Document ymzy0qnm271aj1XmVdgqwJ7Gn

/v A "" "A am C$jSfntca^ mc ESCAMBIA PLANT INTEROFFICE MEMORANDUM To: Distribution. From:B. A'. Gebbia Re: Process Hazards Review Methanol Plant Dote: 3 December 1980 Copies: R. B. Chandler G. D. Cooper R. L. Duggan W. T. Johnson R. E. Jones T. J. Regan The Process Hazards Review for the Escambia Methanol Plant is scheduled for December S - 12. Please plan to attend the review on each day from 8:30 a.m. to 4;30 p.m. Meetings will be in the old administration building conference room. A Process Hazards Review of the Methanol Plant Revamp was completed in November and would be a helpful reference for these meetings. Also, guidelines for reformer design from PSG were developed for this review. Copies of both items are available from B. Gebbia. The '`What If" method will be used along with the fault tree analysis (for quantitative analysis where necessary) during this review. M. Brown will ensure up-to-date copies of the plant P&ID`s are available for the meeting. B. Cooley will serve as the review committee secretary. BAG :dl Distribution: E. Barnes M. Brown C. Blackledge B. Cooley W. Henry D. Henn J. Marts D. Singer B, A. Gebbia AP00036286 )J"'V Reliability inEtfluent and Water Treatment Plants A. R. Churchley* The value and importance of the continued operation of effluent and water treatment plant is often underestimated until a major breakdown occurs. Occasionally, in common with most other industrial plant, a relatively minor failure wili cause an extensive shutdown, it is usual that some plant equipment is more vulnerable in this respect than others and it is characteristic that there may be an imbalance in the reliability of components and processes making up the plant. In extreme esses, and at longer intervals, enforced shut down ofthe whole treatment facility may be expected to occur. Consequently, considerable benefits may be obtained from a study of the reliability make up of such plant. The major effluent discharging industries are': chemical, petroleum refining, electricity generation, iron and steel, paper and board manufacture, food processing, textiles (natural and synthetic!, quarrying, gas and coke production, soap and detergent manufacture and brewing. Each of these industries produces one or more liquid effluents in large quantity. Water is treated with any one of four objectives in mind, namely: treatment of fresh water for potablet purposes; treatment of sewage to yield innocuous, effluent; treatment of industrial feedwaters; treatment of industrial waste for recycling or discharge. Each of these treatments has a different influent, is directed towards a different end product and is carried out in a different environment. With due regard to these differences, it may also be said that there are considerable similarities between them, for example, in all of these treat* ments settlement, dosing and mixing sometimes play an important role and filtration is often used to polish the end product. Even in sewage works where materials of comparatively high concentration can occur and complex reactions Involving mixed bioloqicai. chemical and physical mechanisms are found, the same basic pumping, piping, vaiving and storage operations are used. Ail of the treatment processes ultimately rely on human supervision to manage, operate and maintain the plant, even on highly computerised units. No matter how capable the plants ere of performing their intended task, as with sny engineer* inb process, they are also vuinerabie to human and equipment failures. The benefits of reliability studies may be condensed down to a simple statement of reduced costs. All enforced shut downs involve undesired expenditure, in the case of an industrial effluent treatment plant, its total failure could prevent production elsewhere inthe works and incur heavy losses of profit. In the case of a water authority, a loss of facility causes inconvenience and iI . r Fig. 1: Principal components of a potable water treatment plant i: Dr. Churchley is manager of Safety. Availability and Reliability Assessment (SARA) Ltd., !i Warrington. UK. Suitable for human consumption and for Oomattie usa. AP00036287 annoyance to the public together with consumption of public funds in restoration expenses. Furthermore, serious environmental and health consequences may occur. The consequences of failures may be severe. In a detailed reliability study fault free techniques are used to estimate the frequency of such end events, and to highlight possible areas for improvement to reduce the frequency of their occurrence to an acceptable level. It is emphasised that these methods of reliability analysis are equally applicable to the other three kinds of water treat ment facility for which different and events would be identified. Human error is not strictly an and event in itself but one element in a chain leading to a serious end event and it is often seen as a very immediate problem by plant management. Figure 1 shows a typical treatment system. Toinvartthis diagram into a failure logic diagram (or fault traa). it is necessary to imagine that each of the stages shown does not perform in the manner intended. Some ingenuity is required on the part of the reliability assessor in order to foresee all the ways in which a Jliven undesired end result may be reached. At the same time, some of the more ikely combinations of events may be eliminated. 