trouble areas and to provide specific information for accident preven tion programs, and quarterly and monthly safety meetings. Until this year, analysis of the accident reports was carried out manually. We have developed a computer program which we are now using to assist in this analysis. Specific objectives for disabling in juries, index injuries, and vehicle ac cidents are established each year and ap proved by the marketing vice president for each operating division. Every month a report is published showing perfor mance to date as compared to objectives. The method used to establish these ob jectives is a least squares curve calcula tion based upon past performance with zero accidents at infinity as the ultimate objective. For example, the disabling in jury objective for this year is 1.8. Currently we are at 2.1 and working hard to meet this objective by year end. In.addition to the pocket safety booklet, there is also the service station safety poster. We recognize the problem of providing a safety program for company-operated locations spread through 13 states. This problem is com pounded by a high turn-over of per sonnel, mostly in the 18- to 25-year old group. To offset this problem a program was developed with the single purpose of providing station employees with infor- l s i \j jium/au( uujcty uy/lg/ CJJ mation on the types of accidents he may or at the pumps is a constant danger. be involved in and a means by which he What can you do? Never assume that a himself could avoid injury. customer is a skilled driver. Also, never On the surface, this might appear very assume that the customer's vehicle can i i elementary. However, accident analysis stop when it should. Be polite and tact confirmed by the recent computer data ful; however, stand out of the danger demonstrates that failure of the zone until the vehicle has stopped and its employee to recognize hazards and to engine is turned off. Think about it! Are take personal action accounts for 95 per a few weeks in traction worth the risk? cent of the personal injuries in the Sohio 3. Cuts. Many vehicles on the road marketing department. today are poorly maintained and are This has led us to the development of loaded with sharp edges waiting to rip an audio visual program for use at the open an unwary finger, hand, or leg. The station level. This program is entitled: sharp edges themselves cannot jump out "Service Station- Safety--What Can you and attack you; only your failure to Do? Think About It!" The program con recognize them can cause you to be hurt. sists of 30 slides covering three separate What can you do? Recognize the sections and a cassette which runs 18 potential danger of broken head lights, minutes. The sections covered are: On rusty torn fenders, loose chrome, or the Driveway; In the Lube Bay; Fire sharp edges on license plates. And hav Prevention. ing recognized them, stay clear of them. Following are ten subjects which I feel Think about it! A cut hurts, but the are frequently overlooked in discussions stitches will hurt even more. of service station safety. This is the way 4. Radiator Burns. Modern these ten subjects are presented at the pressurized cooling systems operate nor service station. mally at temperatures of about 240 F. 1. Blow Backs. Factors relating to The temperature is above the boiling weather conditions, the condition of the point of the coolant. This is the reason customer's gas tank, etc., lead to the con for the pressure under the cap. If the cap dition known as blow. back. Simply, gas is opened under these conditions, there is oline blows back out the fill opening. a good chance the coolant will boil out. The major danger of this type of incident What can you do? Do not check a is that this may develop into a serious radiator unless the customer absolutely burn case. When you have gasoline on insists. If you must check a radiator, your clothing, you are a potential walk then let it cool for at least one-half hour. . ing torch. You can be set on fire either by Place the radiator safety tarp over the .. your own actions or by a customer or cap before you attempt to open it. Work another employee. on top of, not under, the tarp. ;i What can you do? Avoid standing Think about it! Scalds are extremely J| directly in front of the fill opening. Stand painful. Do you want one? If you should 3 to the side. Should you get gasoline on ever be burned, remember the best first your clothing and/or skin, remove the aid for a bum is to cool it down with clothing immediately and wash the skin cold water for 15 minutes by the clock; area with soap and water. Stay away then go to a doctor or hospital from anything that can ignite gasoline emergency room. vapors. 5. Slips. Another way to have a bad Think about it! Would you like a fall in a service station is to slip on oily, vacation in the bum ward? Remember, greasy, wet, or soapy floors. Oil and gasoline-soaked clothing can burn like a grease are a part of service station torch. operations; however, they do not belong 2. Struck By Customer's Vehicle. Be as spots on the floor. Spills, when they. ing struck by a car either on the driveway happen, must be cleaned up im- 28 mediat part ol never c falls wi when p Wha grease, happen alert employYou ha' others t Think traction posedly 6. Str year seve in the Iu you star motor a wheel, yc You are ; to drive Frankly, pletely fa What c side, pref always rei room. Ne' for that n guiding it. Think a l ing off to legs crushe 7. Work ing under a exposed tc materials ft road salts, will fall do inches awa Where do } Other eq can cause grinder, too impact wrer air chisel, i' What can tlon in the something i decide to we ' Think abo M JJ.VftiV* I pumps is a constant danger,;^ i you do? Never assume that av$ jfis a skilled driver. Also, neveri lat the customer's vehicle caniv< i it should. Be polite and tact-/ ver, stand out of the danger*".*? the vehicle has stopped and itsf Tf turned off. Think about it! Are* ;ks in traction worth the risk?-; . Many vehicles on the road54 : poorly maintained and ar^ ! ,th sharp edges waiting to rip-;-, nwary finger, hand, or leg. The es themselves cannot jump ou|, 4 :k you; only your failure tor f*' them can cause you to be hurt.* t can you do? Recognize the danger of broken head lights^. . n fenders, loose chrome, or es on license plates. And havlized them, stay clear of them,- ^ out it! A cut hurts, but the ill hurt even more. .diator Burns. Modern* d cooling systems operate noremperatures of about 240 F. >erature is above the boiling he coolant. This is the reason jssure under the cap. If the cap under these conditions, there is ance the coolant will boil out. an you do? Do not check a inless the customer absolutely you must check a radiator, cool for at least one-half hour, radiator safety tarp over the ; you attempt to open it. Work v , not under, the tarp. ibout it! Scalds are extremely >0 you want one? If you should jrned, remember the best first bum is to cool it down with r for 15 minutes by the clock; to a doctor or hospital 1 room. . Another way to have a bad ;rvice station is to slip on oily, ' et, or soapy floors. Oil and e a part of service station 5; however, they do not belong >n the floor. Spills, when they must be cleaned up im mediately. Wet or soapy floors are also a I part of service station operation. They -B never caused a fall; they do contribute to falls when they are left unattended and B when proper caution is not used. I What can you do? Clean up oil, I grease, and water spills when they B happen. When scrubbing floors, stay B alert and caution others (other B employees and customers) to stay clear. B You have a responsibility to yourself and B others to avoid this type of accident. B Think about it! Are the months in B traction worth the few seconds sup B posedly saved by not cleaning up a spill? B 6. Struck By Customer's Car. Every B year several of our employees are pinned B in the lube bay by the customer. When B you stand in front of a car with the B motor running and a driver behind the B wheel, you are gambling with your life. B You are assuming the driver knows how B to drive and that the car will stop. B Frankly, both assumptions are com B pletely false. | What can you do? Stand off to the side, preferably the driver's side; but always remember, give yourself running B room. Never stand directly in front of or, for that matter, behind a vehicle when guiding it. Think about it! Which is better? Stand ing off to the side or risk having your legs crushed? 