failure frequency guidance

DNV SERVING THE process industry quantified risk assessment

failure frequency guidance Process equipment leak frequency data for use in QRA

02 I Process industry I quantified risk assessment I

GLOBAL SERVICES TO THE MARITIME, OIL & GAS AND ENERGY INDUSTRIES Asset risk management Enterprise risk management ■■ Offshore classification ■■ Safety, health and environmental risk management

Ship classification Technology qualification ■■ Verification

■■

■■

■■

■■

I quantified risk assessment I Process industry I 03

CONTENTS 04 1. Introduction 05 2. Background on HCRD 06 3. Application of Data 08 4. Methodology

12 5. Calculating Release Rate 16 6. Leak Frequency Datasheets 38 7. References

The world must lean on new sources to create a sustainable energy system. Wind energy is one of the most important. But technical, environmental, and financial barriers must be removed to release its full-blown potential. With over 25 years experience in wind energy, DNV understands the industry and what is required for success throughout the entire lifecycle of wind energy projects. Through our services and competent people, we can help get more out of wind energy projects.

The power to generate results


1. Introduction Frequency estimates are recognised as one of the largest sources of uncertainty in Quantitative Risk Assessment studies. There are a few sources of data for failure frequency data for process equipment loss of containment: Netherlands and Belgium have issued two different onshore frequency datasets for use in Seveso Directive risk assessments, and some companies and consultants have their own data. In many cases the provenance of these data is uncertain and examples exist of frequencies that are too low and do not match historical accident frequencies. It is detrimental to QRA methodology that such old or inconsistent data is routinely used. DNV is therefore publishing this booklet in order to put best practice process equipment leak frequency data into the public domain. DNV’s data is derived from the Hydrocarbon Release Database (HCRD) which has been compiled by the UK Health and Safety Executive (HSE) over a 20 year period. The database [2] contains details of over 4000 leak events at oil and gas installations in the UK Continental Shelf. It identifies 78 different types and size categories of process equipment, and records the quantity of the release and the release hole size. This is considered the most extensive dataset of its type and superior to current published datasets which often have much smaller and older data which do not reflect current integrity management programs. DNV has assessed this data for several years on behalf of a major oper-

ator. While the data is ‘noisy’, typical for real data, DNV has applied a smoothing function to cover all leak sizes. Leaks are differentiated for 17 equipment types. This analysis and interpretation is complex: considerable effort is required to obtain generic leak frequency data that are suitable for use in Quantified Risk Assessment (QRA). DNV believes that process industry will benefit from consistent application of these generic leak frequencies in QRA, and is therefore publishing its preferred dataset in order to encourage standardisation across different users. The booklet describes DNV’s methods for interpreting the HCRD and describes its application to offshore, onshore and LNG plant. The booklet also presents tabulations of generic leak frequency data for different equipment types. The booklet compares the DNV dataset against some alternative sources of leak frequency data and describes some important reasons why the DNV interpretation of the HCRD should be preferred. DNV was commissioned by Statoil to define the model presented in this document, involving contractors Scandpower and Safetec in the work.


2. BACKGROUND ON HCRD The Piper Alpha incident occurred in 1988 and resulted in 167 people losing their lives. As a result of this disaster Lord Cullen conducted an inquiry [1]. The inquiry made 106 recommendations including the requirement to report leaks to the HSE through the reporting of injuries, diseases and dangerous occurrences regulations (RIDDOR). The HSE organises this data and makes it publically available through the Hydrocarbon Release Database (HCRD). The database started to be compiled in October 1992 and now contains 20 years of experience in hydrocarbon releases from the UK Continental Shelf. Figure 1 shows the number of recorded leaks per year since 1992. (For each bar in the graph, the reporting period for leaks is between April 1st and March 31st in the following year. The first bar contains data for a six month period 1st of October 1992 and the 31st of March 1993). Figure 2 shows the breakdown of these leaks into three categories as defined by HSE: minor, significant and major. The average number of minor leaks per year between April 1st 1993 and March 31st 1998 is 82, whereas there were an average of 109 minor leaks per year between April 1st 2006 and March 31st 2010. This increase may be due to an actual increase in minor leaks or an improvement in the reporting of these leaks. Detection of smaller leaks may also have improved. The number of significant and major leaks has decreased over the period, particularly major leaks which have decreased by a factor of 12. In 2010 in the UK the oil

and industry committed to reducing its number of hydrocarbon emissions by 50% in 3 years. Two years into the three year programme there has been a 40% reduction in the number of leaks [1]. Determining the number of leaks that have occurred offshore provides only one part of the data that is required to calculate leak frequency. The number of different types of equipment offshore has also been recorded and quantified since 1992, although HSE has recorded no change in the equipment count since 2003 (Regarding system and equipment population data, HSE notes that the responsibility for maintaining the currency of this data rests with duty holders. The population data in HCRD is provided by duty holders on a voluntary database and it is not HSE’s role to update, or verify this particular data. Use of this population data would need to be made with caution). It is questionable that the amount of equipment has remained the same offshore since 2003. Maintaining an accurate equipment count is not straightforward, for example the count of equipment on mobile rigs would require the database operators to keep track of the position of MOUs and their movements. The equipment count on the UKCS is provided by the operators on a voluntary database and it is not part of HSE’s role to monitor or verify the equipment count. Therefore there are uncertainties associated with the equipment count.

250 350

200

300

No. of Leaks

250

150

200

100

150 100

50

50

0

0 1993

1995

1994

1997

1996

1999

1998

2001

2000

2003

2002

Year

Figure 1 Number of Leaks per year

2005

2004

2007

2006

2009

2008

2010

1993

1995

1994

1997

1996

1999

1998 Minor

2001

2000

2003

2002 Significant

2005

2004

Major

Figure 2 Number of leaks per leak category per year

2007

2006

2009

2008

2010


3. APPLICATION OF DATA 3.1 Offshore The data in this booklet is based upon data from the UK sector of the North Sea which has been collected by the HSE in the hydrocarbon release database (HCRD). The data is based on approximately 4000 recorded leaks recorded between October 1992 and March 2010. This database has been analysed by DNV to produce generic leak frequencies applicable for use with offshore installations that are operated to UK North Sea standards. The methodology of DNVs approach is presented in the 2009 HAZARDS XXIΙ Conference [4].

3.2 Onshore DNV normally also use these data for QRA at onshore facilities. In general, the HSE data set gives higher leak frequencies than most of the onshore sources of data. There are several possible explanations for this. Process equipment on offshore installations might experience higher leak frequencies than on onshore plants. Possible reasons might be extra external corrosion from salt-water spray, internal erosion from entrained sand, or impacts resulting from the more compact equipment layouts. However, offshore installations have safety management systems that would be expected to counter such evident hazards. The HSE data set on leak causes shows that corrosion/erosion is a minor contributor, with operational/ procedural faults and mechanical defects being the primary causes. Table 1 indicates the causes of leaks offshore between the 1st October 1992 to the 31st of March 2002.

Another possible explanation could be differences in data quality. The HSE offshore data set is a high-quality database, collected recently, covering a large population, with welldefined hole sizes, comprehensive equipment counts, and open for scrutiny by the operators and their consultants. Most of the available onshore leak frequencies come from small sample sizes. In fact in the case of onshore pipe leak frequencies it is concluded that the most widely accepted data set is of eight leaks in U.S. nuclear plants in 1972, or earlier collections whose size and origin are now unknown. [3]. A third factor affecting the comparison is that the HSE offshore data set includes some leaks that occurred while the equipment was depressurised, and others that were quickly isolated. The onshore frequencies are applicable to holes with process fluid at the full operating pressure. The frequencies based on HCRD data should be used with outflow models that take account of the variation in operating circumstances at the time of the leak. A further complicating factor is that onshore and offshore management systems in the UK must address different regulatory requirements. The Offshore Safety Case requirements are more onerous than those required of onshore refineries (e.g. offshore requirements for identification of safety critical elements, performance standards and written schemes, plus the rigorous leak reporting requirements). Overall, it is considered that the HSE offshore data provides the best available estimate of leak frequencies for both

Category

Causation Factor

Instances

Category Totals

Design fault

–

321

321

Equipment Fault

Corrosion/Erosion

277

Mechanical Defect

920

Material Defect

76

Operational Fault

Procedural Fault

Other

89

Incorrectly fitted

267

Improper Operation

495

Dropped/Impact

36

Left Open/Opened

237

Other

81

Noncompliance

231

Deficient Procedure

323

Other

34

Table 1: Causation factors in HSE offshore data [2]

1362

1116

588


onshore and offshore process equipment. However, it does require that the outflow model should take account of the possibility of the equipment being depressurised or quickly isolated at the time of the leak.

