628 CCPS Management Guidelines

The CCPS has published several sets of guidance on management. The first of these was A Challenge to Commitment (CCPS, 1985) addressed to senior management in the industry. There have since followed a number of guidelines on management systems, as described below. 6.28.1 Guidelines for Technical Management of Chemical Process Safety The Guidelines for Technical Management of Chemical Process Safety (the Technical Management Guidelines) by the CCPS (1989 7) give a comprehensive treatment of the...

1

The final states of the system may be obtained by solution of Equations 7.10.37 and 7.10.38. An account of discrete-state, discrete-time Markov models is given by Sandler (1963). An application involving a discrete-state, discrete-time Markov model is given in Section 7.13. An alternative to Markov models is the use of the joint density function approach. The joint density function method is convenient where a failure distribution other than the exponential distribution has to be handled. It is...

220

The frequency of all bad outcomes is 22 x 10-6 year. The overall attenuation factor for all bad outcomes is Overall attenuation factor (5.1 x 10-2) (22 x 10-6) 2300 The individual attenuation factors which make up this overall attenuation factor may be derived from the report as follows. For each location there is an estimated length of pipeline which is capable of giving rise to an interaction. The total length for all locations is 32.5 km. Hence Attenuation factor for location 220 32.5 6.8...

12Pressure System Design 121

12.2 Pressure System Components 12 9 12.3 Steels and Their Properties 12 9 12.4 Pressure Vessel Design 12 16 12.5 Joining, Fastening and Welding 12 18 12.6 Pressure Vessel Standards and Codes 12 19 12.9 Fired Heaters and Furnaces 12 36 12.12 Overpressure Protection 12 46 12.13 Overpressure Protection Pressure Relief Devices 12 49 12.14 Overpressure Protection Relief System Design 12 52 12.15 Overpressure Protection Fire Relief 12 55 12.16 Overpressure Protection Vacuum and Thermal Relief 12 58...

27Information Feedback 271

27.3 Operations Monitoring 27 4 27.5 Accident Investigation 27 5 27.7 Explosion Investigation 27 l2 27.8 Accident Investigation CCPS Guidelines 27 l4 27.9 Public Accident Inquiries 27 l5 27.10 Organizational Memory 27 l8 27.12 Information Exchange 27 l9 27.14 Safety Performance Measurement 27 2O 27.15 Safety Performance Monitoring 27 2O 27.16 Near Miss Reporting 27 24

Drainage and spill control

For a flammable material in a process unit with a flash point < 140 F or being processed above its flash point (1) Diking which exposes all equipment within dike to potential fire, penalty 0.50 (2) Flat area around process unit which allows spills to spread, exposing large areas to potential fire, penalty 0.5 (3) Diking around process unit which surrounds three sides of the area and directs spills to an impounding basin or non-exposing drainage trench, no penalty provided (i) Slope to basin...

49Process Hazard Control Advisory Committee on Major Hazards

The Advisory Committee on Major Hazards (ACMH) under the chairmanship of Professor B.H. Harvey, a former Deputy Director-General of the HSE, was set up soon after Flixborough in 1974. The terms of reference of the committee were To identify types of installations (excluding nuclear installations) which have the potential to present major hazards to employees or to the public or the environment, and to advise on measures of control, appropriate to the nature and degree of hazard, over the...

210 Trend of injuries

The long-term trend of injury rates in the process industry is downwards. The trend in the UK may be seen in Table 2.6, which shows that between 1970-74 and 1987-90 the fatal accident rate for the chemical and allied industries fell from 4.3 to 1.2. However, as Table 2.4 shows, the actual number of fatalities in this industry is small. Consequently, one large accident would have a significant effect on the figures, even if its effect were absorbed over quite a long period such as 10 years....

Inherently safer design

Bell (1971) Anon. (1979 LPB 30) Hearfield (1979, 1980a) Kletz (1978c, 1979e,f,n, 1980j,k, 1983 LPB 51, 1984d, 1985h,k, 1988k, 1989c, 1990f-h, 1991f,h,i, 1992c) Middleton and Revill (1983) Anon. (1985t) Anon. (1986q,r) Goodfellow and Berry (1986) Zanetti (1986a) Caputo (1987) S.E. Dale (1987) Anon. (1988f) Barrell (1988a) Gerritsen and van't Land (1988) Lihou (1988) Anon. (1989 LPB 87, p. 21) Hathi, Sengupta and Puranik (1990) Schaller (1990) Englund (1991) Jacob (1991a) R.L. Rogers and...

92026

One limiting case is where both input coefficients have the same value (C0 C1 C2) Another limiting case is where one input coefficient becomes very small. Hence the output coefficient of variation has the range Relation 9.20.39 shows that for an AND gate the coefficient of variation of the output is more than the larger of the coefficients of variation of the inputs, and that in this sense this gate increases the uncertainty in the input data. These results show that strictly in terms of error...

40

FP, Flash point, closed cup BP, boiling point at standard temperatures and pressure (STP). bWill not burn in air when exposed to a temperature of 1500 F (816 C) for a period of 5 min. c -KSt values are for a 16 l or larger closed test vessel with strong ignition source. See NFPA 68 Guide for Venting of Deflagrations. d Includes wood - 2 in. nominal thickness, magnesium ingots, tight stacks of solids and tight rolls of paper or plastic film. Example SARAN WRAP . e...

13 Safety and Environment Concerns

There was also at this time growing public awareness and concern regarding the threat to people and to the environment from industrial activities, particularly those in which the process industries are engaged. Taking the UK as an illustration, the massive vapour cloud explosion at Flixborough in 1974 highlighted the problem of major hazards. This led to the setting up of the Advisory Committee on Major Hazards (ACMH) which sat from 1975 to 1983, and to the introduction of legislation to...