7. Working Under Lifts. When work ing under a car on the lift, you are always exposed to the possibility of foreign materials falling into your eyes. The dirt, road salts, rust, metal chips, stones, etc. will fall down. You're looking up only inches away from the falling materials. Where do you think they will go? Other equipment in the lube bay that can cause eye injuries are the bench grinder, tools with mushroom heads, the impact wrench, the tire machine, and the air chisel. What can you do? Wear eye protec tion in the lube bay. Don't wait until something is in your eye before you decide to wear eye protection. Think about it! How valuable are your Petroleum Sessions eyes? If you don't know, ask a blind man. 8. Wearing Jewelry. Have you ever met anyone with the ring finger on their hand missing? Chances are the finger is gone because they wore a ring while working. Ring fingers have been lost because of contact with electric circuits, catching on moving machinery, or catching on projections while using a wrench. What can you do? Isn't it obvious? Don't wear jewelry while at work. This includes rings, metal watch bands, earrings, and necklaces. Think about it! You have ten fingers. Would you like to try for less? 9. Servicing Batteries. Every year at least 100 batteries explode in our stations. The general reason given is a defective battery. However, the real cause is that someone created a spark near the battery. Batteries give off an explosive mixture of gases, and a spark can set it off. The most common mistake happens when taking a battery out of a car or when installing one. Do it right; take the ground off first, and put the ground on last. This eliminates the danger of a spark caused by the wrench shorting out. Also, do not smoke near batteries. What can you do? Follow the rule-- ground off first, ground on last. 10. Smoking. Have you ever smoked in the lube bay? Did you ever realize the risk you are taking? If gasoline spills from a ruptured gas line, a spilled gas can, a leaking tank, or any one of many possibilities, that smoke could be your last. A spill without a source of ignition is a mess to clean up. A spill with a source of ignition can result in an explo sion or a fire., What can you do? Don't smoke in the lube bay. Don't permit anyone (customers or other employees) to smoke in the lube bay. Think about it! If you survive the ex plosion or fire, you're in for a long vaca tion in a bum ward. Is it worth lighting up? 29 inv nuuunut oujeiy congress PIPELINE EXPLOSION - BACK IN THE "NOT-SO" GOOD OLD DAYS By WAYNE KINISON Safety Supvr., Shell Pipe Line Corp., Houston, TX On December 12, 1939, a section of it was economically available; (3) the 26.8 miles of 10-inch crude oil pipeline ease of locating and detecting leaks by was literally blown out of the ground in using air when the air is odorized with' eight minutes time with explosions or ethyl mercaptan; (4) the ease of detecting detonations running up to 60 to 70 per air escaping from leaks by using an mile. This section of line is in Shell Pipe audiophone; (5) the disproportionate Line Corporation's McCamey, Texas-- cost of other test methods; and (6) the Cushing, Oklahoma 10-inch, a part of presumption that displacement of oil by the system that furnished crude oil to air was a common and safe practice in Shell's Wood River, Illinois and East the oil pipeline industry. r Chicago, Indiana Refineries. About two weeks before the air tests,- Specifically, the damaged section was Go-Devil type scrapers were run in the : between Archer Station in Texas and oil stream to clear the line of debris. The Ryan Station in Oklahoma. first scraper stuck in one section of the At the time of the explosions, com line, but was freed by striking the line pressed air was being used to push with a wooden skid. It was believed an scrapers to displace the crude oil in the accumulation of debris had stopped the pipeline.' This practice, with some scraper. Other scrapers ran without variations, had been used by other delays. organizations in the industry. The air procedure was started at. Immediately after the disaster, an in Baylor Station, and a 38-mile section of vestigating committee was appointed the line to Archer Station was air and sent to the scene to make a complete pressured, oil vacated, and leaks found investigation. The facts and conclusions and repaired. The procedure was to in- j presented here are, for the most part, sert two hard rubber disc scrapers into those of this committee as taken from the line. The hard rubber discs were the / their official report. same diameter as the inside of the 10- $ The reason for displacement of the oil inch pipe. Air was then forced into the with air was to find and repair external line back of the scrapers by five air conn corrosion leaks in the Pipeline. The soil pressors with a capacity totaling apt I 3! was extremely dry, and leaks in the line proximately 600 cubic feet of air per f had been difficult to locate. The oil loss minute at a maximum discharge pressure I was significant. On many occasions, the of 90-95 psi. The lubricator of one of the v oil would disappear into the dry ground compressors was used to feed ethyl men# without evidence of a leak on the sur captan into the line. As the air column face. The decision to use air to displace moved the scraper through the line, the oil and to test the line for leaks was maintenance men followed them to keep based upon: (1) the failure of locating all track of their position and to search for; the leaks by other standard methods leaks. As each successive point of tried; (2) the lack of available water to leakage was passed, with resulting loss of use water for pressuring the line--the air, the scrapers moved more slowly and: | difficulty and time limitation in ob finally stopped 27 miles from Baylor Sta-, | taining water from the Brazos or Red tion at 10 a.m. December 9, 1939. *:i -J Rivers, the only available sources where Eight leaks had been found. They were | 30 m dug out ant clamps, whe travel. Plugg shutdown ol pressors at moved at ft speed) and : miles from I The follow was removec stream side further cleant the scraper t number of n and a subst; Most of the had been roll> the size of bir Next, a Goserted at Bayl column at 6:1 crude oil was: back of the s being forced c section by the being forced ii tion of the line The mainten scraper all nig miles per hour, line at Baylor v 8:14 a.m. on I 8:00 a.m. gaug< 19,000 barrels t the line, push scraper to withii and the front scr of Ryan Statior On the mom division superii with the crew. C feet ahead of the the scraper (one one being the si fourth man superintendent's the scraper. Th( which the line tra standing on top of a mile away, v of the explosion s driven up and U - BACK IN >LD DAYS i Houston, TX Dnomically available; (3) the :ating and detecting leaks by .vhen the air is odorized with aptan; (4) the ease of detecting ng from leaks by using an .. .e; (5) the disproportionate ner test methods; and (6) the '> jn that displacement of oil by . common and safe practice in eline industry. i. wo weeks before the air tests, type scrapers were run in the to clear the line of debris. The ;r stuck in one section of the vas freed by striking the line ; oden skid. It was believed ah ion of debris had stopped the;- Dther scrapers ran without w procedure was started at; .tion, and a 38-mile section of :o Archer Station was air oil vacated, and leaks found ed. The procedure was to in- ard rubber disc scrapers into he hard rubber discs were the ieter as the inside of the 10- Air was then forced into the ^ >f the scrapers by five air com-t s