3.3 LNG Facilities The main risk drivers on an LNG site are events that are unlikely to be within the direct experience of individual plants and terminals. Establishing the frequency of such events is difficult, precisely because of their rarity. It requires systematic data collection, for leaks and the exposed equipment population, over many plants for many years. Such data collection is time-consuming and hence unusual. Alternative methods such as fault tree analysis are possible for plant-specific applications, but have not yet delivered generic leak frequencies suitable for routine use in QRA studies. Data sets exist that attempt to provide failure rates for cryogenic pipework, or for LNG-specific operating experience in general, but these are not considered to be sufficiently robust to justify any modification to the generic data derived from the HCRD. That is, any argument that offshore data such as the HCRD is not relevant to LNG facilities is considered to be compensated for by the weight of statistical data supporting the derived failure rates for specific equipment items, compared to the very limited data supporting any specific cryogenic / LNG failures that have occurred. Given the perceived risks associated with LNG it is often the case that fully welded pipelines and connections are employed, at least for the cryogenic part of the facility. Hence, where QRA of a ‘typical’ facility would assume that all valves are flanged (even where not shown on the P&IDs) this may not be the case for LNG facilities. It is important to confirm the extent to which this applies for any given facility – the default should be to assume flanged connections. A common aspect of uncertainty in QRA is associated with the frequency of inter-unit pipework / pipeline releases. It is widely accepted that the application of process pipework failure data will tend to give overly conservative values with respect to longer inter-unit pipe segments. This can be of particular relevance to LNG facilities, where the loading lines are often several kilometers long. In the course of conducting a large number of QRA studies, DNV has had the opportunity to draw on the experience of a range of operators. On the basis of these discussions, it is considered appropriate to apply a factor of 10 reductions in the pipework failure frequency for inter-unit piping. It should be

recognised that this is an engineering judgement assumption, based on acknowledging operational experience that inter-unit pipework fails very rarely (in comparison to the process pipework within the main process areas). This revised basis can be of particular relevance to loading lines, although should not substitute for consideration of all potential loads (and hence potential frequency modification factors) that may apply to a particular facility, or particular loading line. In summary: ■■

The DNV analysis of HCRD is recommended as the basis for the process and pipework failure data – as per all QRA studies.

■■

There is no statistically sound basis for modifying the source failure data to account for cryogenic or LNGspecific application.

■■

It is considered justifiable – albeit by engineering judgement – to reduce the process pipework failure rates by a factor of 10 for inter-unit piping.

■■

It should not be assumed that valves are flanged but this is an area where LNG applications may have the opportunity to reduce the parts count and hence the calculated leak frequency.


4. METHODOLOGY 4.1 General This booklet provides hole size frequency data for use in Quantified Risks Assessment (QRA) of process facilities. The methodology shown in this document was developed as in conjunction with Statoil and was presented in the 2009 HAZARDS XXI conference. The frequency data highlighted in this document uses the same methodology but uses data up to 2010 from the database. The booklet contains generic leak frequencies for each of the following process equipment types: 1. Compressors • Centrifugal compressors • Reciprocating compressors 2. Filters 3. Flanges 4. Heat Exchangers (Including Coolers, Heaters and condensers) • Air Cooled Heat Exchangers • Plate heat exchangers • Shell side heat exchangers • Tube side heat exchangers

DNV equipment type

HCRD equipment types

Steel pipes

Piping, steel (3 sizes)

Flanged joints

Flanges (3 sizes)

Manual valves

Valve, manual (10 types & sizes)

Actuated valves

Valve, actuated (18 types & sizes)

Instruments

Instruments (including connecting tubing, valves and flanges)

Process vessels

Pressure vessel (14 types)

Atmospheric vessels

Vessels at atmospheric pressure

Centrifugal pumps

Pumps, centrifugal (2 seal types)

Reciprocating pumps

Pumps, reciprocating (2 seal types)

Centrifugal compressors

Compressors, centrifugal

Reciprocating compressors

Compressors, reciprocating

Shell side heat exchangers

Heat exchangers, HC in shell

Tube side heat exchangers

Heat exchangers, HC in tube

Plate heat exchangers

Heat exchangers, plate

Air cooled heat exchangers

Fin fan coolers

Filters

Filters

Pig traps

Pig launchers & pig receivers (4 sizes)

Table 2: Equipment Type Groups

5. Pig traps 6. Process Pipes 7. Pumps • Centrifugal pumps • Reciprocating pumps 8. Instruments 9. Valves • Actuated valves • Manual valves 10. Pressurized process vessels 11. Atmospheric storage tanks This analysis represents the leak size distribution by an analytical frequency function, which ensures non-zero leak frequencies for all holes size ranges between 1 mm and the diameter of the inlet pipe to the equipment. In the following paragraphs a short presentation of the analysis is given. The methodology for obtaining leak frequencies from HCRD consists of three main steps: ■■

Grouping data for different types and sizes of equipment, where there is insufficient experience to show significant differences between them.

■■

Fitting analytical leak frequency functions to the data, in order to obtain a smooth variation of leak frequency with equipment and hole size.

■■

Splitting the leak frequencies into different leak scenarios, in order to promote compatibility with different approaches to outflow modelling in the QRA.

Grouping data The DNV analysis covers 17 different types of process equipment. Wellhead equipment, drilling equipment, pipelines and risers are all excluded from the analysis, since other more extensive data sources are available for these types of equipment. The remaining types of equipment are characterized as “process equipment”. The HCRD and the Statistics Report [2] allow 78 separate types and sizes of process equipment to be distinguished. In some cases, there is relatively little leak experience, and differences in leak frequencies between certain types and sizes of process equipment have no statistical significance. Analysis of results where there are only few reported results may be misleading. To avoid this, it is desirable to combine equipment types and sizes with relatively little leak experience.


Most HCRD equipment types have been used as defined by HSE, but some, with relatively little leak experience, have been combined into the following groups: ■■ All types of manual valves (bleed, block, check and choke). ■■ All other types of non-pipeline actuated valves (block, blowdown, choke, control, ESDV and relief, but not pipeline ESDV and SSIV). ■■ All types of pressure vessel (horizontal/vertical adsorber, KO drum, other, reboiler, scrubber, stabiliser, separator and stabiliser).

F(D) = f(D) Dm + Frup

Eqn 3

For pipes, flanges, valves and pig traps, HCRD provides data for different equipment size groups. Analysis of these showed significant variations in leak frequency with equipment size for pipes, flanges valves, whereas the population was too small to show any significant variation of leak frequency with equipment size for pig traps. Size dependence is represented in the leak frequency function using the following general form: f(D) = C(1 + aDn)

Eqn 4

Leak frequency function HCRD data, being real data, is very noisy as can be seen in Figure 3. DNV overlays a realistic distribution function that fits the data to obtain average leak frequencies. In raw data there are gaps (e.g. a particular equipment dimension may have zero leak events, but we would not predict zero for its actual leak likelihood) forcing a distribution of results in realistic predictions.

where: C, a, n = constants for each equipment type

A feature of the distribution function is that it allows any hole size distribution to be employed without biasing the result. In early coarse risk assessments we may only choose two hole sizes – small and large, whereas in detailed studies we may choose to employ 5 hole sizes for greater resolution. The equations allow any number of hole sizes to be selected and the total release frequency will always stay the same. The actual hole size feeds the consequence modeling and the more sizes used the greater the computational effort.