Special process hazards

Low pressure (below 15 psia) High pressure Low temperature 1 Carbon steel +10 C to -10 C 2 Carbon steel below -10 C 3 Other materials High temperature 1 Flammability 2 Construction materials Corrosion and erosion Joint and packing leakages Vibration, load cycling, etc. Processes or reactions difficult to control Operation in or near flammable range Greater than average explosion hazard Dust or mist hazard High strength oxidants Process ignition sensitivity Electrostatic hazards

58 Other Insurance Aspects

The extent to which insurers should encourage and give credit for good practice in general and for loss prevention measures is a long-standing question. Insurers have always been very much involved in measures against fire. The first organized fire brigades in the UK were run by the insurance companies. In the nineteenth century, the Fire Offices Committee was formed and set standards for fire construction, alarms, sprinklers, etc., which have been widely applied. In 1946, the Joint Fire...

57 Business Interruption Insurance

The need for business interruption insurance did not really arise until the mid-1960s, when the first large, single-stream plants began to make their appearance. It then became apparent that consequential losses could be as costly as, or more costly than, property damage losses. The problem of business interruption insurance is a difficult one. The viability of such insurance depends on the charging of realistic premiums, on the evolution of effective assessment methods and on the progress made...

612 Will to safety

There is a danger that the emphasis in loss prevention on technological aspects of hazards and their control may obscure a very simple truth, that in many cases an accident occurs because the will to avoid it was lacking. This theme has been developed by W.B. Howard (1983, 1984), who has given a number of telling examples of accidents where the essential cause was a conscious decision to take a particular course of action. Among the accidents cited are instances where an operator was put in...

0551

A F, failure W, withdrawal original no. 15. a F, failure W, withdrawal original no. 15. The least squares method may be applied in various 1 Ji ways. It is often used to fit data in the form of hazard a 7.2 .32 rates. The hazard rate z(f) and the failure density (f) over each time interval Af between failures are

Info

Stability of reaction mixtures - 2 (Grewer, 1974) (a) adiabatic storage test of diazotization solution of 2-cyano-4-nitroaniline in H2S04 (b) adiabatic reaction test of solution of p-nitro-N-methylaniline and carbyl sulphate in nitrobenzene (Courtesy of Elsevier Publishing Company) There are numerous accounts of particular tests and of their interpretation. Some of these are described below. A typical set of tests has been described by Grewer (1974). These tests are illustrated in Figures 11.23...

9898

An illustrative example of the determination of mean and median ranks of a set of censored failure data is given in Table 7.23. The data in Table 7.23 can be fitted by a Weibull distribution and are plotted in Figure 7.27. The characteristic life and the shape parameter p are obtained in the same way as in the previous example. Further discussions of ranking are given by L.G. Johnson (1964) and Bompas-Smith (1973). 7.20.14 Fitting of failure distributions observation window These methods are...

Iso

Figure 8.7 Exotherm profile in test for thermal stability of individual substances using accelerating rate calorimetry (ARC) (N. Gibson, Rogers and Wright, 1987) (Courtesy of the Institution of Chemical Engineers) Figure 8.7 Exotherm profile in test for thermal stability of individual substances using accelerating rate calorimetry (ARC) (N. Gibson, Rogers and Wright, 1987) (Courtesy of the Institution of Chemical Engineers) in a temperature controlled environment and with a heating element on...

0024

Figure 7.29 Variation of failure rate with observation time, in terms of the Weibull distribution (Blanks, 1977) Figure 7.29 Variation of failure rate with observation time, in terms of the Weibull distribution (Blanks, 1977) The method is to equate the moments of the data and those of the distribution Mi tAexp(-At)dt m1 Jo A m and k are obtained from the implicit equations

11113

Where e is the base of the natural logarithm. The authors define the degree of conversion X in terms of the fraction of heat released. They state that experience shows that for a batch reaction the following equation may be used where c is the concentration, k is the reaction velocity constant, r is the rate of reaction, n is the effective order of the reaction and subscripts A, B and 0 indicate reactant A, reactant B and initial conditions, respectively. For a semi-batch reaction a relation...

113 Detailed Design

The detailed design stage involves the detailed process and mechanical design together with detailed design from a large number of supporting disciplines. The treatment given here is confined to a broad outline, with particular reference to features bearing on SLP. Elements of the detailed design of the process and plant include (7) pressure relief, venting and disposal (8) control and instrumentation (11) toxic emission protection (16) plant reliability, availability and maintainability (17)...

Regulations 21 41 and

This Schedule applies to storage of dangerous substances and or preparations at any place, installation, premises, building, or area of land, isolated or within an establishment, being a site used for the purpose of storage, except where that storage is associated with an installation covered by Schedule 4 and where the substances in question appear in Schedule 3. The quantities set out below in Parts I and II relate to each store or group of stores belonging to the same manufacturer where the...

92510

The total integrated dosage Aid ((kg m3) s) is wC2u ut)2-n Setting n 0.25 Atid 1 x175 For a continuous release the concentration at ground level on the centre line of the cloud is where Q is the mass rate of release (kg s) setting n 0.25 The total integrated dosage is obtained by assuming that the continuous release lasts for some finite time. Then the total integrated dosage is Models of heavy gas dispersion tend to be too complex to permit the analytical...

323 Enforcement Practice

Information on the enforcement activities of the HSE is given in the annual reports of the HSC and in the annual Health and Safety Statistics. The Health and Safety Statistics 1990-91 (HSE, 1992b) gives data for enforcement by all enforcement agencies. The number of notices issued in 1990-91 was as follows improvement notices 19079 deferred prohibition notices 467 immediate prohibition notices 6222 and total notices 25764. Data more specific to industry are given for the Field Operations...