vith a capacity totaling ap-. -J ly 600 cubic feet of air per | t maximum discharge pressure ;i. The lubricator of one of the ' rs was used to feed ethyl mer- o the line. As the air column e scraper through the line, ce men followed them to keep ieir position and to search for: each successive point of ts passed, with resulting loss of apers moved more slowly and j >ped 27 miles from Baylor Sta-> i a.m. December 9, 1939. iks had been found. They were Petroleum Sessions dug out and repaired with leak repair walking toward the group about 500 feet clamps, whereupon the scrapers resumed away at the time of the explosion. travel. Plugging the eight leaks permitted Suddenly, the scraper stopped, the shutdown of three of the five air com superintendent motioned the car driver pressors at Baylor. The scrapers then to stop, which he did and shut off the moved at four m.p.h. (double former engine. The two men knelt down to try speed) and reached Archer Station, 38 to pick up any sound of the scraper. miles from Baylor, on December 10. Within an estimated 10-15 seconds, there The following day, the leading scraper was a muffled explosion right under was removed and reinserted in the up them and almost immediately a second stream side of Archer Station. It had explosion. The first muffled explosion further cleaned the.line by carrying into occurred at the scraper and merely the scraper trap and strainers a large bulged the pipe but did not burst it. The number of rocks and metallic objects, second explosion, or rupture in the pipe, and a substantial amount of sludge. occurred directly under the car. The car, Most of the latter, largely iron oxide, with the driver in it, was hurled ahead in had been rolled into small pellets about the direction the scraper had been travel the size of bird shot. ing and over the heads of the two men Next, a Go-Devil type scraper was in listening for the scraper. The car landed serted at Baylor Station behind the air on its side 145 measured feet from its column at 6:15 p.m., December 11, and previous position and about 20 feet off crude oil was started through the line in the pipeline. When it landed, it was fac back of-the scraper. Now the air was ing in the opposite direction. being forced out of the Baylor-Archer While the car was in the air, a third section by the oil and, in turn, air was explosion occurred toward Ryan Station being forced into the Archer-Ryan sec about 145 feet from the second explosion tion of the line. and just 20 feet from where the car The maintenance crews followed the landed. Almost simultaneous with the scraper all night, averaging about two third explosion, a similar eruption oc miles per hour. Oil was pumped into the curred some 240 feet back of the scraper line at Baylor without interruption until and in the direction of Archer Station. 8:14 a.m. on December 12, 1939. The Other explosions then occurred, travel 8:00 a.m. gauge showed that just over ing at a high rate of speed the entire 19,000 barrels of oil had been put into almost 27 miles back to Archer Station. the line, pushing the rear Go-Devil At about the time of the explosion, the scraper to within one-half mile of Archer foreman glanced at his watch and and the front scraper to within nine miles noticed it was 7:50 a.m. He ran to his of Ryan Station. truck, drove a short distance to the com On the morning of the accident, the pany telephone line which paralleled the division superintendent was walking pipeline, and tried to ring Archer Station with the crew. One man was walking 25 but was unable to get them because the feet ahead of the scraper, two men with explosions had torn down the telephone the scraper (one on each side of the line, lines. However, in the other direction the one being the superintendent), and the telephone lines were intact and he was fourth man was driving the able to get the dispatcher at Cushing, superintendent's car about 20 feet behind who called Archer Station on the Bell the scraper. The owner of the ranch System. While the dispatcher was issuing which the line traversed at this point was instructions to the Archer Station standing on top of a hill three-quarters operator, the line blew up in the station of a mile away, where he had a"fall view yard. The records at Archer Station of the explosion scene. The foreman had showed there was an elapsed time of driven up and left his truck and was about eight minutes between the initial 31 im( ou;ci t LW/JAfW and final explosion, indicating an moment of fracture, two very clean and average velocity of about 300 feet per bright irregular metallic surfaces which second for the 26.8 miles. were ideal for the discharge of an elec The ranch owner previously men trostatic spark, and which had not tioned, who was standing on a hill with a previously occurred in this section of the good view of the scene of the first series line. of explosions, stated that they threw up The resulting spark ignited a com great clouds of dust and "traveled faster bustible mixture of hydrocarbons and than an airplane." He was about three- air, which started the instantaneous quarters of a mile from the line and saw propagation of a flame in both directions the last visible eruption go over a hill to and caused a muffled explosion at the the west, a distance of 1 Vi miles, before he scraper. This first explosion was not suf heard the sound of any explosions. ficient to rupture the pipe but bulged the Two men who worked for a gas com bottom of it. The flame, traveling at the pany were driving down a highway. They relatively slow rate of approximately two said the explosions threw dust and white feet per second, ignited a highly com smoke 60 feet into the air and "looked bustible gaseous mixture approximately like a locomotive coming down the line 16 feet back of the scraper, and produced belching white smoke, but traveling an extremely violent explosion which much faster than a locomotive." deformed the scraper plates, burst the There was only one slight injury in this pipeline, and threw the automobile accident, and it was a minor one. It about 145 feet along the line toward wasn't the driver of the superintendent's Ryan Station. Furthermore, it produced car. He was not injured at all but plenty tremendous pressure waves along the in scared. The injured person was a WPA side of the pipe in each direction, worker on his way to work on the road. resulting in a series of explosions which He drove into one of the big craters that reached Archer Station, 26.8 miles to the had blown out in the pipeline road cross southwest, about eight minutes later. In ing. He was taken to a hospital, ex addition, it caused an explosion some amined, treated for minor injuries and 145 feet toward Ryan Station. However, shock, and dismissed. However, his 1931 the waves from this explosion en Studebaker car was pretty well wrecked. countered a body of oil which prevented Several theories as to the cause were their further propagation in that direc examined; however, the theory that tion. The reactive pressure surge through seems to correspond most precisely to all the blasted open end of the pipe carried the known facts is that the scraper, with it a few barrels of oil which were propelled by air estimated to have been sprayed against the car and upon the compressed under approximately 125 psi superintendent, and also upon the at the point of explosion, struck and nearby ground. attempted to pass a large wooden object in the bottom of the line; namely, a pipe Conclusion skid, originally 4" x 4" x 36" long but 1. It is evident from the resulting ex tapered to 2" x 4" at one end due to wear. plosions and detonations that the line The scraper was an insulated metallic contained a concentration of hydrocar- i object, by virtue of the four neoprene bons in the explosive range. discs and the oil film on the wall of the 2. Although further investigation in pipe. This object was electrostatically dicated that several other companies had charged by slight leakage of high used air to clear short sections of pressure air at high velocity. The wedge petroleum lines in a similar fashion, it effect of the wooden obstruction was suf was not-an industry-wide practice. ficient to fracture the sleeve on the shaft 3. This accident, and maybe one or of the scraper, thereby presenting, at the two similar since, as well as Bureau of 32 Mines e use of a uid hyd safe pro< or an h recomme 4. Yoi procedur pressed a have bee In Sect Procedun r f fracture, two very clean and&yi gular metallic surfaces whichi;/: for the discharge of an elec-^ -4 spark, and which had not$ !