1 lMFMin#v OMMI'HkMM Occurwif 3 |ubyi(wiifillm 3 t Ulior tymw 4urw 4 UMtMWl'Ina Cvwki' Fig 2: An aarfy stage in thm construction of the taiiura logic diagram To give an idea of the appearance of a failure logic diagram during the course of its construction. Figure 2 has been drawn. This represents a small and incomplete fragment of a much larger picture, which could easily run into tens of pages in the case of a real-world assessment. The columns drawn in Figure 2 (but not normally shown in such a diagram) indicate that the flow of the Logic diagram occurs from a detailed breakdown of individual subsystems, or component failures, through more generalised failures to the end event which they cause. It gives rise to a number of unanswered questions. How, for example, can the treated water be contaminated in storage? Can a severe oil leak in the pumping house cause serious contamination in the storage reservoir, or is this a comparatively trivial event? Are there other ways in which contamination can occur in the pump house? The diagram refers to excessive treatment by chemicals, but can under treatment occur? How is the treatment controlled? Questions such as these are best answered by the experienced management and operating staff responsible for me water treatment plant. This emphasises the need to build a good working relationship between the reliability assessor and the plant staff and management. When the detailed assessment is prepared, the completed fault tree includes faiiura rates for individual equipment, in terms of failures per year. These failure rates may be deduced if sufficiently detailed maintenance records are available and some of this data may be available in the open literature, although good quality reliability data is often sparse or non-existent, One survey of Instrumentation reliability in wastewater applications has been completed2. Most commonly a recognised data base would be consulted for suitable data, such as the Systems Reliability Service Data Bank. From the fault traa. the areas most liable to failure may be deduced and the mean frequencyof undesired end events determined. It is then possible to test various proposed improve ments for effectiveness in-reducing this frequency. In this way, a very useful cost/benefit analysis may be prepared for the improvement alternatives. AP00036288 The fault tree and failure rate information assists in planning a number of maintenance functions, for example, spares holding and ordering levels may be optimised. Routine maintenance and inspection may also-be planned so that a balance is struck between effort required and the acceptable frequency of end event. The mathematics required3 taderive these benefits are a part of the skillof reliability assessment. To further exemplify the application of reliability techniques to water treatment and at the same time to introduce the safety aspect1 a fault tree is constructed for operator safety, where sludge get is generated in the following example. During the process of sludge digestion on any sewage works a gas mixture consisting mainly of methane and carbon dioxide is produced. Mixed with air, this sludge gas is explosive over a range of concentrations. Figure 3 shows a generalised fault tree leading to a safety hazard to the operator. The safety hazard to the operator arises from two main routes. The lower part of Figure 3 shows some of the ways in which an explosion may occur within vessels or pipe* work if an internal leak occurs, if an operator is present and an internal gas/air mixture has formed, then an internal ignition source causes an operator hazard. A hazard to plant and potential costs due to damage also exists, but this hazard would normally be dealt with on a separate logic diagram. liU Ire*. Vrt-- or EiimwCO Fa.luxef tin CemrolSi Hum teWOUt. OWBW 1 ftorn *MS Central tv<M fipndawiaii Tri0 BIom Out UmtfwSulM Law Sludft Otttlk* Fig. 3: A fault tree for operator safety The internal gas/air mixture may result from a falling sludge level in the digester which reduces the gas pressure, any leeks in valves, flanges or vessels will then admit air. The logic diagram shows that both a leak and a pressure reduction must occur at the same time. Referring again to Fig ure 3. falling sludge t;vel results from an unscheduled opening of the low sludge off-take in the digester, which could occur as a result of a control system failuro, human error or a valve failure in the open position. With a lithe thought this fault tree can be extended further to cover the individual futures which may occur. The upper half nf Figure 3 shows how gas