7118

F (t) t V exp(-X2t) - exp(-X1t) 7.11.10 R(t) 1 X1 exp(-X2t) - X2 exp(-X1t) 7.11.11 Many practical problems cannot be solved by any of the analytical methods described and are only soluble by simulation. The principal method used is Monte Carlo simulation. Accounts of Monte Carlo simulation include that given in Method of Statistical Testing. Monte Carlo Method (Shreider, 1964)and that by Shooman (1968a). The application of Monte Carlo simulation to the determination of the reliability and or...

Computer system software reliability

NRC (Appendix 28 Computer Software Reliability) Aiken (1958) Anon. (1960b) Naur (1966) Floyd (1967) London (1968) Manna and Pnueli (1969) Bauer (1975a,b) Goos (1975) Poole (1975) Tsichritzis (1975a,b) Fagan (1976) Meyers (1976) Shooman (1976, 1983) J.L. Peterson (1977) R.B. Anderson (1979) Daniels (1979 NCSR 17, 1983, 1986, 1987) Boulton and Kittler (1979) Daniels and Hughes (1979 NCSR 16) Glass (1979) Kopetz (1979) Cho (1980) Longbottom (1980) Dempster et al. (1981) Kersken and Ehrenberger...

009 0065

LPG, liquefied petroleum gas URL, United Refineries Ltd LPG, liquefied petroleum gas URL, United Refineries Ltd event at unit k. In order to do this, it is necessary to make some assumption about the probability distribution of the failures induced at unit j. The assumption made is that the distribution between catastrophic failures and leak failures at unit j is the same as that at unit i. The relation for the probability of loss of containment at unit k due to an event at unit j, itself due...

Reactivity instability explosibility

ASTM (STP 394, 1975) Lothrop and Handrick (1949) Boynton et al. (1959) Steele and Duggan (1959) van Dolah (1961, 1965, 1969a) Mackenzie (1962, 1974) van Dolah et al. (1963) Shabica (1963) Wilcox and Bromley (1963) Cruise (1964) J.R Marshall (1964) G.T. Austin (1965b) Garn (1965) Snyder (1965, 1982) Platt and Ford (1966) Sykes (1966) Prugh (1967) Silver (1967) Wankel (1967) Adams (1968) Lindeijer (1968) Roburn (1968) Settles (1968) Woodworth (1968) Brinkley (1969) Lindeijer and Pasman (1969)...

72 Reliability Engineering in the Process Industries

7.2.1 Applicability of reliability techniques It is entirely right that the process industries should seek to apply the techniques and obtain the benefits of reliability engineering. But it is important to recognize that reliability engineering as a discipline has grown up outside these industries. It is to be expected, therefore, that the techniques will need to be adapted to and developed for the problems of the process industries. A similar adaptation was necessary in order to apply control...

105 Layout Planning and Development

Plant layout is usually divided into the following The layout developed typically goes through three stages The sequence of layout development described by Mecklenburgh is Table 10.3 Typical stages in the development of a plant layout (after Mecklenburgh, 1985) (Courtesy of the Institution of Chemical Engineers)

CCPS Hazard Evaluation Guidelines

Physical and chemical properties (see also Tables 11.1, 11.18, 16.1, 16.3, 17.3, 17.62, 18.1) NIOSH (Crit. Doc. series, 1990 18) Nuckolls (1929) Mellan (1950, 1957, 1977) MCA (1952-SD series, 1972 21) NSC (1952-, Safety Data Sheets) DECHEMA (1953-) Sax (1957-, 1981, 1986) Kirk and Othmer (1963-, 1978-, 1991-) Marsden (1963) Shabica (1963) D.T. Smith (1965a) Anon. (1966-67) CGA (1966 1) G.D. Muir (1971-) Bahme (1972) FPA (1972-, CFSD, Vol. 4, H series) Meidl (1972) Bretherick (1974, 1975, 1976,...

5

NMLs give different L values an average may be taken. C factor Low risk plants Medium risk plants Relatively high risk plants High risk plants C factor parameters 1. Nature of raw materials 1 Flammable solids, heavy flammable liquids or vapours, dusts at ambient conditions. 2 Light flammable liquids, hydrogen, combustible gases, naphtha, petrol 3 LPG-type materials or other flashing liquids (any substance which is flashing above 5 ) materials in use above their ignition point...

Quality assurance ISO 9000 BS 5750

ASME (Appendix 27) ISO (Appendix 27, 1987 ISO 9000) IMechE (1982 65, 66,1990 122) ASCE (1985 22) IChemE (1986 124, 1988 133) Allen and Nixon (1987 SRD R455) ICI (1987) Atkinson (1988a) Weismantel (1990) CIA (1991 CE6) Holmes (1991) Fouhy et al. (1992) Graham (1992) Hockman and Erdman (1993) P.L. Johnson (1993) Love (1993) Owen and Maidment (1993) Weightman (1993) Mancine (1994) BS (Appendix 27 Quality Assurance), BS Handbook 22 1990, BS PD series

Categories of substances and preparations not specifically named in Part I

The quantities of different substances and preparations of the same category are cumulative. Where there is more than one category specified in the same entry, the quantities of all substances and preparations of the specified categories in that entry shall be summed up. Categories of substances and preparations (Column l) For application For application of of Regulation 4 Regulations 7-12 (Column 2) (Column 3) 1 Substances and preparations that are classified as 'very toxic' 2 Substances and...

30 O

Storage facilities meet minimum fire regulations. Most containers carry name of contents. Approved equipment generally is used. Supply at work area is limited to one day requirement. Containers are kept in approved storage cabinets In addition to 'fair', ventilation rates are periodically measured, recorded and maintained at approved levels The majority of workmen instructed concerning skin-irritating materials. Workmen provided with approved personal protective equipment or devices. Use of...