It is important to be aware that the leak frequency form is imposed on the data and that this is a mathematical representation of historical data. The data itself does not directly support a separate frequency for ruptures. The historical data related to releases from large hole sizes is very limited and the uncertainty related to estimation of such leaks is therefore considerable The additional rupture frequency Frup and the slope parameter m are assumed to be constants, i.e. not to be dependent on equipment size, for any equipment type.

The analysis represents the variation of leak frequency with equipment and hole size by the following general leak frequency function:

The function is used to calculate separate hole size frequencies for three types of leak scenario: ■■ Total leak frequency ■■ Full pressure leak frequency ■■ Zero pressure leak frequency

F(d) = f(D)dm + Frup

for d = 1 mm to D

Eqn 1

where: F(d) = frequency (per year) of holes exceeding size d f(D) = function representing the variation of leak frequency with D D = equipment diameter (mm) d = hole diameter (mm) m = slope parameter = additional rupture frequency (per year) Frup Hence the frequency of holes within any range d1 to d2 is: F(d1) – F(d2) = f(D)(d1m – d2m)

for d = 1 mm to D

Eqn 2

The frequency of full-bore ruptures, i.e. holes with diameter D, is:

The HCRD provides sufficient data to determine estimates for the a and n parameters for f(D) for pipes, flanges, manual valves and actuated valves. For the other equipment types, f(D) is equal to the constant C.

using separate parameters for C, a, n, m and Frup. These variables are used in DNVs software LEAK to produce the leak frequencies that are presented in the datasheets of this report.

4.2 Leak scenarios Analysis of the HCRD reveals a large number of scenarios with a significant difference between the recorded released mass and the mass that would be estimated by using a standard QRA methodology based on the recorded incident data. The HCRD includes many leaks that have occurred at very low system pressures. In order to account for this the analysis divides the leaks in HCRD into 2 main scenario categories: “full pressure leaks” and “zero pressure leaks”.


Full pressure leaks This scenario category is intended to be consistent with QRA models that assume a leak through the defined hole, beginning at the normal operating pressure, until controlled by isolation and blowdown, with a probability of isolation/blowdown failure. This is subdivided as follows: ■■

Full leaks which are intended to be consistent with QRA models that assume a leak through the defined hole, beginning at the normal operating pressure, until controlled by ESD1 and blowdown, with a small probability of ESD/blowdown failure. These are subdivided as follows

– ESD isolated leaks, which are defined as cases where the outflow quantity is comparable with that predicted for a leak at the operational pressure. – Late isolated leaks, presumed to be cases where there is no effective ESD of the leaking system, resulting in a greater outflow quantity. Late isolated leaks are defined as cases where the outflow is greater than predicted for a leak at the operational pressure controlled by the slowest credible ESD and no blowdown. ■■

imited leaks, presumed to be cases where the outflow L quantity is significantly less than from a leak at the operational pressure controlled by the quickest credible ESD (after 30 seconds) and blowdown (according to API) initiated 60 seconds later. This is presumed to be cases where there exist restrictions in the flow from the system inventory, as a result of local isolation valves initiated by human intervention or process safety systems other that ESD and blowdown.

Normally a quantitative risk assessment will assume that all leaks are full leaks because these have the potential of developing into serious events endangering personnel and critical safety functions. From these leak frequencies the analyst can use a standard event tree approach for the subsequent consequence assessment. This includes probabilities for ESD and BD failure.

reflect that these events involve reduced release mass and durations.

Zero pressure leaks This scenario includes all leaks where the pressure inside the leaking equipment is virtually zero (0.01 barg or less). This may be because the equipment has a normal operating pressure of zero (e.g. open drains), or because the equipment has been depressurised for maintenance, but not de-inventoried. These leaks may typically be very small gas releases, short lasting oil spills, or liquid releases from atmospheric tanks. Most likely they represent a significantly reduced major accident risk potential relative to a pressurised release through the same hole size (although they do pose occupational safety issues) and the contribution to the overall risk level as predicted in QRA studies is considered negligible.

4.3 Uncertainties There are several significant uncertainties in fitting a curve to the available leak data. Some uncertainties are due to the way that leak data is reported to the HSE and the lack of data for larger events. Other sources of uncertainty include issues about an accurate population count and the accuracy with which the operator records the data. The sources of uncertainty are illustrated in Figure 3.

4.4 Allocation of Leak Events The method of allocating leak records in HCRD into the scenarios is as follows: ■■ Identify the zero pressure leak events in order to discount them from the analysis ■■ Estimate the initial release rate Qo from the hole, based on parameters recorded in HCRD, ■■ Estimate a range of plausible release quantities, REmin to REmax, based on typical ESD and blowdown response ■■ Compare the recorded release quantity in HCRD to the estimated release quantity range to determine the scenario. The scenario allocation criteria are (in order): Zero pressure leaks – actual pressure in HCRD < 0.01 barg. ■■ Limited leaks – recorded release quantity in HCRD < REmin·/ D ■■ ESD isolated leaks – recorded release quantity in HCRD in the range REmin / D to REmax· ■■ Late isolated leaks – recorded release quantity in HCRD > REmax· ■■

Limited leaks may be of as much concern for personnel risk as full leaks in the period immediately following the start of the release, but they will have a shorter duration. Hence the potential for them developing into any major concern for other safety functions, such as structural integrity, evacuation means, escalation, etc. Any consequence calculations should be modelled as for ESD and late isolated leaks, but

1

With the assumption that process shutdown (PSD) is the shutdown of a particular section rather than the whole platform the PSD system may have the same effect as far as QRA modelling of release rates is concerned.


Holes<1mm only reported since 2001

All releases

LEAK Function

Holes1-2mm probably under reported

Frequency Exceeding (/year)

1.E-03 1.E-04 1.E-05 Exposed population declines as d approaches D 1.E-06

Holes >100mm not specified since 2001 Uncertainty increases for largest events probably under reported

1.E-07 0.1

1

10

100

Hole Diameter (mm) Figure 3 Uncertainties in applying curve to data

D is a disproportion factor. It is used to ensure that the classification of limited leaks is appropriate. The value given to D is typically 4 and this is the value which has been used in the calculation of the tables in this document.

As a simple indication of the relative importance of each leak scenario using the methods and criteria above, Figure 4 shows the breakdown of all leaks in HCRD for the period 1992-2010. This shows that approximately 6% of leaks are at zero pressure, and that 48% are limited leaks. Of the remaining 3% Late isolated 7% 46% leaks, 3% are consistent with late isolation.

Full leak 49%

Full pressure leaks 94% HCRD Leak (Total)

ESD isolated 93%

48%

Limited leak 51%

Zero pressure leaks 6%

6%

Figure 4: Event Tree of Leak Scenarios [4]

Release Type Zero Pressure leak Full pressure leak

Total

43%

SUM

Figure 4 can be further subdivided to produce Table 3. This indicates the effects that each individual fluid has on the leak type. Table 3 can be used in conjunction with the data sheets to obtain an estimate of the frequencies of limited leaks.

100%

Total

GAS LEAK

OIL LEAK

CONDENSATE LEAK

2-PHASE LEAK

NONPROCESS

6%

6%

7%

7%

2%

8%

Limited leak

48%

33%

75%

64%

67%

53%

Full leaks

ESD isolated

43%

57%

16%

27%

30%

36%

Late Isolated

3%

4%

2%

2%

1%

3%

100%

100%

100%

100%

100%

100%

Table 3 Proportion Distribution of leak incidents in the HCRD database2 (%)3 [4]

2

The figures take account of HCRD data until March 2010.

3

The data that supports the distribution of 2-Phase leaks and condensate leaks are not very comprehensive and the uncertainty in these numbers is therefore larger than for the other phases e.g. gas and oil leaks. The given distribution for 2-phase leaks and condensate leaks represents a best estimate.