may enter operational areas and Cause a hazard to the operator if, for any reason, the gas detection system fails. There are two reasons why a hazard exists. Even though no ignition source is shown sludge get is an asphyxiating mixture and the operator himself mey cause an ignition since he n. not normally expected to wear anti-static clothing or use spark-free tools. The remainder of the diagram shows some of the locations where feaks may occur, und general causes. In this type of analysis the mean frequency of hazards to the operator may b: n>5essed, and significant improvements are usually derived from the failure `ogic diagrams. The main difference in practice is that a much more comprehensive analysis is carried out. Liaison with management and staff is vitally important in the early stages, particularly when gathering data on how the iiidnt works. Some of the sources and means of recording information are shown Figure 4. When a complete description of the plant operation has been obtained and understood, failure modes can be deduced and fault treee ''(instructed. The configuration of the fault trees and the failure retes used are discussed Fig. 4: Typical assessment procedure with plant management to ensure that errors and omissions are removed and that there ie no conflict with practical plant experience. Finally, the work is written up as a report. The conclusions contain suggestions for improvement and recommendations for optimal routine maintenance testing and inspection. Some reliability studies have already been carried out in the field of industrial feed>waters but it is liketv that considerable scope for reliability improvement exists in this area, particularly on the newer, more highly automated plant. The differing types of treatment plant, (potable, sewage, feedwater and industrial wastewater) share in a common need for reliability and many similarities exist between their equipment and control systems. In all cases, a lack of reliability can be translated into increased costs. Improvements in reliability will increase availability, reducing losses and important benefits may be seen in the maintenance function. Safety may also be improved by increased reliability and, in some cases, an analysis is carried out with this sole objective in mind. The day may come when reliability is regarded as an ingredient of plant success which has the same importance as the hardware from which the plant is constructed. Reliability today is a quantitative discipline, rather than a black art, and may be considered as a commodity which can be purchased through the services of the reliability consultant. In the future, increased reliability will undoubtedly be demanded from water treatment plants and this article is intened to show that reliability engineering assessments would be of value in all the water treatments discussed. Hence, in future, we are likely to see increasing use being made of reliability prediction techniques. Probably the most important message that the reliability engineer would wish to convey is that reliability engineering as stasady well developed in many of the engineering industries and it is now vastly to take on the challenge of water treatment industries. References 1. Royal Commission on Environmental Pollution. Chairman Sir Eric Ashley. 3rd report September 197Z. Cmnd 5054 HMSO. 2. Molvar. A E. et al. How reliable is instrumentation in wastewater applications? Instruments and control systems. October 1977 p29. 3. Green, A. E. and Bourne. A. J. Reliability technology, Wiley Interscience, reprinted March 1977. 4. Health and Safety Commission, consultative document. Hazardous installations (notification end survey) regulations. 1978. Acknowledgement This is an abridged version ot the paper that first appeared In tha Effluent and Water Treatment Journal of May 1979. Quarterly Digest wishes to thank theodltor forgiving permission to make use AP00036290 Courses on Reliability Technology The National Centre of System* Reliability runs a number of regular and short specialised courses on various aspects of Reliability-Technology. Some courses are held at Liverpool University, some at the Centre's headquarters at Culcheth and others (by arrangement! at an organisation's own premises. A number of courses are constructed as standard packages whilst other*, particularly short courses of one ortwodaysduration, can be tailored to meet the specific requirements of individual organisations. Standard-courses include: An introduction to reliability assessment -- theory and practice Location: Liverpool University Duration: Two weeks Frequency: Twice per year. Spring and Autumn Topics: Mathematics of probability. Probability distributions. Reliability testing and maintenance. Equipment failure analysis and prediction. System reliability analysis and synthesis. Data banks. Case studies. Tutorials. Project work and Workshops. An introduction to the reliability evaluation of engineered systems Location: Culcheth Duration: Two days