200

Annual number of incidents 39 and annual loss US 572 million. At the level of the individual plant, some examples of claims for property damage and business interruption have been given by di Gesso (1989). He quotes the following three cases. Case 1, in 1981, involved a tower which failed during a hydrostatic test with water at a temperature of 10oC rather than at 20oC as recommended. In Case 2, also in 1981, a compressor trip led to a furnace tube burst and damage to a common stack. In Case 3,...

67 Management of Change

A large project is subject throughout its life cycle to numerous changes, or modifications. It is necessary to have systems which will manage these changes satisfactorily. In particular, it is necessary keep control of modifications, which may occur in design or operation. These modifications may be to the process or the plant. Either way, modification control systems are required to detect intent to modify, to refer the modification to the appropriate function for checking, to record a...

2570

Town and Country Planning (Use Classes) Order Collection and Disposal of Waste Regulations Control of Industrial Major Accident Hazards (Amendment) Regulations Control of Substances Hazardous to Health Regulations (COSHH) Docks Regulations Pressure Vessels (Verification) Regulations Town and Country Planning (Assessment of Environmental Effects) Regulations Town and Country Planning General Development Order Air Navigation Order Air Quality Standards Regulations Building Regulations (Amendment)...

781

The term (2A-2) is a normalizing factor such as to yield erf(ro) 1 7.7.57 The error function complement erfc is defined as erfc(x) 1 erf(x) 7.7.58 The integral which constitutes the error function is similar in form to those which occur in the distribution function of the normal distribution and in the log-normal distribution, as comparison of Equation 7.7.56 with Equations 7.7.18 and 7.7.21 shows. A similar integral also occurs in the definition of the probit Y described in Chapter 9....

755

P(X) P(Ai)+P(A2) +P(A3) - P(AA) 5 - P(A1A3) - P(A A3) + P A ) . . P(X) P(Ax) + (PA2) + + P(An) - P(A ) - P(A ) -----P(An-1An) + P (A1A2A3) + P (A1A2A4) + +P(An_2An_1An) (-1)n-1P(A A2 An) If the events are mutually exclusive, Equation 7.5.5 simplifies to The probability obtained in Equation 7.5.15 is a joint probability. Marginal probabilities may also be obtained from conditional probabilities 7.5.5 Independence and conditional independence If events A and B are independent

Notification of New Substances Regulations van

Den Heuvel (1982) HSE (1982 HS(R) 14) Fire, fire certificates Langdon-Thomas (1967a,b) Everton (1972) HSE (1976 6) BRE (1986 BR85) FPA (1989 CFSD FPDG 1, OR 1) Everton et al. (1990) Storage H. Watts (1951) J.R. Hughes (1970) Anon. (1985cc) M.R. Wright (1985) Classification, labelling and packaging, CLP Regulations HSE (1978 HS(R) 1, 1983 HS(R) 17, 1985 HS(R) 22) IBC (1983 44) Transport British Rail (n.d.) HSE (1979 HS(R) 3, 1987 HS(R) 24) Ministry of Transport (n.d.) British Transport Commn,...

9Hazard Assessment

9.7 Cause-Consequence Diagrams 9 34 9.10 Rare Events and External Threats 9 49 9.11 Human Factors and Human Error 9 52 9.16 Population Characteristics 9 60 9.17 Modification of Exposure 9 70 9.19 Presentation of Results 9 76 9.20 Confidence in Results 9 79 9.24 Simplified Assessment Methods 9 109 9.28 Risk Assessment Debate 9 125

37Health and Safety Executive

The HSE forms part of the Department of Employment. It is responsible for enforcing the relevant legislation and for providing advice. The HSWA 1974 provides a unifying framework for a number of different existing Acts. The HSE brings together a number of inspectorates and other functions which previously operated separately under these Acts. The structure of the Commission and Executive is shown in Figure 3.2. There has been a tendency for other inspectorates to be consolidated within the HSE....

826 Hazard Warning Analysis

In some cases analyses are carried out to identify the lesser events which may serve as precursors to, or warnings of, more serious events. Such analyses clearly rank as hazard identification techniques. One such method is the technique of hazard warning analysis described by Lees (1982b). Like fault tree analysis, on Table 8.52 Rapid ranking of hazards hazard categories (Gillett, 1985) (Courtesy of Process Engineering) Table 8.52 Rapid ranking of hazards hazard categories (Gillett, 1985)...

99 Expert Judgement

It is sometimes difficult to obtain data on event frequency or probability. In such cases a possible solution is the use of expert judgement to obtain the required estimates. This approach has been used particularly to obtain estimates for human error rates, where it is especially difficult to obtain field data, and for equipment failure rates, but it also has other applications. An account of expert judgement is given in Eliciting and Analysing Expert Judgement A Practical Guide (M.A. Meyer...

X

Subsection 7.20.11 as Sections 7.1-7.13 plus a parameter C logarithm of likelihood function ( ln L) Subsection 7.20.16 as Sections 7.1-7.13 plus v random number zi ith influencinbg variable, or explanatory factor zo(t) baseline hazard function coefficient of zi Ci parameter representing complexity of ith sub Wi weighting factor for ith subsystem Subsection 7.23.3 T period of test q proportion of defective items mo replacement time of component a standard deviation of life of component

Availability

Smith and Grace (1961) Gibbons (1962) Sandler (1963) R.E. Jackson et al. (1965) H.L. Gray and Lewis (1967) W.N. Smith (1968) Jenkins (1969) Sherry (196970) Buzacott (1970a, b) Dailey (1970) Henley (1971) Jenkins et al. (1971) Konoki (1971) Kuist and Fife (1971) Roth and Fiedler (1971) Cason (1972) McFatter (1972) G.H. Mitchell (1972) Ufford (1972) Gaddy and Culbertson (1973) Kardos and Vondran (1973) Locks (1973) Yaro (1973) Inone et al. (1974) Kafarov et al. (1974) Kardos et al. (1974-)...