5. Calculating Release Rate Where: Where:

- QL = initial liquid release rate (kg/s) 5. CALCULATING RELEASE RATE In order to estimate the initial release rate Q o from the hole and a range of plausible release quantities REmin Where: 5. CALCULATING RELEASE RATE - Q liquid release rate (kg/s) CDL = initial discharge coefficient 5. CALCULATING RATE to REmax aRELEASE series of equations are used. 2 5. CALCULATING RELEASE RATE - C discharge coefficient A hole area (m ) In order to estimate the initial release rate Q from the hole and a range of release D==plausible o 5. CALCULATING RELEASE = initial liquid raterelease (kg/s) In order to estimate the initialRATE release rate Qo from the -holeQLand a range of plausible 2 release In order initial rate Q hole andofaa the range plausible release - fluid A hole area ) (kg/m3) quantities seriesrelease of equations used.the The phase refers to the(m initial o from ρof = liquid density min to REthe max a L= In order to toREestimate estimate the initial release Qare from the hole release o of - phase CofD =plausible discharge coefficient quantities REmin to REmax rate a series equations areand used. range The of the fluid refers to3 the initial 2 quantities RE to of equations are used. The phase ofintheand refers to the initial order to RE estimate the initial release rate Qostate fromof the aL range plausible state of In fluid in the equipment before a leak. - fluid ρ = liquid density ) (N/m The phase ofseries the fluid refers to the initial fluidhole = of atmospheric pressure = 1025)N/m min max a P initial pressure ofrelease liquid (absolute) 2 (kg/m quantities RE min to REmax a series of equations are used. The phase of the fluidao refers to the initial

- A = hole area (m ) state of fluid in the equipment before a leak. 5 state of fluid in the equipment before leak. g the = acceleration to gravity 9.81 m/s2 the equipment aa leak. quantities RE a series of equations are used. The phase fluid refers topressure, the initial - ofP pressure ofdue liquid atmospheric N/m2)=2 (absolute) min to REbefore max o a = initial 3 10(N/m state of fluid in the equipment before a leak. - ρby: For gas state releases the initial release rate from high pressure equipment is given L = liquid density (kg/m ) 5 2 2 = liquid surface above hole (m) ofFor fluid in releases the equipment before a leak. P atmospheric pressure, 10 N/m gh a==acceleration due to gravity=9.81m/s gas the initial release rate from high pressure-equipment isheight given of by: 2 For gas releases the initial release rate from high pressure equipment is given by: P = initial pressure of liquid (N/m ) (absolute) o 2 releases the rate initial release from high pressure For gas releasesFor thegas initial release from highrate pressure equipment is given - g = acceleration duesurface to gravity=9.81m/s h by: height of liquid above hole (m) 2 !!! release rate from high pressure equipment - By P = atmospheric pressure, 10h,5 N/m For gasequipment releases the initial isaneglecting given by: the liquid head, and replacing the pressure given !" is ! !!! !!!by: - h = height of liquid surface above hole (m) 2 !!! !! = !! !!! Eqn 5 ! !!! !" - the g =liquid acceleration to gravity=9.81m/s By neglecting h due andpressure replacing the pressure with the gauge pressure of !!! term with head, the gauge of the liquid thisterm can be simpli!!! !! ! = ! !" !!! ! ! !! Eqn 5 ! !!! !"! ! ! !!!! !!!!! Eqn 5 ! !!! h = height of liquid surface above hole (m) By neglecting the liquid head, h and replacing the pressure term with the gauge pressure of liquid this can be simplified to: !!! = =! !! !! Eqn 5 fied to: !!!!" ! ! !! ! !!! !!! ! !!! ! = ! !! Eqn 5 ! ! ! liquidEqn this5 can be simplified!to: Where: !!! !!! By neglecting the liquid !! head, h and replacing the pressure term with the gauge pressure of Where: Eqn 9 !! = !! 2 !! !! E Where: Where: ! ! to: !! liquid this can be simplified - Q g = initial gas release rate (kg/s) Where: ! = ! 2 ! ! E Where: ! ! ! ! - Qg = initial gas release rate (kg/s) ! - Q initial gascoefficient release rate (kg/s) C = initial gas release rate(kg/s) (kg/s) = 0.61 andthe liquid head h, the equ QgDgg = As a simple !! approximation, As a simple approximation, substituting Csubstituting ! D = 0.61 andCDneglecting - Q = discharge initial rate - gasCrelease 2D = discharge coefficient = !! thesubstituting 2 !! !! Ch, the !!to: - C discharge Qarea gas release 2rate (kg/s) can A (mcoefficient ) coefficient D= g = initial = = discharge CDD neglecting equation can be simplified and neglecting the liquid head h, the equE As abe simple approximation, simplified ! liquid head D = 0.61 - C =-hole discharge coefficient - A 2= hole area (m 2) 2 coefficient 2 - A = hole area (m ) C = discharge P = initial pressure of gas (N/m ) absolute A o= hole= area hole (m area (m ) to: to: D can 2 be simplified - A - Po )= initial ) absolute 2 of 2 pressure !! ! a simple approximation, substituting 2 gas (N/mAs ! !! !! CD = 0.61 and neglecting the liquid head h, the equ E 2) absolute - P = initial pressure of gas (N/m = initial pressure of gas (N/m ) absolute P - molecular A = hole area (m ) M weight of gas o ! = 2.1 x 10 ! O - Po = initial pressure of gas (N/m ) absolute - M = molecular weight of 2gas !! ! can be simplified to: ! = 2.1 x 10 ! ! ! E = weight ofofgas Eqn 10 - M =ratio weight of gas gas -molecular Pof =specific initial pressure gas (N/m ) absolute γM == heats omolecular ! ! ! - M molecular - of γweight =specific ratioof ofheats specific heats where: g == ratio = ratio -universal M = molecular weight of gas - γR of specific heats gas constant = 8314 (J/kg mol K) - γ = ratio of -specific heats !! = 2.1 x 10!! ! ! !! !! E where: R =gas universal gas constant = 8314 (J/kg mol K)where: R ==-universal = universal constant = 8314 J/kg - R gasofconstant constant =heats 8314 (J/kg molmol K)K γ= ratio specific T initial temperature of gas (K) o= universal - d = hole diameter (mm) - R gas = 8314 (J/kg mol K) - Ttemperature of gas (K) o = initial temperature = of (K) gas (K) To = initial d = hole diameter (mm) 3 - T temperature ofconstant gas R initial =temperature universal gasof = 8314 (J/kgwhere: mol K) - d hole diameter (mm) ρrL= =Liquid density (kg/m ) 3) - Too =- initial !!gas (K) ! = liquid density (kg/m L 3 - Tand initial temperature gas (K) !! o = noting ! Rearranging the above that ! = of gives: ρ =Liquid density (kg/m ) (bar gauge) release of liquid ! ! noting = initial pressure PdLL =initial Rearranging the gives: - P !! Rearranging the above aboveand and noting = gives: !! = hole diameter (mm)of liquid (bar gauge) !that !that Rearranging the above and noting that = gives: !! ! - PL=initial release of liquid3 (bar gauge) ! ! Rearranging the above and noting that !! = !! ! gives: ! - may ρL=Liquid ) This equation be useddensity for oil, (kg/m condensate and non-process releases. !! that !!!! != gives: Rearranging the above and noting !!! This equation may be used oil,gauge) condensate and non!! ! ! ! - may PL=initial liquidfor (bar This equation be usedrelease for oil,ofcondensate and non-process releases. !! !! ! ! !!! !!! !! !!! Eqn 6 process releases. ! ! Eqn 6 ! Two-phase releases are less amenable to simple approximation, but since they form a small !! ! ! = !! ! !!! ! ! ! ! =!!! !! !! !!! !!! !!!! !! ! Eqn 6 !! ! !! ! ! !!! ! !! ! ! = ! Eqn 6 This equation may be used for oil, condensate and non-process releases. !!!!!! ! Two-phase releases are less amenable to simple approximation, but since they form a small proportion of HCRD, they are represented by: ! ! !! = !! ! !!!!!! !!! !!!!! Eqn 6 ! ! !! ! ! ! Two-phase releases areEqn less 6amenable to simple approxima!! = !! !!!! ! proportion of HCRD, they are represented by: ! !! substituting γ = 1.31, CD !!! Approximating the gauge pressure to absolute pressure, =less 0.85, and! to simple approximation, but since they form a smal !"# Two-phase are Approximating the to absolute pressure, tion, since they of HCRD, they ! + form E Approximating thegauge gaugepressure pressure to absolute pressure, releases substituting γ !=!amenable 1.31, C =a!small 0.85, proportion and ! =but !"#!! !"#!! Approximating the gauge pressure to absolute pressure, substituting γ = 1.31, C = 0.85, and converting the units of pressure to bar and noting that the units of the diameter are in mmrepresented we D ! by: D proportion HCRD, they are Approximatingconverting the gauge the to 0.85, absolute pressure, γof= units 1.31, CD!"# = 0.85, andto converting the units of the substituting gpressure = 1.31, are represented by:and!are!in D = pressure ! ! + E unitsCof bar and substituting noting that the diameter mm we ! =of ! ! converting the units units of of the pressure bar and and to noting that the the units of ofsubstituting the diameter diameter are in C mm we γ =!"#!! 1.31, 0.85, and gaugetopressure absolute pressure, have: Approximating D = !"#!! converting the pressure bar that are in mm we pressure to bar andtonoting thatnoting the units of the units diameter are where: the have: !"# ! have: converting theweunits pressure to bar and noting that the units of the diameter!are mm !we +in E ! ! = ! have: in mm have:of Eqn 11 where: !"#!! ! !! ! !! !! !! ! Eqn rate 7!"#!! have: !! = 1.4 x 10 ! ! = 1.4 x 10 Q =initial release (kg/s) o ! !! !! Eqn 7 !! !! = 1.4 x 10!! Eqn !! ! ! !! !! where: ! !! !! Eqn 77 (kg/s) rate =releaserelease rate (kg/s) o g=initial ! = 1.4 x 10 ! Eqn 7 - Q where: !! ! Where: 7 - Q rate release (kg/s)Eqn Where: !! = 1.4 x 10 ! !! !! = initial rate (kg/s) QogL =release Where: Q =initial release rate (kg/s) o Where: - Q rate (kg/s) GOR = = gas oil ratio gas per kg oil) release rate(kg (kg/s) QgL =release Where: - d = hole diameter (mm) Where: - Qg=release rate (kg/s) - d = hole diameter3 (mm) release rate(kg (kg/s) QL = = d g==hole = diameter hole diameter (mm) - GOR gas oil ratio gas per kg oil) - d (mm) ρ initial density of gas ( kg/m ) 3 - d = hole diameter (mm) - QL=release rate (kg/s) - ρdensity density 3 of3)gas ( kg/m ) g = initial = initial of gas (kg/m GOR = gas oil ratio (kg gas per kg oil) rgg =-initial d = hole diameter (mm) 3) - ρ density of gas ( kg/m P pressure of gas (bar gauge) - ρg = initial density of gas (pressure kg/m ) of gas - GOR = gas oil ratio (kg gas per kg oil) - Ppressure 3 (bar gauge) g = initial = of of gas (bar gauge) Pg = initial - P pressure of gas gas (bar gauge) ρginitial =pressure initial density gas ( kg/m ) - Pgg =- initial of (bar gauge) For liquid releases,-thePinitial release rate isofgiven by: gauge) The initial release rate is assumed to continue at a constant initial pressure gas (bar For liquidg = releases, the initial release rate is given by: For liquid releases, the initial release rate is given by: For liquid releases, the rate initial rate is given by: rate until the inventory is isolated. After isolation, the For liquid releases, the initial release is release given by: !! releases, = !! ! 2! !! −release !! + !rate Eqn 8 as the isolated section is depressurised release rate declines For liquid the! initial ! !ℎis given by: !! = !! ! 2!! !! − !! + !! !ℎ Eqn 8 ! = ! ! 2! ! − ! + ! !ℎ Eqn Eqn 8 through the leak. of isolated sections can then ! = !! ! 2!! !! − !! + !! !ℎ ! EqnBlowdown 88 ! ! ! ! ! ! !! = !! ! 2!! !! − !! + !! !ℎ Eqn further increase the rate at 8which the section is depressur-