Frequency: Several times per year Topics: Need for reliability evaluation. Quantification of reliability, Active and passive systems, Simple probability concepts. Reliability parameter evaluation. Synthesis of system reliability. Data requirements, Case histories andTutoriats. An introduction to fault-tree methodology and applications Location: Culcheth Duration: Two days Frequency: Two to three times per year Topics: Needfor fault-tree analysis. Various methods of approach. Reliability and probability aspects, System evaluation. Logical analysis and Boolean algebra, Fault-tree symbolism and methodology. Cut-sets, Fault-tree evaluation. Computer program strategies. Case studies and Tutorials. Basic reliability calculations Location: Culcheth Duration: Two days Frequency: Two to three times per year Topics: Historical development, Basic probabilistic concepts, Reliability in time domain, System reliability parameters. Simple distributions. Mean-time to failure, Failure-rate, Unavailability, Probability of failure, System reliability calculations. Series systems, Parallel systems. Majority-vote systems, Tutorials. An introduction to fault modes and effects analysis (FMEA) Location: Culceth Duration: One day Frequency: Two to three times per year Topics: Need for FMEA, Various methods. Fault analysis and classification, FMEA methodology. Tabular presentations, Codes, Case studies and Tutorials. Computer codes for system reliability analysis Location: Culcheth Duration: One day Frequency: Once per year Topics: Need for codes, Straiegies of approach, Common aspects. Review of codes available. Examples of application. Advantages and disadvantages of methods. Tailored short courses or presentations (half to two-day duration) on subjects which include: Reliability evaluation techniques: Reliability data acquisition and analysis; Svsiem reliability modelling; Availability modelling; Common-mode failure analysis: Mechanical reliability analysis; Human reliability factors; Software reliability with other subjects in the field of reliability technologywhich may be included to meet particular requirements. For further technical information or discussions on particular aspects please contact. ; Mr. A. J. Bourne. B.Sb/.' C.Eng., M.LE E. M.lnst.M.C., F.I.QA, Manager. Reliability Technology. National Centre of Systems Reliability. U.K.A.E.A. Wigshaw Lane, Culcheth, Warrington WA3 4NE. Telephone: Warrington 31244, Extension 345, or for further details of. and applications for. the two-week standard reliability course and other standard courses please contact: Miss Mary Sutton, National Centre of Systems Reliability, U.K.A.E.A., Wigshaw Lane, Culcheth, Warrington WA3 4NE. Telephone: Warrington 31244, Extension 313. Comment Mr. G. Volta.t an additional comment on "Computer Programsfor Fault Tree Analysis" (by Andy Cross*. Quarterly Digest January 1980}. I think it will be of interest to SRS and fetiow Associate Members to know about the computer programs developed and in regular use at the Joint Research Centre, Ispra and freely available to all the European organisations through the JRC Eurocopi Service. The following table lists five programs, which I think will be of interest, under the same headings as used by Andy Cross so that these, too, can be addedto his list. MOGMM MAMC ItfUENCt AWE-1 AWE-2 BflUNA DYCOMICS M. AST0LF1 "Pro^ams AWE-1. AWE-2 and 8RUNA lor th ealeuMon of system availability- Calculation methofe and how to um" EUR SS38 *0976) _ M. ASTOLFI. F. NALDI "DYCOMICS Program. How to um." EUR SS11 SALP-3 M. ASlOtFI. S. CONTIM. C. L VAN MUYZENBER, G. VOLTA "SALM. A computer program for fault tree analyse'' EUR 6tS3 EN Computer program* for for* troo analysis. See also COMMENT in the April 1960 Quarterly Been by Richard Store* of the Institute of Engineering, Technical Univareity of Berlin. t Commission of the Eureoeen Communities. Joint Reaeerch Centre, Ispra. Italy * Reliability Reaeerch Engineer, Reliability TechnofogyResearch Unit, NCSR. AP00036292 Personalities Mr. E. S. tondon Director and founder of Hazards Evaluation and Lou ftmMian (Conaultants) Limited. 40 ItwHy Sinai, Wamngtdn. U.K. 6. S. (Selwyn) London, C.Chem. FRIC. who trained at Manchester Collage of Technology. started his sateet in chemtcel plants in 1930 with Jos. Crotltaid & Sons Ltd., m vtfamngtorr where soaps and adibla oils wars processed m IMI ha moved to s government war-gas factory m Runcowi and talar iq on* in North Wales. Attar the war ha pined Ward. Blenktnaop Si Co. Lid., at Widnas, where tins end pharmaceutical chemicals war* made. He than turned to heavy organic chenveels. dyestufta mtermediatu. insecticides, are., with Hickson & Welch Ltd., at Castteford. Yorkshire, m 1948 Mr. London was invited to |0in the UK Ministry of Supply's Atorjuc Energy projtcHnowtneUKAEA) eta plant managarpn the Spnngfields plant m Lancashire. In 1950 he moved to their Windscale works in Cumbria,where ne managed each of mo new eradiated fuel reprocessing plants in turn. Selwyn considers the