7178

Which itself is the density function of a normal distribution with a mean value Then the probability that the load exceeds the strength is given by the distribution function where (x) is the density function of the load x, f(y) is the density function of the strength y ( x z), and z is the difference between x and y. A safety parameter may be defined as

48Process Hazard Control Background

The growth of major industrial hazards, especially those associated with the process industries, and the means of controlling them, was one of the main concerns of the Robens Committee. The subsequent Health and Safety at Work etc. Act 1974 (HSWA) provides a much improved framework for control. Even before this, a Department of the Environment (DoE) circular DoE 72 1 had drawn attention to the need to take account of major hazards in planning. The disaster at Flixborough in 1974 caused a great...

86

9.5.14 Illustrative example crystallizer system As a third illustrative example, consider the proposed crystallization plant system described by Lawley (1974b). This system is shown in Figure 9.8. The problem is to estimate the expected frequency of emission of the toxic and corrosive crystallizer slurry via the stack to the atmosphere. There are two ways in which this can occur - by discharge at about atmospheric pressure or by discharge at high pressure - but only the former is considered...

T

Figure 11.15 Course of a kinetic controlled exothermic reaction in a closed adiabatic system (a) heat generated vs time (b) temperature vs time (c) heat generated vs temperature Another limiting factor is the relation between the reaction rate and temperature. Some of the relations which occur are shown in Figure 11.17. Figure 11.17(a) gives the normal curve of rapid rise in reaction rate with temperature Figure 11.17(b) shows diffusion control with a slow increase of reaction rate with...

Transfer processes fluid flow heat transfer

NRC (Appendix 28 Fluid Flow, Thermal Hydraulic Analysis) Lander (1942) L.F. Moody (1944) D.Q. Kern (1950) Streeter (1951) Lewitt (1952) McAdams (1954) Treybal (1955) Crank (1956) G.C.F. Rogers and Mayhew (1957) Emmons (1958) Boucher and Alves (1959-) Landau and Lifshitz (1959) Bird, Stewart and Lightfoot (1960) Duncan, Thomas and Young (1960) C.O. Bennett and Myers (1962) Kutateladze and Borishanskii (1966) Streeter and Wylie (1967, 1975) Massey (1968) Holland et al. (1971) D.S. Miller (1971)...

94 Event Data

The methods which are used to estimate the frequency of an event depend somewhat on the nature of the event. The aim is usually to base the estimate on historical experience, but the precise way in which this is done varies. In the simplest case there may be available frequency data which apply directly to the event in question. This is likely to be the case if the event of interest is one which does sometimes occur, e.g. the failure of a single item of equipment such as a pipeline fracture or...

112 Conceptual Design

At the conceptual design stage the process concept is developed, its implications are explored, and potential problems are identified. Design in general and conceptual design in particular is generally regarded as an art. Much work is going on, however, to put it on a more systematic basis. Early work is described in Strategy of Process Design (Rudd and Watson, (1968). A systematic approach is stated in The Conceptual Design of Chemical Processes (J.M. Douglas, 1988). A further discussion of...

C

F (t), f2 (t) failure density function of subsystem Fd(d) downtime distribution function Fr(tr) repair time distribution function k counter state counter parameter in Weibull distribution (alternative form) parameters defined by Equations 7.13.28 and 7.18.5 m mean life parameter in Weibull distribution (alternative form) mean of normal distribution number of repair-men m* location parameter in log-normal distribution md mean downtime mi th moment (experimental) mr mean repair time mean of...

7Reliability Engineering 7l

7.1 Development of Reliability Engineering 7 5 7.2 Reliability Engineering in the Process Industries 7 6 7.3 Definition of Reliability 7 7 7.4 Meanings of Probability 7 7 7.5 Some Probability Relationships 7 8 7.6 Some Reliability Relationships 7 10 7.8 Reliability of Some Standard Systems 7 19 7.9 Reliability of Complex Systems 7 22 7.11 Joint Density Functions 7 29 7.12 Monte Carlo Simulation 7 30 7.17 Models of Failure Strength-Load Interaction 7 42 7.18 Models of Failure Some Other Models 7...

Regulation 8 Updating of reports under Regulation

(1) Where a manufacturer has made a report in accordance with Regulation 7(1), he shall not make any modification to the industrial activity to which that report relates which could materially affect the particulars in that report, unless he has made a further report to take account of those changes and has sent a copy of that report to the Executive at least 3 months before making those changes or before such shorter time as the Executive may agree in writing. (2) Where a manufacturer has made...

2000

Cellulose nitrate (containing > 12.6 nitrogen) Lead azide Mercury fulminate Nitroglycerine Pentaerythritol tetranitrate Picric acid Flammable substances as defined in Schedule 1, paragraph c(i) Flammable substances as defined in Schedule 1, paragraph c(ii) Flammable substances as 200 defined in Schedule 1, paragraph c(iii)

7745

These quantities are shown in Figure 7.5(a). The distribution is also called the 'uniform distribution' and has two parameters, the location parameter a and the scaling factor b. The rectangular distribution is used in reliability work mainly to give random variables a uniform distribution across a specified interval. It is used, for example, in Monte Carlo simulation work. Computers generate pseudo-random numbers with a rectangular distribution. Random numbers...

Enclosed or indoor process units

For enclosed area, defined as any roofed area with three or more sides or an area enclosed by roofless structure with walls on all sides (1) Dust filters or collectors in an enclosed area, penalty 0.50 (2) Process in which flammable liquids are handled above flash point in an enclosed area, penalty 0.30 for quantities in excess of 10 M lb, penalty 0.45 (3) Process in which LPG or flammable liquids are handled above boiling point in an enclosed area, penalty 0.60 for quantities in excess of 10...