Where: = initial liquid release rate (kg/s) QL = discharge coefficient CD A = hole area (m2) = liquid density (kg/m3) rL = initial pressure of liquid (N/m2) (absolute) Po

ized and hence decrease the release rate through the hole.


Isolation

Blowdown

QT

Equipment

No.

Size

Process Vessel

0.5

8”

1

6”

Shell and Tube Heat Exchanger

1

6”

Flange

11

8”

Flange

5

6”

2

8”

2

½”

3

8”

10

8”

5

6”

Centripetal Compressor

QO

Release Rate

QB Leak Flow

Actuated Valve The initial release rate is assumed to continue at a constant rate until the inventory is isolated. The The initial initial release release rate rate is is assumed assumed to to continue continue at at aa constant constant rate rate until until the the inventory inventory is is isolated. isolated. Bore Fittings After isolation, the release rate declines as the isolated section is depressurizedSmall through the leak.

After isolation, the release rate as isolated section depressurized through the The initial release is assumed to continue a constant rateis the inventory is isolated. After isolation, therate release rate declines declines as the the at isolated section isuntil depressurized through the leak. leak. Blowdown of isolated sections can then further increase the rate at which the section is Manual Valve Blowdown of isolated sections can then further increase the rate at which the section is After isolation, the release rate can declines the isolated depressurized through Blowdown of isolated sections then as further increasesection the rateis at which the section is the leak. t Ithe release tB Time depressurized and hence decrease rate through the hole. Process Pipe depressurized hence decrease release rate through the hole. Blowdown of and isolated sections canthe then further increase at which the section is depressurized and hence decrease the release rate throughthe therate hole. Figure Decline rate of release rate with depressurized and hence decrease the 55release through the time hole. Figure Decline of release rate with time Figurerate 5 Decline of release rate with time Figure 5 Decline of release with time Figure 5 Decline of release rate with time

The expected release quantity is calculated as follows: The quantity is The expected expected release release quantityrelease is calculated calculated asisfollows: follows: The expected quantityas calculated as follows: ! ! The expected release quantity is calculated as follows: ! ! ! ! ! I 1− !! + ! ! = ! !! ! = !!! ! !!! + + ! ! + II 1 1− − !!!!!! + +! !!! !!!!!! Eqn 12 ! = !!! !! !! !! !! = !! !! + I 1 − ! + !! ! !! !! Where: Where: Where: Where: !!! !! !!! Where: !!! !!! !! !!! ! ! !! ! = ! exp ! I! ! = ! exp ! ! ! Eqn 13 ! ! = !! exp I !!! !I! !!! !! = !!!!exp I ! ! ! = Eqn 14 ! = II !!!!!! ! ! ! =I! ! !!! ! = I !!! !!!! ! ! ! Eqn 15 ! = ! !! !! = !!! !! !! ! = ! ! ! ! ! !!!

! ! !! ! ! !! !! ! !! !