pioneering period from 1948 to 1956 as the meet eicttmg ot his career and considers himsatt fortunate to havebeen oneofthe few involved in separsimgout tne test catenet ot British piutoniumproeueed in the old Windscale Atomic Piles. Great things were accomplished than despite shortages of people end knowledge. In 1960 Selwyn moved to the UKAEA'i Heellh and Safety Branch (now the Safely and ReiiatMtity Directorate! There he helped to develoo and Introduce the modern oUnt assessment methods end procedure still used throughout the nuclear industry and rapidly being edooted by non-nuclear concerns He became Head of Chemical Plants and Laboratory assessments, put in 1978 Selwyn left to form his present consultancy Ha it also a director of Safety, Availability and Reliability Assessments (SARA) Ltd., which, together with "HELP", provides complementary consultancy service in eefety and reliability assessments thru is available to all Selwyn is also a founder director ot the Warrington Research Centre, one of Europe'* foremost fire testing and consultancy organisation*. Mr. N. Jones Senior consultant with Hassids Evaluation and Less Prevention (Consultants) Ltd.. 40 Bewaay Street, Winington, UK. J ~\ I O'V. -v Ls.................hewi^J N. (Noel) Jones. L.R.S.C.. began his career in 1941 with CourtauU* Limited. Flint. North Wales. UK. where they manufactured viteoee rayon. From t Add to 1948 he served in the Aircrew Brsnchct tnc novel Air Force after which he returned to Courteulds Limited aa laboratory assistant, then shift chem at and (malty production (oraman. Mr. Jon jb joined the UKAEAln 1959 et the Douhrsay Experimental Reactor Establishment. ThurSo. Caithness. Scotland, a* en enatyuoel end development enemiat in the irradiated fuel reprocessing laboratory. Curing mis period Noel ned been iiudylng Chemistry st Thurso.Technicel College and in 1964 ha gained corporate membership of the Royal Institute of Chemistry (now the Royal Society of Chemistry). In 1966 he took over Technical management of en integrated Nuclear Fuel Reprocessing Plant end the associatedHighlyActive Liquor Evaporation and Storage Complex Noel joined the Svaiems Reliability Service, at its mcepl ion in 1970, ase Field Enomeer. He wassetivelv involved in development of the SRS Data Bans, in me organisation of Event Data Collecting schemes at Associate Memoere work*, ihe supervision end guidance of engineering students abstracting reliability dale from maintenance and operational record* at chemical plant* throughout Ihe UK arid, lanerly, ne was engaged on SRS reliability assessment protect work, in 1975 Noel transferred to the Chemical Plants. Ls&roatones and Transport Section oi the Safety and Reliability Directorate (3RD] where he carried out Quantitative safety assessments of new chemical plants He also advised Design engineers working on new plants end reoiesenied SRDbv serving on Safety Working Parties Noe) joined his present company m July 1979. Dr. A. R. Churchley Santor consultant with Safety. Availability and Reliability Assessments (SARA) Limited, 40 Beweey Street, Warrington, UK. ^3>/2__ l Dr. A R (Andrew) Churchley, BSe. Tech , Ph D.. C.Eng., M.lnst.E., A ) Ini.Sc., graduated m Fuel Technology and Chemical Engineering from Sheffield University m 1967. His research work concerned oscillating combustors for gas turbines, in conjunction witn the National Gas Turbin* Establishment. Prestock, Hampshire. During this oenod Dr. Churchley gained experience as an engineering consultant through his asioeiatiori with the development department of the National Carbonising Company. North Nottinghamshire Division In 1970 he joined the engineering department of that company at * project engineer, constructing Europe's most modern lowlemperaiure smokeless fuel production plant ai Srubson. Leicestershire. UK. At mistime ne became a charteredfuel technologist and, later, he became research leaoer in tn* Scientific Services Department, improving the'Qusliiy end profitability of smokeless fuel production In 1979 Andrew joined Unilever Research at Port Sunlight. Wirral. Marseysid* asan Information Scientist He built ud a practice serving the diverse interests of Unilever engineers including, significantly, t group of reliability engineers He became editor of Satety News, an international publication with world wide circulation, and assumed responsibility for the laboratory data preparation group Andrew jorned AE Electronics, e reliability engineering consultancy based et Stockton Heath near Warrington, in 1978 where he was engaged in ltudies of water treatment plant in May 1980 Andrew joined his present company which is the sister company of Harard Evaluation and Loss Prevention (Consultants! Limned. Ha has presented a numoer of technical paper* at National conferences end tymootie end published reliability paper* in the technical prase (see elsewhere m this edition pf Ouanerly Digest) Amongst his spar* time activities Andrew ia keen red* "Ham" and he operates station G4EAQ. AP00036293