7715

For the range 0 < t These quantities are shown in Figure 7.4(a). The distribution is characterized by a single parameter, the hazard rate A. The exponential distribution, which has a constant hazard rate, is the distribution usually applied to data in the absence of other information and is the most widely used in reliability work. For the normal distribution the characteristics are

92025

Hence the output coefficient of variation has the range (1 21) min Equation 9.20.18 shows that for an OR gate the coefficient of variation of the output is less than the larger of the coefficients of variation of the inputs, and that in this sense this gate reduces the uncertainty in the input data. For an AND gate with an input of frequency A1 and another input of probability p2, with corresponding standard deviations a1 and s2, respectively, the frequency A of the output is The corresponding...

914Domino Effects

Another aspect which needs to be taken into account in hazard assessment is the possibility of knock-on or 'domino' effects, leading to escalation. An event at one unit may be the cause of a further event at another unit, and so on. The possibility of domino effects was one of the principal concerns which led to the hazard assessment of the complex of installations at Canvey, which was the subject of the two Canvey Reports described in Appendix 7. The incident at Feyzin in 1968 (Case History...

269

It is usually a straightforward matter to obtain the necessary information on the on-site workforce, but this is not so for the off-site population. Here the data required include (3) population changes by time of day For estimation of societal risk information is required on all these items, but even for estimation of individual risk only it is still necessary to have data on the vulnerable population and the population outdoors. The accuracy of estimation of the density and other...

815 Failure Modes Effects and Criticality Analysis

Another major method of hazard identification is failure modes and effects analysis and its extension failure modes, effects and criticality analysis. Accounts of failure modes and effects analysis (FMEA) and of failure modes, effects and criticality analysis (FMECA) are given in the CCPS Hazard Evaluation Guidelines (1985 1, 1992 9) and by Recht (1966b), J.R. Taylor (1973, 1974c, 1975a), Himmelblau (1978), Lambert (1978a), A.E. Green (1983), Flothmann and Mjaavatten (1986), Moubray (1991),...

Industrial law employers liability

Legge (1934) Allsop (1962) Fife and Machin (1963, 1972, 1982, 1990) Whincup (1963, 1968, 1976) Wedderburn (1965) Ison (1967) F.A. Robinson and Amies (1967) Samuels (1967, 1969) Winn (1968) Gold (1969) Atiyah (1970) Law Commission (1970a,b) Society of Labour Lawyers (1970) Broadhurst (1971) Munkman (1971) Robens (1972) Kinnersly (1973) D. Farmer (1977) Locke (1977) W.E. Cox and Walker (1978) J. Jackson (1979) Anon. (1980 LPB 35, p. 13) Staunton (1982) Farquhar (1986) Anon. (1987d) A.L. Jones...

7111

Where 1(r) is a density function, and , 2(t r) is a conditional density function. The associated marginal density function fit) is It may be noted that there is a difference between the number of repair men that can usefully be used, depending on whether it is reliability or availability which is being calculated. In a system with n equipments which fails when n - r + 1 equipments have failed, it is possible to use n repair men to maximize availability, but only n - r repair men to maximize...

51 Economics of Loss Prevention

Loss prevention is concerned with the avoidance both of personal injury and of economic loss. In both spheres there is an economic balance to be struck. But there are also quite difficult problems in making the economic assessments. There is no doubt, however, about the economic importance of loss prevention. Some costs arise through failure to take proper loss prevention measures others are incurred through uninformed and unnecessarily expensive measures. Both types of cost are numerous and...

Total quality management

Ishikawa Kaoru (1976, 1985) Crosby (1979, 1984, 1986) Juran (1979) Taguchi (1979, 1981) Juran and Gryna (1980) Deming (1982, 1986) Laporte (1982) Feigenbaum (1983) F. Price (1985) DTI (1986) Singo Shigeo (1986) Shell Int. (1988) Baguley (1989, 1990) Bond (1989b, 1990a) Hodge and Whiston (1989) Oakland (1989, 1993) K.H. Roberts (1989) K.H. Roberts and Gargano (1989) Taguchi, Elsayed and Hsiang (1989) Whiston and Eddershaw (1989) Cairns and Garrett (1990) Institute of Materials (1990 PR 1004)...

72020

Fmed + 0.4 It is usual to calculate the failure distribution F(f) as either the mean rank Fmean or the median rank Fmed. The use of ranking is essential when the number of items under test is small. The mean is commonly taken as a representative descriptor of a sample from a distribution. In the case of a highly skewed distribution, however, a better description may be the median. A discussion of the difference between mean rank and median rank and of the choice between them is given by Kapur...

620 Control

The purpose of the measurement activities just described is to provide the basis for control actions to correct deviations and for review of the performance of the system. If a deviation is detected, control action is required to correct it. In some cases the corrective action needed is obvious and it can then be taken. In other cases it is necessary to carry out an investigation. The purpose of the investigation is to discover not a single 'cause' but the total situation which has given rise...

15 Large Single Stream Plants

For some decades up to about 1980, there was a strong trend for the size of plants to increase. The problems associated with large, single-stream plants are a major reason for the development of loss prevention. These problems are now considered in more detail in order to illustrate some of the factors underlying its growth. Selected references on large, single-stream plants are given in Table 1.3. The increase in the size of plant in the period in question for two principal chemicals is shown...