- occurred in this section of the^ % " jlting spark ignited a corn** ixture of hydrocarbons and** i started the instantaneous^ ' n of a flame in both directions;*,, 1 a muffled explosion at thet as first explosion was not suf-|' jpture the pipe but bulged the-r\ it. The flame, traveling at then ow rate of approximately two a cond, ignited a highly com-* seous mixture approximatelyrf f c of the scraper, and produced ely violent explosion which the scraper plates, burst the > and threw the automobile feet along the line toward on. Furthermore, it produced.s pressure waves along the inle pipe in each direction, i a series of explosions which cher Station, 26.8 miles to the about eight minutes later. In t caused an explosion some vard Ryan Station. However, . from this explosion eni body of oil which prevented ;r propagation in that direc:active pressure surge through open end of the pipe carried ew barrels of oil which were ;ainst the car and upon the ident, and also upon the und. vident from the resulting exad detonations that the line i concentration of hydrocar; explosive range, ugh further investigation int several other companies had o clear short sections of lines in a similar fashion, it industry-wide practice, iccident, and maybe one or - since, as well as Bureau of reiroieum sessions Mines experiments, have proven that the use of air under pressure to displace liq uid hydrocarbons in a pipeline is not a safe procedure. Instead, the use of water or an inert gas such as nitrogen is recommended. 4. You can be sure that in our safety procedures, we forbid the use of com pressed air to vacate pipelines after they have been in liquid hydrocarbon use. In Section B-94 of our Pipeline Safety Procedures Manual we read: "Pipelines have literally been blown out of the ground for many miles by injecting air into them to replace the crude, product, or gas. Therefore, pipeline scrapers, pigs, or balls must not be forced through a line with air after the line has been in crude or product service, nor shall air ever be injected into a pipeline which contains or has contained a flammable liquid or gas. Only water or an inert gas such as nitrogen may be used safely for such purposes." 33 ly/u national oajeiy congress HIGH LEVEL TANK ALARMS ON PIPELINE TANKS By R. L. JOHNSON Dir., Safety & Environmental Affairs, Continental Pipe Line Co., Houston, TX In September 1972 Conoco ex encompasses errors in calculating space perienced a catastrophic fire at a product available for delivery by the use of the terminal. The subsequent accident in wrong tank table and also simple sub vestigation revealed that the cause of the traction errors. For instance, the safe fill fire was the overfilling of a gasoline level for a tank may be 80,107 barrels storage tank while receiving a pipeline and the gauge is 32,433 barrels; when the delivery. The investigation also revealed latter is subtracted from the former, the that the overfill may have been caused by answer, which should be 47,674, can utilizing the tank tables for a 26,000 sometimes come out 57,674. barrel tank in the calculation of space The problems in the category of com available in a 20,000 barrel tank. While munication between pipeline operators the exact reason for the miscalculation and terminal or tankage operators in was not definitely known, it was ap volves changes in the pipeline schedules parent that a human failure did occur in or rate changes during a delivery. These this case. problems are illustrated as follows: Following this accident, we reviewed 1. A terminal operator may be an our entire operations and discovered ticipating a 65,000 barrel deliveiy as in that overfills into terminal and pipeline dicated by the pipeline delivery schedule; tankage were not as infrequent as we however, the tender arrangement may would have supposed. An in-depth in have changed since the schedule was vestigation as to the causes of tank over published and the proposed delivery may fills was launched and came up with have grown to, say 72,000 barrels. It is basically two categories of human easy to see that, without proper com failures. The first category was human munication between the pipeline failure on the part of terminal employees operating personnel and the terminal in calculating the amount of space operating personnel, you can arrive at a available in terminal tankage to receive situation where the pipeline is trying to pipeline deliveries. The second was an stick 72,000 barrels into tankage that area of communication between pipeline only has space for something slightly operators and terminal operators. over 65,000 barrels. Under the category of human failures 2. Varying pipeline rates, if not com on the part of terminal operators, we municated to the terminal operators, can isolated the following possibilities: result in earlier than anticipated switches 1. Incorrect tank gauges or incorrect at a terminal: For example, a terminal recording of tank gauges. Anyone who operator, based upon a pipeline delivery has ever gauged tanks has occasionally rate of 3600 barrels per hour, may figure made errors in reading the gauge tape or that he can continue the delivery into a in recording the gauge in the gauge given storage tank until, say, 8:00 a.m. book. For some reason it seems to be the next morning. However, if during the extremely easy to look at a tape gauge evening the pipeline rate increases to and see a gauge of 10 feet 6 inches and 4200 barrels per hour, it may be that the somehow write down in the gauge booJL terminal operator will be out of tank 6 feet 10 inches. space at 7:00 a.m. rather than 8:00 2. Improper outage calculations. This A.M. ALARMS lNKS N al Affairs, ouston, TX sses errors in calculating spacev for delivery by the use of the;: ink table and also simple sub-: errors. For instance, the safe fill. a tank may be 80,107 barrels^ ;auge is 32,433 barrels; when thef .ubtracted from the former, thef which should be 47,674, cans ;s come out 57,674. ; oblems in the category of com- on between pipeline operators ainal or tankage operators in- langes in the pipeline schedules hanges during a delivery. These s are illustrated as follows: terminal operator may be an il a 65,000 barrel delivery as in- ty the pipeline delivery schedule; , the tender arrangement may anged since the schedule was d and the proposed delivery may >wn to, say 72,000 barrels. It is ' see that, without proper com- tion between the pipeline g personnel and the terminal , g personnel, you can arrive at a < i where the pipeline is trying to & ,000 barrels into tankage that i s space for something slightly $ 000 barrels. I tying pipeline rates, if not com- ; ed to the terminal operators, can earlier than anticipated switches ' minal. For example, a terminal based upon a pipeline delivery 600 barrels per hour, may figure can continue the delivery into a orage tank until, say, 8:00 a.m. morning. However, if during the the pipeline rate increases to . rrels per hour, it may be that the j 1 operator will be out of tank i it 7:00 a.m. rather than 8:00 Petroleum Sessions The above examples are somewhat ex operated terminals which are served by aggerated for the purpose of illustration, Continental Pipe Line Systems, the but these are basically the factors we ran pipeline dispatchers keep an updated into when we investigated the problem. record of terminal stocks and thus could Our solution to the problem was two quickly spot any gauging or calculation fold. We initiated procedural changes errors. Where our operated terminals are which, if