Process Pipe

Table 4 Parts Count of Isolatable System

Cooler

Eqn 12 Eqn Eqn 12 12 Eqn 12 Eqn 13 Eqn Eqn 13 13 Eqn 13 Eqn Eqn 14 14 Eqn 14

KEY

HP Section Scrubber

HP Gas Compressor

Eqn 14 Eqn 15 Eqn Eqn 15 15 Actuated Valve Eqn 15 Manual Valve

= !!= ! rate through leak when blowdown starts Q !!B rate !!release - QB = release through leak when blowdown starts (kg/s) !! -- Q rate through leak QBB = = release release rate through leak when when blowdown blowdown starts starts (kg/s) (kg/s) Flange (kg/s) - QT = total release rate through leak and blowdown valve when blowdown starts (( kg/s) -- Q release rate through leak and blowdown valve when blowdown starts release rate through leak when blowdown starts (kg/s) QTTB = = total total release rate through leak and blowdown valve when blowdown starts ( kg/s) kg/s) total release rate(kg) through leak and blowdown QT in= isolated - I = inventory section Small Bore Fitting Figure 6: Sample Isolatable Section -- IIQ= inventory in isolated section (kg) = total release rate through leak and blowdown valve when blowdown starts ( kg/s) =T inventory in isolated section (kg) valve when when blowdown starts ( kg/s) - MB = mass remaining blowdown starts -- M = mass remaining when blowdown starts I =BB inventory in= isolated section (kg) section M = mass when starts (kg) I remaining inventory inblowdown isolated - t = time from start of leak (s) -- ttM= from start of leak (s) = mass remaining when blowdown starts =B time time from start of leak (s) = mass remaining when blowdown starts MB - tI = time from start of leak to isolation (s) -- ttI = = time from start of leak to isolation (s) to isolation (s) time from start of leak (s) 5.1 Sample Calculation t = time from start of leak (s) I - tB = time from start of leak to blowdown (s) -- ttBI = = time from start of leak to blowdown = time start of leak (s) to(s) isolation (s) Figure 6 shows a sample isolated section and Table 4 displays of leak isolation timetI from from start start of from leak to to blowdown (s) B = time - ρd = density factor -- ρ density factor = time of leak to(s) blowdown (s) tB from the part count for this simplified system. Process pipe is start of from leak start to blowdown ρtBdd == =time density factor - RE = expected ESD-limited outflow (kg) = density factor rd factor length in metres, not number of pipes. There are 2 pipe -- R ESD-limited outflow ρEEd = density R == expected expected ESD-limited outflow (kg) (kg) - b = blowdown valve diameter = expected ESD-limited outflow (kg) RE sizes in the system. The 6” pipe connects the scrubber to the -- b valve diameter expected ESD-limited outflow (kg) bRE= ==blowdown blowdown valve diameter b = blowdown valve diameter compressor and on to the cooler. The remaining pipework - density b = blowdown valve The factor is set to 11diameter for gas and 2-phase releases, but for liquid releases the following The the following is 8”. The analysis is being conducted on a gas stream. The density density factor factor is is set set to to 1 for for gas gas and and 2-phase 2-phase releases, releases, but but for for liquid liquid releases releases theleak following formula is used: formula is used: The is density set to gas and 2-phasebut releases, The density factor set tofactor 1 forisgas and1 for 2-phase releases, for liquid releases the following formula is used: !! but for liquid releases the following formula is used: There are a number of assumptions that are made in countformula is used: ! ! ! Eqn 16 !! ! =! ! !!! = =! ! !!!!! Eqn 16 16 ing the parts: Eqn !! ! Eqn 16 ! Eqn 16 ! =!! Where: ! Where: ■■ For parts that are independent of equipment size (all Where: items except for valves, pipe and flanges) the largest pipe Where: - fWhere: = density number (0.5) -- ff = number (0.5) = density density number (0.5) 3 (0.5) diameter that is connected to the piece of equipment is f = density number 3 - ρg = gas density (kg/m3) ρfrgg= = =density gas= density density (kg/m number (0.5) -- ρ gas (kg/m )) 3 3) gas density (kg/m taken to be the size. g - ρl = liquid density (kg/m ) 3 3 -- ρ density (kg/m gas= density (kg/m ) 3)) 3) liquid density (kg/m ρrlllg = ==liquid liquid density (kg/m 3 ■■ Only half the scrubber is counted since only the top half - ρl = liquid density (kg/m ) Once the frequency of a hole size occurring is determined the release rate for that particular diameter of hole can be calculated thereby finding the frequency of that release rate.

is in contact with gas under normal operating conditions. (The lower half of the scrubber is included in a separate count for the liquid stream).


■■

■■

wo flanges are counted on each valve. (The base leak T frequency of the valve does not account for flange connections).

The results are shown in Table 3. Then by applying Eqn 1 to each type of equipment, the contribution of each hole size can be examined.

he actuated valves at the boundaries of the system are T ESDVs. These isolate the section. Only half these valves are counted and one flange connection.

Equipment

No.

Size

Frequency [ /equipment year]

Total [Leaks/year]

Process Vessel

0.5

8”

2.155 x 10-3

1.077 x 10-3

Centrifugal Compressor

1

6”

1.061 x 10-2

1.061 x 10-2

Shell and Tube Heat Exchanger

1

6”

3.446 x 10-3

3.446 x 10-3

Flange

11

8”

1.286 x 10

1.414 x 10-3

Flange

5

6”

1.117 x 10

5.585 x 10-4

Actuated Valve

2

8”

5.921 x 10

1.184 x 10-3

Small Bore Fittings

2

½”

-4

5.894 x 10

1.178 x 10-3

Manual Valve

3

8”

1.437 x 10-4

4.311 x 10-4

Process Pipe

10

8”

6.945 x 10

6.945 x 10-4

Process Pipe

5

6”

7.349 x 10

3.674 x 10-4

-4 -4 -4

-5 -5

Total

0.021

Table 5 Sample Calculation

5.2 Alternative sources of leak frequency data The leak frequency data and methodology presented in this document are based on analysis and application of data in the HSE Hydrocarbon Release Database. A number of other data sources and methodologies have been published, but DNV considers the HSE database to be the best available for most QRA applications. Other available databases include the handbook for failure frequencies which was developed by the Flemish (Belgian) Government [5], and the Reference Manual Bevi Risk Assessments which was developed by the Dutch National Institute of Public Health and the Environment[5] [6]. DNV has compared the leak frequency result from these sources against the DNV methodology. A comparison of results based on the sample isolatable section is shown in Table 4. The comparison of results shows that the leak frequency estimated using DNV’s method for this case is greater than that obtained by the Dutch and Belgian methodologies. The Belgian and Dutch methodologies present leak frequency data for equipment systems; and omit any explicit counts of the flanges, valves and instruments associated with major equipment items. This is the main reason why the Dutch and Belgian methodologies produce lower estimate of leak frequency.

Data source & methodology

Leak frequency (per year)

HCRD (DNV)

0.021

Dutch government

0.006

Belgian government

0.012

Table 6: Comparison of Estimated Leak Frequencies

The experience base of the Dutch and Belgian methodologies does not match the large experience of leaks contained in the HCRD; leak frequencies derived from the HCRD are more accurate primarily because HCRD is the largest database with information for over 70 different sizes and types of equipment, collected systematically over the last 20 years. DNV notes that the absence of separate frequencies for flanges, valves and instruments in the Belgian and Dutch methodologies also means that risk assessments performed using these methods are insensitive to some design decisions, such as the benefits of all-welded designs. Figures 7 and 8 illustrate these points. They show the ratios of frequencies for each type of equipment obtained by dividing the DNV frequency by the Dutch/Belgian frequency for each type of equipment. The figures show for the majority of equipment types the DNV methodology quotes higher frequencies. The figures highlight the large difference between the frequencies. The explanations of the differences are as discussed in Section 3.2.


Storage Vessel Centrifugal Compressor Heat Exchanger Plate Heat Exchanger (HC in tube) Heat Exchanger (HC in shell) Recipricating Compressors Centrifugal Pump Process Vessel 20(in.), Im in LengthProcess Pipeline 6(in.), Im in LengthProcess Pipeline 2(in.), Im in LengthProcess Pipeline

0.1

1

10

100

1000

Figure 7: Ratio of frequencies - DNV data to Belgium tabulation

Storage Vessel Recipricating Compressors Centrifugal Compressor Recipricating Pump Centrifugal Pump Process Vessel 20(in.), Im in LengthProcess Pipeline 6(in.), Im in LengthProcess Pipeline 2(in.), Im in LengthProcess Pipeline

0.01

0.1

Figure 8: Ratio of frequencies - DNV data to Netherlands tabulation

1

10

100

1000


6. LEAK FREQUENCY DATASHEETs The following pages include leak frequency data for 17 types of process equipment. These process equipment types are split into two categories: ■■ Diameter Dependent ■■ Diameter Independent As explained in section 4.1 there is enough information in the HCRD to determine all the constants of the leak frequency data equation for the diameter dependent equipment types (Process pipe, Flanges, Manual and Actuated Valves). Leak frequencies for other types of equipment are considered to be independent of equipment size. For equipment considered independent of equipment size the leak frequencies are quoted to an equipment size of 6 inches. This is because the leak frequencies remain the same for the larger diameters.