Hazard assessment applications and case histories

McGillivray (1963a,b) Kletz (1971, 1972a) Ybarrando, Solbrig and Isbin (1972 AIChE 119) Siccama (1973) Lawley (1974a,b, 1976, 1980) N.C. Rasmussen (1974) AEC (1975) Dicken (1975) J.R. Campbell and Gaddy (1976) Moser, Moel and Heckard (1976) Sellers (1976, 1988) Shell UK Exploration and Production (1976) Lundquist and Laufke (1977) D.S. Nielsen (1977) D.S. Nielsen, Platz and Kongso (1977) Okrent (1977) Rasbash (1977b) HSE (1978b,d, 1981a) Blokker et al. (1980) Joschek et al. (1980) de Ruiter and...

111034

Where mos is the measured self-heat rate at temperature T0 and 0ms is the measured time to maximum rate at temperature T. Further methods of characterizing the reaction and the reactor have been described by Grewer et al. (1989). It is common for a product to be brought to an elevated temperature for some 24 hours. The adiabatic storage test may be used to determine a suitable temperature limit. The most useful information yielded by the adiabatic storage test is the adiabatic induction time...

73 Definition of Reliability

Definitions of reliability are given in various British Standards dealing with terminology on quality and reliability. BS 4778 Part 1 1987 defines reliability as The ability of an item to perform a required function under stated conditions for a stated period of time. This supersedes the definition given in BS 4200 Part 2 1974 which has now been withdrawn. Further, essentially similar, definitions are given in BS 4778 Section 3.1 1991 and BS 4778 Section 3.2 1991. BS 4778 Section 3.1 1991...

Diers

Nolan and Barton (1987) Cronin, Nolan and Barton (1987) Pantony, Scilly and Barton (1989) ABPI (1981, 1989) Pilz (1986) Fierz et al. (1983, 1984) Regenass (1984) Kohlbrand (1985, 1987a, 1990) Gordon et al. (1982) O'Brien et al. (1982) Hoffmann (1985) Berkey and Workman (1987) Kunzi (1980) Barton and Rogers (1993) McNab (1981) N. Gibson (1984, 1986b) Christen (1980) Snyder (1965, 1982) Brannegan (1985) Duval (1985) Coates and Riddell (1981a,b) Lambert and Amery (1989) Figure 11.20 Phase plane...

L

Figure 7.19 Some patterns of equipment hazard rate, including and additional to the bathtub curve (Moubray, 1991) (Courtesy of Butterworth-Heinemann) exhibit a constant hazard rate. This is readily understood intuitively in terms of a smoothing out of particular hazard rate characteristics of individual components. The phenomenon is well illustrated by the so-called lightbulb curve' shown in Figure 7.20. If a building is provided with new lightbulbs throughout, then assuming that these fail by...

8211 Hazcheck

The first element in the system is HAZCHECK, which might be described as an enhanced checklist system. Accounts of HAZCHECK are given by Reeves, Wells and Linkens (1989), Hurst and Reeves (1990), Wells, Phang and Reeves (1991) and Wells (1992 LPB 105). HAZCHECK is essentially a structured checklist, or set of checklists. The main headings are given in the master list and are expanded in subordinate lists. The master list has been given in a number of versions. It is convenient here to give that...

Fire fighting

Hand fire extinguishers Water supply Installed sprinkler, water spray or monitor systems a This table lists the offsetting factors but not the associated numerical values. See Technical Manual (D.J. Lewis, 1979) for full details. application. It has been described by D.J. Lewis (1980, 1984) and is considered in more detail in Chapter 10. The instantaneous fractional annual loss (IFAL) index is a separate index developed by the Insurance Technical Bureau primarily for insurance assessment...

Hazard Accident and Loss

2.2 Standard Industrial Classification 2 4 A rational approach to loss prevention must be based on an understanding of the nature of accidents and of the types of loss which actually occur. Therefore, in this chapter, first the nature of the accident process is considered and then the accident and loss statistics are reviewed to give an indication of the problem. Selected references on accident and loss experience are given in Table 2.1. In addition, many other tables of data are given in other...

21 The Accident Process

There are certain themes which recur in the investigation of accidents and which reveal much about the accident process. First, although in some reporting schemes the investigator is required to determine the cause of the accident, it frequently appears meaningless to assign a single cause as the accident has arisen from a particular combination of circumstances. Second, it is often found that the accident has been preceded by other incidents which have been 'near misses'. These are cases where...

83Checklists

One of the most useful tools of hazard identification is the checklist. Like a standard or a code of practice, a checklist is a means of passing on hard-won experience. It is impossible to envisage high standards in hazard control unless this experience is effectively utilized. The checklist is one of the main tools available to assist in this. Checklists are applicable to management systems in general and to a project throughout all its stages. Obviously the checklist must be appropriate to...

12 Industrial Safety and Loss Trends

About 1970 it became increasingly recognized that there was a world-wide trend for losses due to accidents to rise more rapidly than gross national product (GNP). This may be illustrated by the situation in the UK. The first half of this century saw a falling trend in personal accidents in British factories, but about 1960 this fall bottomed out. Over the next decade very little progress was made in fact there was some regression. Figure 1.1 shows the number of fatal accidents and the total...

10 100 1000

Figure 9.37 Groningen FN curve risk criterion (Hagon, 1984 after PWG, 1979) (Courtesy of the Institution of Chemical Engineers) F3300 2.2 x 10-5 This indicates that large accidents are a significant contributor to the average risk of death. and Hagon takes as a prime determinant of the risk d30_2 4 x 10-3 aversion index a the parameter F , which is the 1 . frequency of all accidents between one with N fatalities D3 oo 52 x 10-3 and one with the maximum number M of fatalities. Figure 9.23 shows...

817 Computer Hazop

The hazop study method just described was developed for plants in which predominantly the control system was based on analogue controllers. The advent of computer control has created the need for some method which addresses the specific problems of this form of control. There are now the beginnings of a methodology for the hazop-like study of computer controlled process systems, or computer hazop (CHAZOP). This is quite distinct from computer-aiding of hazop. Accounts of approaches to computer...