followed, would eliminate the served by other pipeline companies, communication problem as well as some of the companies follow the above minimize exposure to calculation errors. procedures and some vary from these The second part of our solution was the procedures. installation of high level alarms on ter Reinforcing the above procedures is minal tankage. We feel that both facets our system of high level tank alarms of the solution are equally important, which have been installed on all of our because we recognize that we can still product terminal tankage. Most of the have failures in both personnel and installations which we have made in equipment. By super-imposing the alarm volve the installation of a standpipe ex system on our procedural changes, we ternal to the tank. This standpipe is feel that the odds are increased that both generally 114-inch diameter pipe con type of failures would not occur sim nected to the tank by couplings which ultaneously. have been welded to the shell of the tank The procedural changes which were near the bottom of the tank and at some initiated are as follows: point near the top of the tank. Two float 1. One or two days prior to a switches are mounted in the standpipe. scheduled pipeline delivery, the terminal The first switch will give a warning operator will gauge the tank into which alarm. The second switch is positioned at the delivery will be received and the safe fill level of the tank and gives a calculates the space available for the final alarm. The first switch is positioned delivery. such that after it activates, the pipeline 2. The terminal operator then gives and terminal personnel have a minimum the tank gauge, in feet and inches, to the of 15 minutes to switch tanks or close pipeline delivery personnel. The pipeline out the pipeline delivery. When the final deliveryman, who has been furnished alarm sounds, the instructions are to copies of the tank tables, then re shut down the delivery immediately. calculates the space available for the Valves are installed in the connection delivery and assures himself that the to the tank and an outlet is provided at tank will accommodate the scheduled the bottom of the standpipe for testing delivery. purposes. 3. On the day of the delivery, the ter At some of our installations we have minal operator regauges the tank to utilized auxiliary contacts on the ground receive the delivery and recalculates the ' reading automatic gauges instead of the space available. standpipe arrangement. On these in 4. The terminal operator then again stallations we also have a warning level furnishes the gauge information to the and a critical alarm level. pipeline delivery personnel and they The standpipe method of installing the recheck'the outage available in the tank alarm switches on the tankage en against their scheduled delivery and the countered some problems at facilities pipeline delivery personnel and the ter subject to extreme temperatures. The minal operator discusses .jates, exact problems were incurred primarily on fuel time of delivery, and all details of the oil tanks during extremely cold weather. delivery. The temperature of the fuel oil in the This procedure varies somewhat standpipe would be much lower than the between pipelines and terminals. At our temperature in the tank itself and 35 ...... - vn*wj*v`jw 1976 National Safety Congress problems resulted in getting a represen tative level in the standpipe due to the increased viscosity of the fuel oil in ex treme temperatures. This problem has been solved by installing insulation on the standpipes. Alarm annunciators were installed in the terminal offices which give visual as well as an audible indication of both the warning level and final alarm condition. Alarm annunciators were also generally installed in the local pipeline delivery office and in many cases were tied into leased line circuits and taken back to the pipeline operating center. The alarm points of the product level in the tanks were selected based upon maximum pipeline delivery rates and safe fill levels. The 15-minute period between the warning alarm and critical alarm gives adequate time to the pipeline as well as the terminal personnel to make the necessary valve lineups-and switches. There is some variation in the points to which these tank alarms are fed. At terminals served by Continental Pipe Line Company systems, the alarms are all brought into the centralized control center, as well as having local indication at the terminal. At terminals served by other pipelines, we have brought the alarms into the dispatching centers where practical. In all cases this has not been possible, and in most cases they are at least brought into the local pipeline delivery facility. In a very few cases where the delivering pipeline carrier could not monitor the alarms, we havebrought tank alarms into Conoco's cen tralized control center and monitored the alarms in that fashion. A testing procedure has been initiated under which all of the tank alarms are actuated on a frequent basis to assure that the alarm devices are properly working. These tests also check the transmission of the alarm indications and the alarm displays at the various points remote from the terminal. When testing the standpipe type installations, the standpipe is isolated from the tank by closing the valves on the connections to the tank and then pumping product into the stand pipe with a hand pump until the product level reaches the alarm points. On the ground reading automatic gauges, an alarm level is manually simulated. In both cases, the whole alarm system is tested. In summary, we feel that the procedural changes, as well as the in stallation of the high tank level alarms, served to minimize our exposure to tank overfills on our systems. At some times in the past, there has been talk by various regulatory agencies of requiring automatic capabilities for shutting a pipeline stream out of a tank in a high level alarm position. We have purposely avoided this concept in our systems because we feel strongly that any system utilizing a high tank level alarm for the purpose of automatically closing a valve to terminate a delivery does nothing to solve the real problem, but merely dis places the problem from the area of the terminal to the area of the pipeline. In other words, we could visualize an automatic system as providing the ultimate in protection for terminals, but f resulting in pipeline overpressure con- ditions and subsequent ruptures, which ; could result in worse problems. 36 Explc volumes pected he the fire p the poter staggers There systems i today, bi very sma area, whe sor can t scale sup[ area had n The U.S. ( old, ready a series of The pu 18,000 cub was four fe area was f piping, am ship had b service. T1 door and that type ai were closec The purp is possible such a large what type c how the va would react Pressure because of pressure ac and 1301 w high concen successful, . most success I want t research pro design a co Petroleum Sessions .ich all of the tank alarms are on a frequent basis to assure alarm devices are properly These tests also check the ion of the alarm indications alarm displays at the various note from the terminal. When e standpipe type installations, pipe is isolated from the tank.. ; the valves on the connections ik and then pumping product stand pipe with a hand pumpt*;Droduct level reaches the alarm?, 1 the ground reading automatic | in alarm level is manually?? 