Typically the parts count is multiplied by the total leaks to determine the overall leak frequency. For a more in-depth analysis the parts count may be multiplied by the values in the full, limited and zero pressure leak columns, but for most purposes it will be sufficient to use only the “full” leak frequencies. The frequency of limited leaks can be obtained using data in Table 3. It may be noted that the sum of frequencies for full, limited and zero pressure leaks do not necessarily equal the total leaks. The small difference is due to total, full and zero pressure leaks being determined using different equations. The tables presented in this section have been generated using commercially available DNV LEAK software which implements the methodology described in this document.


Process Equipment Leak Frequencies

Equipment Type:

Centrifugal Compressors

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

The scope includes the compressor itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded.

Frequency Data:

Equipment Size

Category 1 -‐ 3 mm 3 -‐ 10 mm 10 -‐ 50 mm 50 -‐ 150 mm > 150 mm Total 1 -‐ 3 mm 3 -‐ 10 mm 10 -‐ 50 mm 50 -‐ 150 mm > 150 mm Total 1 -‐ 3 mm 3 -‐ 10 mm 10 -‐ 50 mm 50 -‐ 150 mm > 150 mm Total 1 -‐ 3 mm 3 -‐ 10 mm 10 -‐ 50 mm 50 -‐ 150 mm > 150 mm Total 1 -‐ 3 mm 3 -‐ 10 mm 10 -‐ 50 mm 50 -‐ 150 mm > 150 mm Total

0.5 in

1 in

2 in

4 in

6 in

Total 5.802E-‐03 2.462E-‐03 1.435E-‐03 0.000E+00 0.000E+00 9.699E-‐03 5.802E-‐03 2.462E-‐03 1.435E-‐03 0.000E+00 0.000E+00 9.699E-‐03 5.802E-‐03 2.462E-‐03 1.057E-‐03 3.772E-‐04 0.000E+00 9.699E-‐03 5.802E-‐03 2.462E-‐03 1.057E-‐03 3.772E-‐04 0.000E+00 9.699E-‐03

5.802E-‐03 2.462E-‐03 1.057E-‐03 2.257E-‐04 1.516E-‐04 9.699E-‐03

Full Pressure 5.583E-‐03 2.316E-‐03 1.300E-‐03 0.000E+00 0.000E+00 9.199E-‐03 5.583E-‐03 2.316E-‐03 1.300E-‐03 0.000E+00 0.000E+00 9.199E-‐03 5.583E-‐03 2.316E-‐03 9.686E-‐04 3.309E-‐04 0.000E+00 9.199E-‐03 5.583E-‐03 2.316E-‐03 9.686E-‐04 3.309E-‐04 0.000E+00 9.199E-‐03

5.583E-‐03 2.316E-‐03 9.686E-‐04 2.008E-‐04 1.300E-‐04 9.199E-‐03

Zero Pressure 1.324E-‐04 1.052E-‐04 2.624E-‐04 0.000E+00 0.000E+00 5.000E-‐04 1.324E-‐04 1.052E-‐04 2.624E-‐04 0.000E+00 0.000E+00 5.000E-‐04 1.324E-‐04 1.052E-‐04 9.519E-‐05 1.672E-‐04 0.000E+00 5.000E-‐04 1.324E-‐04 1.052E-‐04 9.519E-‐05 1.672E-‐04 0.000E+00 5.000E-‐04

1.324E-‐04 1.052E-‐04 9.519E-‐05 4.428E-‐05 1.229E-‐04 5.000E-‐04



Equipment Type:

Reciprocating Compressors

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

The scope includes the compressor itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in


3.641E-‐02 1.581E-‐02 6.973E-‐03 1.532E-‐03 1.067E-‐03 6.179E-‐02


3.685E-‐02 1.512E-‐02 6.238E-‐03 1.275E-‐03 8.107E-‐04 6.029E-‐02

Zero Pressure 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00

0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00



Equipment Type:

Filters

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in


1.870E-‐03 9.307E-‐04 4.820E-‐04 1.258E-‐04 1.109E-‐04 3.520E-‐03


1.608E-‐03 7.559E-‐04 3.661E-‐04 8.894E-‐05 7.089E-‐05 2.890E-‐03

Zero Pressure 2.453E-‐04 1.605E-‐04 2.232E-‐04 0.000E+00 0.000E+00 6.290E-‐04 2.453E-‐04 1.605E-‐04 2.232E-‐04 0.000E+00 0.000E+00 6.290E-‐04 2.453E-‐04 1.605E-‐04 1.150E-‐04 1.082E-‐04 0.000E+00 6.290E-‐04 2.453E-‐04 1.605E-‐04 1.150E-‐04 1.082E-‐04 0.000E+00 6.290E-‐04

2.453E-‐04 1.605E-‐04 1.150E-‐04 4.219E-‐05 6.598E-‐05 6.290E-‐04



Equipment Type:

Flange

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

The following frequencies refer to a flanged joint, comprising two flange faces, a gasket (where fitted), and two welds to the pipe. Flange types include ring type joint, spiral wound, clamp (Grayloc) and hammer union (Chicksan)

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in

Total 3.725E-‐05 1.364E-‐05 1.227E-‐05 0.000E+00 0.000E+00 6.316E-‐05 4.037E-‐05 1.479E-‐05 1.279E-‐05 0.000E+00 0.000E+00 6.795E-‐05 4.628E-‐05 1.695E-‐05 6.126E-‐06 7.661E-‐06 0.000E+00 7.701E-‐05 5.745E-‐05 2.104E-‐05 7.605E-‐06 8.062E-‐06 0.000E+00 9.415E-‐05 6.816E-‐05 2.496E-‐05 9.023E-‐06 1.594E-‐06 6.852E-‐06 1.106E-‐04

Full Pressure 3.538E-‐05 1.239E-‐05 1.031E-‐05 0.000E+00 0.000E+00 5.808E-‐05 3.771E-‐05 1.320E-‐05 1.066E-‐05 0.000E+00 0.000E+00 6.156E-‐05 4.229E-‐05 1.480E-‐05 5.076E-‐06 6.269E-‐06 0.000E+00 6.844E-‐05 5.133E-‐05 1.797E-‐05 6.161E-‐06 6.540E-‐06 0.000E+00 8.200E-‐05 6.028E-‐05 2.110E-‐05 7.235E-‐06 1.206E-‐06 5.603E-‐06 9.542E-‐05

Zero Pressure 1.156E-‐06 8.773E-‐07 2.367E-‐06 0.000E+00 0.000E+00 4.400E-‐06 1.156E-‐06 8.774E-‐07 2.367E-‐06 0.000E+00 0.000E+00 4.400E-‐06 1.157E-‐06 8.780E-‐07 7.519E-‐07 1.616E-‐06 0.000E+00 4.403E-‐06 1.171E-‐06 8.891E-‐07 7.614E-‐07 1.630E-‐06 0.000E+00 4.452E-‐06 1.241E-‐06 9.420E-‐07 8.067E-‐07 3.550E-‐07 1.343E-‐06 4.687E-‐06



Equipment Type:

Flange

Frequency Data:

Equipment Size

Category 1 -‐ 3 mm 3 -‐ 10 mm 10 -‐ 50 mm 50 -‐ 150 mm > 150 mm Total 1 -‐ 3 mm 3 -‐ 10 mm 10 -‐ 50 mm 50 -‐ 150 mm > 150 mm Total 1 -‐ 3 mm 3 -‐ 10 mm 10 -‐ 50 mm 50 -‐ 150 mm > 150 mm Total

10 in

14 in

20 in

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Total 8.880E-‐05 3.252E-‐05 1.176E-‐05 2.077E-‐06 7.110E-‐06 1.423E-‐04 1.088E-‐04 3.984E-‐05 1.440E-‐05 2.544E-‐06 7.360E-‐06 1.729E-‐04 1.379E-‐04 5.051E-‐05 1.826E-‐05 3.226E-‐06 7.724E-‐06 2.176E-‐04

Full Pressure 7.801E-‐05 2.731E-‐05 9.362E-‐06 1.560E-‐06 5.780E-‐06 1.220E-‐04 9.559E-‐05 3.346E-‐05 1.147E-‐05 1.912E-‐06 5.956E-‐06 1.484E-‐04 1.218E-‐04 4.263E-‐05 1.462E-‐05 2.436E-‐06 6.218E-‐06 1.877E-‐04