31 Factory legislation

The current regulatory regime and legislation for health and safety in the UK is best understood as a continuation of developments which have taken place over the best part of 200 years. Accounts of the development of factories legislation are given in Safety and Health at Work (Robens, 1972) (the Robens Report), the HM Chief Inspector of Factories Annual Report 1974 (HMCIF, 1974), Her Majesty's Inspectors of Factories 1833-1983 (HSE, 1983d) and The Work of a Factory Inspector (HSE, 1987b)....

115 Licensors Vendors and Contractors

The description of the design process which has just been given is based essentially on that which takes place in an operating company which is designing its own plant. The ultimate responsibility for the safe design and operation of plant lies with the operating company. It should take the appropriate steps to ensure that the processes designed and the equipment supplied by the other parties are safe. Some design responsibility resides, however, with other parties. These may be (1) licensor,...

Stage Two plot layout

Typical stages in the development of a plant layout are given in Table 10.3. The Stage Two and Final Stage design network is shown in simplified form in Figure 10.1. The process of layout development makes considerable use of guidelines for separation distances. These are described in Section 10.11. Stage One is the preliminary layout, also known as the conceptual, definition, proposal or front end layout. In this stage consideration is given to the various factors which are important in the...

33 Factory Inspectorate

The original inspectorate is the Factory Inspectorate set up under the 1833 Act. This is the largest Inspectorate and that which is of principal interest in the present context. References to accounts of the work of a Factory Inspector were given in Section 3.1. The development of the Factory Inspectorate and its current work are described in the HM Chief Inspector of Factories Annual Report 1974 (HMCIF, 1974), Her Majesty's Inspectors of Factories 1833-1983 (HSE, 1983d) and The Work of a...

3d

Te effective time duration(s) Ie effective radiation intensity (W m2) t time duration of pool burning (s) I radiation intensity from pool burning (W m2) p peak overpressure (N m2) J impulse (Ns m2) C concentration (ppm) T time interval (min) Table 9.19 CPD models principal contents of Yellow Book and Green Book (Committee for the Prevention of Disasters, 1992a,b) Turbulent free jet Spray release Evaporation Heat radiation Dispersion Vapour cloud explosion Consequences of rupture of vessels...

Spacings between two storage units spacing between one tank wall and the other tank wall m

Overall risk Overall risk rating R2 of storage unit B storage unit high Extreme Very extreme Very Very units with frothover or boilover potential and with units located on sloping ground. Another approach to the determination of separation distances is to use hazard models to determine the separation distance at which the concentration from a vapour escape or the thermal radiation from a fire fall to an acceptable level. This is the other side of the coin to hazard assessment of a proposed...

2

And 45 fatalities, giving 0.23 deaths incident. This higher figure reflects the fact that these incidents are biased towards those involving fatalities. Marrs and Lees express the risk in terms of reactor operating time Frequency of fatality 6.7 x 10-5 deaths reactor-year The fatal accident rate (FAR) from this cause alone was estimated as 3.8. The chemical and petrochemical industries are described in The Petroleum Chemicals Industry (Goldstein and Waddams, 1967), The Petrochemical Industry...

Acknowledgements

For permission to reproduce material in this book the author would like to acknowledge in particular Academic Press Adam Hilger Addison-Wesley Publishers AGEMA Infrared Systems the Air Pollution Control Association the American Chemical Society the American Gas Association the American Institute of Chemical Engineers the American Petroleum Institute the American Society of Heating, Refrigerating and Air Conditioning Engineers the American Society of Mechanical Engineers Anderson Greenwood...

212 Case Histories

The generalized statistics may be supplemented by individual case histories. These are treated in Appendix 1, which describes the various sources and gives specific case histories. The sources include the accident reports of the Health and Safety Executive (HSE) and the National Transportation Safety Board (NTSB) the collections of the Manufacturing Chemists Association (MCA) and the American Petroleum Institute (API) the periodic reviews of insurers, such as the 100 Large Losses (Marsh and...

Preface to Second Edition

The first edition of this book appeared in 1980, at the end of a decade of rapid growth and development in loss prevention. After another decade and a half the subject is more mature, although development continues apace. In preparing this second edition it has been even more difficult than before to decide what to put in and what to leave out. The importance of loss prevention has been underlined by a number of disasters. Those at San Carlos, Mexico City, Bhopal and Pasadena are perhaps the...

Pb And Pc Fb And Fc Fb And Pc

Pa Pb+Pc-PbPc pb+pc Not permitted Pa PbPc where p is the probability of failure of the protective action, A is the frequency of the input event, and A0 is the frequency of the output event. For a Type 2 situation, Equation 9.5.3 is again applicable, with the probability p of failure of protective action in this case being obtained from Equation 9.5.2. The evaluation of a Type 3 situation is less straightforward. For this, use may be made of the appropriate parallel system model derived from...

1023 Modular Plants

During the late 1950s and early 1960s there was introduced a type of plant consisting of a number of modules and mounted on skids which could be transported by road from the fabrication to the operating site. The processes were straightforward and the plants were simple and cheap. From these early skid-mounted plants there has developed a whole range of modular and barge-mounted plants, some of which are large and complex. Accounts of modular plants have been given by Glaser, Kramer and Causey...

314 Flammables Explosives and Fire Legislation

3.14.1 The Petroleum (Consolidation) Act 1928 The Petroleum (Consolidation) Act 1928 (P(C)A) was brought in to deal with the storage and transport of 'petroleum spirit'. Petroleum spirit is defined as petroleum with a flash point equal to or less than 22.8oC (73oF). Although old, the Act is still in force. The P(C)A deals mainly with the licensing of premises for the storage of petroleum spirit. By its extension to other flammable substances it represented for a long time the principal means of...