1. In both cases, the wholetern is tested. nmary, we feel that the .1 changes, as well as the in-' of the high tank level alarms, minimize our exposure to tank n our systems. At some times . ist, there has been talk by gulatory agencies of requiring capabilities for shutting a . tream out of a tank in a high n position. We have purposely this concept in our systems e feel strongly that any system high tank level alarm for the f automatically closing a valve ; ite a delivery does nothing to real problem, but merely disproblem from the area of the ' o the area of the pipeline. In | rds, we could visualize an system as providing the ul- | protection for terminals, but | in pipeline overpressure con- >| d subsequent ruptures, which ilt in worse problems. EXPLOSION SUPPRESSION OF LARGE TURBULENT AREAS By W. A. CROSLEY Executive Vice Pres., Detector Electronics Corp., Minneapolis, MN Explosion suppression for large with a capability of suppressing ex volumes in which' turbulence can be ex plosions in large as well as small pected has to be the greatest challenge in enclosures. Work is being done by our the fire protection field today. However, company to accomplish this, and next the potential need for just such a system year we expect to have such a system, staggers the imagination. but none is available at this time. Dry There are explosion suppression chemical cannons used in these tests are systems in service throughout the world experimental devices. today, but in almost every case it is a I will now describe how the test was very small volume or an unobstructed conducted and the information that was area, where the mechanical pressure sen obtained from these tests. sor can.be used successfully. A large- It is extremely important to detect the scale suppression test in an obstructed exploding fireball in less than 75 area had never been tested until last year. milliseconds after ignition, if the explo The U.S. Coast Guard made available an sion is to be suppressed. If it takes longer old, ready-for-the-scrap-heap tanker for than that, the explosion will travel faster a series of tests. than the extinguishing agent can handle, The pump-room had a volume of and the result will be no suppression or 18,000 cubic feet when the bilge water possibly a partial suppression. was four feet below the bottom deck. The In a turbulent area, the only type sen area was filled with the original pumps, sor that can be used is a radiation type. It piping, and valving in place the way the was proven during these tests that ship had been during its long period of pressure sensors are not suitable. If the service. The ventings were the hatch area was unobstructed and if there was door and normal leakage in a ship of no turbulence, the pressure sensor with a that type and vintage. The exhaust vents Yi psi pressure setting would have been were closed during the tests. suitable. However, there are very few The purpose of the test was to see if it large-volume applications that do not is possible to suppress an explosion in have obstructions in them. such a large area, and also to determine The detector best suited for this what type of detection was the best and application is an ultraviolet radiation how the various types of extinguishants type. Photo cells and infrared sensors are would react. just as fast, but are subject to false fire Pressure sensors were not successful signals from artificial light, sunlight, hot because of the time it took to receive a bodies, and other heat producing bodies. pressure actuation. Halon 1211, 2402, The U V system has spurious signals also, and 1301 were successful, but in fairly but they are much more controllable. high concentrations. Water was partially UV is generated by a flame in combus successful, and dry chemical-"was the tion; however, the following subjects most successful on a concentration basis. also generate UV and care must be taken I want to emphasize this was a to eliminate these signals from the detec research program to gain information to tor. design a complete suppression system 1. Since most hazardous areas do not 37 f tsf u ,'uuunai Mjety congress allow welding without a safety permit, ricocheting off a beveled ring, and then this can easily be controlled by placing re-entering the detector housing through the system in bypass during this period the center of the window, striking the and having a guard stand by with a UV sensor and creating a small signal manual override system during the which enters the signal processing sec welding operation. tion of the electronics. 2. X-rays and gamma rays--if there The signal generated by the UV source are enough rays present to cause an ac lamp enters the UV sensor in the same tuation, there are enough rays to injure location when the UV from a fireball personnel, thus again a safety permit enters it. The signal processing portion would be required. The system would be determines which signal it is. It causes put into bypass during the period x-rays either the solid state relays to actuate the and gamma rays were being used. extinguishing agent or, if it is only the 3. Cosmic radiation is very short in signal from the UV source lamp, it duration, and UV detection systems have signals to the system that everything is in built-in compensation. Cosmic radiation operation from the detector window does not become a concern. through the UV sensor and through the 4. High electrostatic forces could be a electronics. If the signal from the UV problem in some unusual application, source lamp fails to get through, the but again this signal will not be a fault relay de-energizes, warning the problem if the detection equipment is operator of a malfunction. electrically shielded properly during the In a detection/suppression sequence, installation. the UV detector responded between 18 5. A UV sensor that is only sensitive and 25 milliseconds after ignition. to a narrow band of UV, such as from The first contact of the agent with the 1850 to 2450 Angstrom, is not affected fireball occurred at less than 100 by solar radiation on earth. The ozone milliseconds after ignition, and the peak layer attenuates all wave lengths below of the suppressed pressure curve oc 2800 Angstrom. curred at 150 milliseconds at a .2 psi In a basic UV detector system, once overpressure. U V from an exploding fireball enters the The following is a description of the sensor's 90 cone of vision, a solid-state test site and how the equipment was switch will close in less than 10 located. The pump room from the top milliseconds. decks to the bilge water consists of a A UV detection system, like any man trunk which is approximately half the made product, is subject to failure. These width of the ship and is located on its could be tube failure, electronic or wir center line. Two wing spaces open in the ing failure, or optical system failure. lower quarter of the trunk .and extend There are two ways to overcome this the full beam of the ship, port and star probability; one is to have redundant board. systems with periodic tests. A second A limitation inherent in full-scale ex way is to have continuous automatic ex plosion testing is the inability to exactly amination of the detector's electronic duplicate all of the variables which affect component, tube, and optical surfaces. the explosion. This leads to the lack of This has been accomplished by designing repeatability.of the results. In this test, a small UV source lamp and locating it all efforts were made to control quantity inside-the detector housing next to the of fuel, quantity of suppression agent, UV sensor, but completely screened and explosion venting ratio (0.1 ft.2 per from the sensor. Every few seconds the 100 jt.3. Variables which could not be UV source lamp pulses briefly, controlled include environmental factors generating UV which can only reach the such as temperature (varied between sensor by passing through the window, 70 F and 90 F) and relative humidity.' 