Zero Pressure 1.884E-‐06 1.430E-‐06 1.225E-‐06 5.388E-‐07 1.779E-‐06 6.856E-‐06 4.148E-‐06 3.148E-‐06 2.696E-‐06 1.186E-‐06 3.316E-‐06 1.449E-‐05 1.454E-‐05 1.103E-‐05 9.450E-‐06 4.158E-‐06 1.037E-‐05 4.955E-‐05



Equipment Type:

Fin Fan Heat Exchanger

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

The scope includes the heat exchanger itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange is also excluded.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in


7.997E-‐04 3.802E-‐04 1.866E-‐04 4.600E-‐05 3.740E-‐05 1.450E-‐03


7.997E-‐04 3.802E-‐04 1.866E-‐04 4.600E-‐05 3.740E-‐05 1.450E-‐03

Zero Pressure 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00

0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00



Equipment Type:

Plate Heat Exchanger

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

The scope includes the heat exchanger itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange is also excluded.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Shell Side Heat Exchanger

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Shell & tube type heat exchangers with hydrocarbon in the shell side. The scope includes the heat exchanger itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Tube Side Heat Exchanger

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Shell & tube type heat exchangers with hydrocarbon in the tube side. The scope includes the heat exchanger itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Pig Trap

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Includes pig launchers and pig receivers. The scope includes the pig trap itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Process Pipe

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Includes pipes located on topsides (between well and riser) and subsea (between well and pipeline). The scope includes welds but excludes all valves, flanges, and instruments.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Process Pipe

Frequency Data:

Equipment Size


10 in

14 in

20 in

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10






Equipment Type:

Centrifugal Pump

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Centrifugal pumps including single-‐seal and double-‐seal types. The scope includes the pump itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Reciprocating Pump

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Reciprocating pumps including single-‐seal and double-‐seal types. The scope includes the pump itself, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Small Bore Fittings

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Includes small-‐bore connections for flow, pressure and temperature sensing. The scope includes the instrument itself plus up to 2 valves, 4 flanges, 1 fitting and associated small-‐bore piping, usually 25mm diameter or less.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

Total 3.092E-‐04 1.373E-‐04 8.644E-‐05 0.000E+00 0.000E+00 5.330E-‐04 3.092E-‐04 1.373E-‐04 8.644E-‐05 0.000E+00 0.000E+00 5.330E-‐04 3.092E-‐04 1.373E-‐04 6.220E-‐05 2.424E-‐05 0.000E+00 5.330E-‐04

Full Pressure 2.998E-‐04 1.287E-‐04 7.643E-‐05 0.000E+00 0.000E+00 5.050E-‐04 2.998E-‐04 1.287E-‐04 7.643E-‐05 0.000E+00 0.000E+00 5.050E-‐04 2.998E-‐04 1.287E-‐04 5.601E-‐05 2.042E-‐05 0.000E+00 5.050E-‐04

Zero Pressure 1.092E-‐05 7.144E-‐06 9.935E-‐06 0.000E+00 0.000E+00 2.800E-‐05 1.092E-‐05 7.144E-‐06 9.935E-‐06 0.000E+00 0.000E+00 2.800E-‐05 1.092E-‐05 7.144E-‐06 5.119E-‐06 4.815E-‐06 0.000E+00 2.800E-‐05



Equipment Type:

Actuated Valves

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Includes all types of actuated valves (block, blowdown, choke, control, ESDV and relief), but not actuated pipeline valves (pipeline ESDV and SSIV). Valve types include gate, ball, plug, globe and needle. The scope includes the valve body, stem and packer, but excludes flanges, controls and instrumentation.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Actuated Valves

Frequency Data:

Equipment Size


10 in

14 in

20 in


Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10





Equipment Type:

Manual Valves

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Includes all types of manual valves (block, bleed, check and choke). Valve types gate, ball, plug, globe, needle and butterfly. The scope includes the valve body, stem and packer, but excludes flanges, controls and instrumentation.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Manual Valves

Frequency Data:

Equipment Size


10 in

14 in

20 in

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10






Equipment Type:

Process Vessel

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

Includes all types of pressure vessel (horizontal/vertical absorber, knock-‐out drum, reboiler, scrubber, stabiliser, separator and stabiliser), but not the HCRD category “other”, which are mainly hydrocyclones. The scope includes the vessel itself and any nozzles or inspection openings, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded.

Frequency Data:

Equipment Size


0.5 in

1 in

2 in

4 in

6 in






Equipment Type:

Atmospheric Storage Vessel

Definition:

Rev.:

1

Date:

26/9/2012

Source:

HCRD 10/92 – 03/10

This datasheet applies to offshore atmospheric tanks. Includes types of vessel at atmospheric pressure (oil storage tanks). The scope includes the vessel itself and any nozzles or inspection openings, but excludes all attached valves, piping, flanges, instruments and fittings beyond the first flange. The first flange itself is also excluded.

Frequency Data: Equipment Size

0.5 in

1 in

2in

4in

6 in






7. REFERENCES 1. Department of Energy, The Public Inquiry into the Piper Alpha Disaster, 1990 2. Health and Safety Executive (HSE), Hydrocarbon release reporting and statistics (www.hse.gov.uk/offshore/ hydrocarbon.htm) accessed 2012 3. Spouge, J. (2005), New generic leak frequencies for process equipment, Process Safety Progress, 24, 4, pp249-257 4. Falck, A., Bain, B., & Rødsætre, L. (2009) Leak frequency modeling of offshore QRA based on the Hydrocarbon Release Database, Hazards XXI Conference, IChemE, Nov. 2009. 5. Flemish Government. (2009) Handbook Failure Frequencies 2009 for drawing up a safety report. LNE Department. Environment, Nature and Energy Policy Unit. Safety Reporting Division. 6. RIVM. (2009). Reference manual BEVI risk assessments. v3.1, Jan 1st 2009. Centre for External Safety, Netherlands National Institute of Public Health and the Environment

Disclaimer: This document describes certain of DNV’s

methods and applications. It was prepared on the basis of the at-the-time knowledge and information and it is provided to you for information purposes only. Nothing herein contains specific advice and DNV makes no representation or warranty regarding the information’s characteristics, fitness-for-purpose or other qualities. You as a reader understand and accept that you are solely responsible for your conclusions, use, or non-use of the information herein and that DNV is not liable for any consequence, damage, or loss arising therefrom, including but not limited to loss of opportunities, profits or otherwise.


Det Norske Veritas as NO-1322 Høvik, Norway I Tel: +47 67 57 99 00 www.dnv.com

THIS IS DNV DNV is a global provider of services for managing risk, helping customers to safely and responsibly improve their business performance. Our core competence is to identify, assess and advise on risk management. DNV is an independent foundation with presence in more than 100 countries.

GLOBAL SERVICES TO THE MARITIME, OIL & GAS AND ENERGY INDUSTRIES Safety, health and environmental risk management ■■ Enterprise risk management ■■

Asset risk management Technology qualification ■■ Verification ■■

■■

■■

■■

Ship classification Offshore classification

Main offices Aberdeen Cromarty House, 67-72 regent Quay Aberdeen AB115AR United Kingdom Tel: +44 (0)1224 335000

Dubai BurJuman Office tower 14th Floor Trade Center road United Arab Emirates Tel: +971 4 3526626

Houston 1400 Ravello Dr Katy, TX 77449 USA Tel: +1 281 396 1000

Rio de Janeiro Rua Sete de Setembro,111/12 Floor 20050006 Rio de Janeiro Brazil Tel: +55 21 3722 7232

Singapore DNV Technology Centre 10 Science Park Drive Singapore 118224 Tel: +65 6508 3750

Stavanger Bjergstedveien 1 NO-4007 Stavanger Norway Tel: +47 51 50 60 00

© Det Norske Veritas AS 03-2013 Design and print: coormedia.com 1301-044

Oslo Veritasveien 1 NO-1322 Høvik Norway Tel: +47 67 57 99 00

failure frequency guidance

Recommend Documents