38 l off a beveled ring, and then ; the detector housing through of the window, striking the and creating a small signal :rs the signal processing sec: electronics. al generated by the UV source s the UV sensor in the same hen the UV from a fireball The signal processing portion which signal it is. It causes olid state relays to actuate the .ng agent or, if it is only the m the UV source lamp, it he system that everything is in from the detector window te UV sensor and through the . If the signal from the UV np fails to get through, the / de-energizes, warning the f a malfunction. :ection/suppression sequence, .'lector responded between 18 lliseconds after ignition. : contact of the agent with the iccurred at less than 100 Is after ignition, and the peak ppressed pressure curve oc- [50 milliseconds at a .2 psi re. owing is a description of the ,nd how the equipment was he pump room from the top the bilge water consists of a ch is approximately half the he ship and is located on its . Two wing spaces open in the rter of the trunk .and extend am of the ship, port and star- ition inherent in full-scale ex iting is the inability to exactly ill of the variables which affect ion. This leads to the lack of ty of the results. In this test, were made to control quantity aantity of suppression agent, sion venting ratio (0.1 ft.2 per Enables which could not be include environmental factors .emperature (varied between 90 F) and relative humidity. 1 Petroleum Sessions Also, most important, they did not milliseconds to distribute the agent. want to blow up the ship. This rapid pressure rise was finally at Test explosions could not be designed tributed to obstruction creating tur for a stoichiometric mixture of propane bulence which increased the flame speed and air throughout the entire volume of and thus the rate of pressure rise. The the pump room because this would evidence is dramatically depicted in the produce approximately 120 psig. Based motion pictures of flame being forced on the design limitations for the pump between obstructions and is further room bulkheads, it was decided that the verified by the reduction in maximum pump room could withstand a maximum pressure and rate of pressure rise in tests of 15 psig. Calculations show that if the U-29 and U-30. stoichiometric mixture of a hydrocarbon In these tests, the independent variable field were placed in 10 per cent of the obstructions were reduced by bringing volume and ignited, then the expansion the bilge water above the machinery in of the explosion into the remaining 90 the lower portion of the pump room per cent of the volume would limit the where almost all of the obstructions were theoretical maximum pressure to 12 located, and this reduced the volume to psig. 15,000 cubic feet. This conclusion has a Also, a successful explosion suppres pronounced effect on the time available sion should hold the maximum explo for detection and actuators of explosion sion pressure to substantially less than suppression systems. It was found that five psig. Thus, it was reasoned that the the UV detectors used in these tests total volume would produce realistic could detect the incipient explosion from results for explosion suppression pur 110 to 400+ milliseconds before a poses. pressure sensor set for lA psig could One would expect the test explosion to detect the explosion. follow the typical pressure versus time The Det-Tronics UV sensors were for stoichiometric mixture of propane then employed in the remainder of the and air curves up to approximately 10 suppression tests. psig. One can further rationalize that this Two explosion suppression systems method is realistic, considering the ex were tested. They were based on the plosion from the fireball's point of view. following principles: As it is ignited and begins to grow, it 1. UV detector sees incipient explo consumes only stoichiometric fuel air sion. mixture at its exterior surfaces. It is not 2. It signals the control circuitry until the fireball has expanded to fully which fires an electrically actuated encompass the mixture that it realizes blasting device. the entire volume of the pump room is 3. The initiators rupture a restraining not included. diaphragm, causing the release of a The Coast Guard then proceeded to suppressing agent. run unsuppressed tests and produced the 4. Driving force throws the agent at predicted maximum pressures on the the flame front. order of 12 psig. 5. The flame is extinguished; the ex The first attempted suppression test plosion suppressed. used pressure sensors set at one-half psi One system consisted of spherical to trigger the suppression system, and high-rate discharge extinguishers to con they were unsuccessful. The pressure rise tain the agent. The driving force was was much more rapid than theoretically provided by pressurizing the charged ex predicted, and thus no more-than 25 tinguisher to 325 psig with dry nitrogen. milliseconds were left from the time of The Detector Electronics UV sensor was detection to maximum pressure. The used for the detection of the Fireball, and suppression system requires at least 100 various suppression agents were used in 39 1976 National Safety Congress this system. They included water, Halon 2402, Halon 1211, and Halon 1301. The other system consisted of cylin drical cannons to contain the agent. The driving force was provided by pressuriz ing the charged cannon to 600 psig with dry nitrogen. Again, the Detector Elec tronics UV detectors were used to sense the fireball, and the suppression agent used was the dry-chemical Purple K. In analyzing these tests, the following factors were considered: size of fireball at agent contact; degree of agent breakdown; after burning subsequent to the initial suppression; and maximum pressure developed. The analysis in dicates that successful suppression re quires the listed minimum application densities for the suppression of explo sion of a propane to air mixture in obstructed spaces similar to a ship's pump room. Water. This agent was only partially successful at application densities of .075 and .071 pounds per cubic foot. While it did reduce the pressure to five psig, it was felt considerably higher application densities would be required to achieve suppression comparable to other agents tested. Halon 2402. This agent successfully suppressed explosions at densities of .071 and .116 pounds per cubic foot, but two factors indicated that these lower densities were approaching the threshold level for suppression: (1) agent breakdown evidenced by the orangeyellow smoke observed; (2) severe after burning occurred. Agent reached fireball in .09 seconds; however, it took 1.03 seconds to obliterate the flame. Halon 1211. This agent was successful at .06 and .09 pounds per cubic foot den sities. While the maximum pressure developed during these tests were slightly higher than.Halon 2402, it is not signifi cant since they were still less than one psig. There was no indication from the data that the threshold application den sity had been approached. Halon 1301. This agent was successful at application densities of .034, .051 and .077 pounds per cubic foot. The lowest application density appeared to ap proach the threshold for suppression. The evidence was similar to Halon 2402. There was severe after burning for .56 seconds after the initial suppression. Purple K. The dry chemical successfully suppressed the propane ex plosion at an application density of .009 pounds per cubic foot. Suppression was marginal at .006 and unsuccessful at .003 pounds per cubic foot. No after burning occurred in any of the three tests. The maximum pressure obtained dur ing these dry chemical tests are plotted against agent application densities. It in dicates that an application density of .007 pounds per cubic foot or greater is required to maintain the pressure below one psig. Increases in density above .01 pounds per cubic foot do not produce a significant reduction in the maximum pressure. In the dry chemical tests, 18 cannons were used. These are the same cannons that Consolidated Coal Corporation used in their methane suppression study for the U.S. Bureau of Mines contract. These are experimental extinguishers, and no off-the-shelf hardware. However, the UV detection system is commercially available at this time. Ten UV detectors were installed throughout the entire pump room, the same as would be necessary in an actual installation. If any one of these testers sees UV radiation, it will fire all 18 can nons. The circuitry was designed to in dicate at any time if there was an elec trical malfunction in the wiring to any of the cannon detonators. 40 OFFIC. General Che Vice Generat TX Program Ch. Secretary are Houston Newsletter E, Director of Petroleui Corporat Technical Put Lake Ch; Standards Re Industrial H\ Petroleurr Research--Da side, IL Legislative--J< TX Director of A Dallas, TJ Activities and Gulf Oil C Activities and national A Activities and man), Atla Activities and , Mobil Oil Director of O Bartlesville Division of Ma, Houston, T Division ofMar Houston, T e was no indication from the the threshold application den'een approached. '301. This agent was successful tion densities of .034, .051 and .ds per cubic foot. The lowest on density appeared to aple threshold for suppression, ice was similar to Halon 2402. > severe after burning for .56 :'ter the initial suppression. e K. The dry chemical ,y suppressed the propane exan application density of .009 :r cubic foot. Suppression was it .006 and unsuccessful at .003 r cubic foot. No after burning n any of the three tests, ximum pressure obtained durdry chemical tests are plotted ent application densities. It inat an application density ofds per cubic foot or greater is o maintain the pressure below Increases in density above .01 ;r cubic foot do not produce a : reduction in the maximum Iry chemical tests, 18 cannons . These are the same cannons | isolidated Coal Corporation-; eir methane suppression study J ,S. Bureau of Mines contract.^5 e experimental extinguishers, ff-the-shelf hardware. ;r, the UV detection system is'*; ally available at this time. V detectors were installed^ it the entire pump room, th.ej /ould be necessary in an actual f n. If any one of these testersf adiation, it will fire all 18 caij-jf ; circuitry was designed to iij,\ any time if there was an elecfunction in the wiring to any {>fj m detonators. M OFFICERS OF THE PETROLEUM SECTION NATIONAL SAFETY COUNCIL 1976-77 General Chairman--L. W. Kinison, Shell Pipe Line Corporation, Houston, TX Vice General Chairman--Walter D. Jenkins, Continental Oil Company, Houston, TX Program Chairman--David E. Miller, Mobil Oil Corporation, New York, NY Secretary and Cameron Award Chairman--J. P. Allison, Frank B. Hall & Co., Houston, TX Newsletter Editor--Roland Gay, Halliburton Services, Duncan, OK Director of Technical Activities--**J. F. McKenna (Temporary), American Petroleum Institute, Washington, DC; E. H. Stockdale (Alternate), Gulf Oil Corporation, Pittsburgh, PA Technical. Publications--John A. Judice (Chairman), Cities Service Oil Company, Lake Charles, LA Standards Review--J. M. Connor (Chairman), Sun Oil Company, St. Davids, PA Industrial Hygiene and Health--Wendell L. Ward (Chairman), American Petroleum Institute, Washington, DC Research--David H. Markland (Chairman), U.O.P., Inc., Process Division, River side, IL Legislative--Joe F. Gay (Chairman), Champlin Petroleum Company, Fort Worth, TX Director of Activities and Membership--Doyle E. Stegall, Sun Oil Company, Dallas, TX i Activities and Membership (Atlantic Region)--George F. Rideout (Chairman), Gulf Oil Company, Philadelphia, PA Activities and Membership (Gulf Coast Region)--W. E. Hine (Chairman), Inter national Association of Drilling Contractors, Houston, TX Activities and Membership (Pacific Coast Region)--Cliff C. Hitchcock (Chair man), Atlantic Richfield Company, Carson, CA Activities and Membership (International Operations)--H. H. Beams (Chairman), Mobil Oil Corporation, New York, NY Director of Operations--Paul F. Lindemann, Phillips Petroleum Company, Bartlesville, OK Division of Marketing--John W^ Keegan (Chairman), Gulf Oil Company U.S.A., Houston, TX Division of Manufacturing--Dean Blackwell (Chairman), Tenneco Oil Company, Houston, TX 41 Division of Pipelines--Ralph E. Powers (Chairman), Williams Pipe Line Com pany, Tulsa, OK Division of Production, Drilling and Exploration--Milton H. Parker (Chairman), Parker Drilling Company, Tulsa, OK Division of Petrochemicals--C. S. Windvand (Chairman), Witco Chemical Cor poration, New York, NY Division of Offshore--Stanley Atherson (Chairman), Getty Oil Company, Los Angeles, CA Director of Motivation and Education--Edward M. Wilson, Jr., Union Oil Co. of California, Palatine, 1L Audio-Visual Aids--Robert J. Cordes (Chairman), Marathon Oil Company, Findlay, OH Training--Bill Goolsby (Chairman), Delta Drilling Company, Tyler, TX Associations--C. W. Listen (Chairman), Phillips Petroleum Company, Bartlesville, OK Off-the-Job Safety--Carroll G. Web (Chairman), Cities Service Company, Tulsa, OK Publicity and Public Relations--P. R. Wrigley (Chairman), National Petroleum Refiners Assn., Washington, DC Advisory and Study Committee--H. G. Rollins (Chairman), Shell Oil Company, Wood River, IL Members, Advisory and Study Committee (Dates of Chairmanship in parentheses)-- *C. H. Klunick (1974-75), Continental Oil Company, Houston, TX; *Frank K. Lightfoot, Jr., (1973-74), Amerada Hess Corporation, Tulsa, OK; *A. G. Aiuvalasit (1972-73), Gulf Chemical Company, Houston, TX; *C. W. Listen (1970-71), Phillips Petroleum Company, Bartlesville, OK; *John G. Lowrance (1969-70), Standard Oil Company (Indiana), Chicago, IL; * Roland Pryor (1966-67), Phillips Petroleum Company, Bartlesville, OK; **Quincy V. Tuma (1961-63), Texaco Inc., Houston, TX Liaison Representatives from Associations--John T. Haggin, Association of Oilwell Servicing Contractors, Dallas, TX; Willard E. Hine, International Association of Drilling Contractors, Houston, TX; Walter H. Johnson, National LP-Gas Association, Chicago, IL; J. D. Manney, American Petroleum Institute, Washington, DC; Herbert E. Robb, Western Oil & Gas Association, Los Angeles, CA; P. R. Wrigley, National Petroleum Refiners Assn., Washington, DC Members-at-Large--Phillip W. Blake, Safety Consulting, Inc., Wichita, KS; James ! Harvey, Texas Pacific Oil Company, Dallas, TX; E. J. Papacek, Texas Eastern Transmission Corp., Houston, TX; C. R. Stevenson, Signal Drilling Company, Inc., Denver, CO; Earl H. Stockdale, Gulf Oil Corporation, Pittsburgh, PA; H. Duane Applequist, The Standard Oil Co. (Ohio), Cleveland, OH; Thomas \ M W. Farrior, Ashland Oil, Inc., Ashland, KY; Wayne E. Wright, Skelly Oil Company, Tulsa, OK; David Cochran, Getty Oil Company, Houston, TX Past General Chairmen (In Addition to Advisory and Study Committee)--Carl B. Adams (U.S. Dept, of Labor) (1965-66); A. G. Aiuvalasit, Gulf Chemicals Co.. 42 Houston, T> Sun Oil Co., W. Breelani H. Brown (R Carr (Retire (Retired) (19< (Consultant) (Retired) (19. (Retired) (195 Myers (Retin Section Administn Past General Chairman `Recipient DSS Award , Williams Pipe Line Com- ton H. Parker (Chairman), man), Witco Chemical Cor- ), Getty Oil Company, Los 'ilson, Jr., Union Oil Co. of Marathon Oil Company, Company, Tyler, TX leum Company, Bartlesville, ies Service Company, Tulsa, irman). National Petroleum irman). Shell Oil Company, airmanship in parentheses)-- any, Houston, TX; *Frank joration, Tulsa, OK; *A. G. 1 ouston, TX; *C. W. Listen e, OK; *John G. Lowrance icago, IL; *Roland Pryor le, OK; **QuiNCY V. Tuma \ Haggin, Association of ard E. Hine, International TX; Walter H. Johnson, J. D. Manney, American . Robb, Western Oil & Gas National Petroleum Refiners ing, Inc., Wichita, KS; James i. J. Papacek, Texas Eastern >n, Signal Drilling Company, Corporation, Pittsburgh, PA; io), Cleveland, OH; Thomas ayne E. Wright, Skelly Oil il Company, Houston, TX Study Committee)--Carl B. /alasit. Gulf Chemicals Co.. Houston, TX (1972-73); **C. D. Attaway, (Retired) (1951-52), G. B. Black, Sun Oil Co., Philadelphia, PA (1959-60); H. W. Boggess (Retired) (1937-38); A. W. Breeland (Retired) (1936-37); Kenneth V. Brooks (Retired) (1971-72); J. H. Brown (Retired) (1938-39); William F. Burris (Retired) (1958-59); S. Ross Carr (Retired) (1963-64); Fred Claiborne (Retired) (1955-56); R. D. Eberly (Retired) (1960-61); D. M. Farrell (Retired) (1949-50); **Parker C. Folse (Consultant) (1957-58); **0. C. Haier (Retired) (1967-68); C. H. Lindberg (Retired) (1956-57); **H. T. Markee (Retired) (1942-43); J. H. McKenzie (Retired) (1954-55); **Fred R. McLean (Consultant) (1941-42); J. Howard Myers (Retired) (1946-47); **George F. Prussing (Retired) (1926-27) Section Administrator--Halit Yurttas, National Safety Council, Chicago, IL Past General Chairman Recipient DSS Award 43