Title: UKOPA Pipeline Product Loss Incidents (1962-2010) : Report Number: UKOPA/11/0076 Issue: Final v1.0

Report Number: UKOPA/11/0076
Issue: Final v1.0
Title:
UKOPA Pipeline Product Loss
Incidents (1962-2010)
Date of Issue:
November 2011
Author:
RA McConnell & Dr J V Haswell
Ref:
UKOPA/11/0076
Reproduction permitted with due acknowledgement

UKOPA Ambergate UK November 2011

Issue: Final v1.0
UKOPA PIPELINE FAULT DATABASE
Pipeline Product Loss Incidents

(1962 - 2010)
8th Report of the UKOPA Fault Database Management Group
Comprising:
National Grid
BP
Ineos
Sabic
Shell UK Limited
Shell EPE
E-ON UK
Wales & West Utilities
Scotia Gas Networks
Northern Gas Networks
Health and Safety Executive
Report prepared by R A McConnell & Dr J V Haswell for FDMG
Report Reference: UKOPA/11/0076

November 2011


Issue: Final v1.0
Comments, questions and enquiries about this publication should be directed to the
UKOPA Pipeline Fault Database Working Group Chairman:
United Kingdom Onshore Pipeline Operators Association
Pipeline Maintenance Centre
Ripley Road
Ambergate
Derbyshire
DE56 2FZ
e-mail:
enquiries@ukopa.co.uk


Issue: Final v1.0
Summary
This report presents collaborative pipeline and product loss incident data from
onshore Major Accident Hazard Pipelines (MAHPs) operated by National Grid,
Scotia Gas Networks, Northern Gas Networks, Wales & West Utilities, Shell UK
Limited (now Essar Oil (UK) Ltd), Shell EPE, BP, Ineos, SABIC and E-ON UK,
covering operating experience up to the end of 2010.
The data presented here covers reported incidents where there was an unintentional
loss of product from a pipeline within the public domain, and not within a compound
or other operational area.
The overall failure frequency over the period 1962 to 2010 is 0.234 incidents per
1000 km.year, whilst in the previous (Formal 6th) report this figure was 0.242
incidents per 1000 km.year (covering the period from 1962 to 2008).
The failure frequency over the last 20 years is 0.079 incidents per 1000 km.year.
For the last 5 years the failure frequency is 0.093 incidents per 1000 km.year, whilst
in the previous report this figure was 0.064 incidents per 1000 km.year (covering the
5 year period up to the end of 2008).
This report also presents data for part-wall damage and defects known as fault data,
and the statistical distributions derived for estimating pipeline failure probabilities due
to external interference events.


Issue: Final v1.0
Contents
1
INTRODUCTION.................................................................................1
1.1 BACKGROUND .................................................................................1
1.2 PURPOSE OF THE DATABASE ............................................................1
1.3 KEY ADVANTAGES ...........................................................................2
2 PRODUCT SYSTEM DATA ...............................................................3
2.1 EXPOSURE ......................................................................................3
2.2 TRANSPORTED PRODUCTS ...............................................................4
3 PRODUCT LOSS INCIDENT DATA ..................................................4
3.1 INCIDENT IGNITION ...........................................................................6
3.2 INCIDENT FREQUENCY .....................................................................6
3.2.1 Trends over the Past 5, 20 and 48 Years................................................. 6
3.2.2 Confidence Intervals................................................................................. 8
3.3
3.4
3.5
INCIDENT FREQUENCY BY CAUSE ...................................................10

GIRTH W ELD DEFECTS ..................................................................13
EXTERNAL INTERFERENCE .............................................................14
3.5.1 External Interference by Diameter Class................................................ 14

3.5.2 External Interference by Measured Wall Thickness Class...................... 15
3.5.3 External Interference by Area Classification........................................... 16
3.6
EXTERNAL CORROSION..................................................................17
3.6.1
3.6.2
3.6.3
3.6.4
External Corrosion by Wall Thickness Class .......................................... 17

External Corrosion by Year of Construction ........................................... 18
External Corrosion by External Coating Type......................................... 19
External Corrosion by Type of Backfill.................................................... 20
3.7 PIPELINE FAILURE CLASSIFIED AS OTHER .....................................21

3.8 PIPELINE FAILURES CAUSED BY INTERNAL CRACKING ......................22
3.9 DETECTION OF PIPELINE FAILURES .................................................23
4 FAULT DATA ...................................................................................24
4.1 PIPELINE DAMAGE DATA ................................................................24
4.2 PART-W ALL DEFECT DATA.............................................................25
4.3 STATISTICAL DISTRIBUTIONS OF DEFECT DIMENSIONS .....................26


Issue: Final v1.0
1 Introduction
1.1 Background
One of the key objectives of UKOPA is to develop a comprehensive view on risk
assessment and risk criteria as they affect Land Use Planning aspects adjacent to
high hazard pipelines. The main multiplier in pipeline risk assessments is the per unit
length failure rate which directly relates to the extent of risk zones adjacent to the
pipelines.
Regulators and consultants who carry out risk assessments for UK pipelines have
generally relied on US and European data to provide the basis for deriving failure
rates due to the shortage of verified published data relating to UK pipelines. UKOPA
published the first report in November 2000, presenting the first set of incident data
for pipeline incidents resulting in the unintentional release of product up to the end of
1998.
A full list of published reports is listed in the table below.
Report
Date
2000
2002
2003
2005
2007
2009
2010
2011
Type
of
Report
Formal
Formal
Formal
Formal
Formal
Formal
Interim
Formal
Covering
Incidents
to end of
1998
2000
2002
2004
2006
2008
2009
2010
Report
Number
Reference
1
2
3
4
5
6
7
8
R 4092
R 4798
R 6575
R 8099
6957
9046
UKOPA/2010/0074
UKOPA/2011/0076
1.2 Purpose of the Database

The purpose of the database is to:
record leak and fault data for UK Major Accident Hazard Pipelines
estimate leak and pipeline rupture frequencies for UK pipelines, based directly on
historical failure rate data for UK pipelines
provide the means to estimate failure rates for UK pipelines for risk assessment
purposes based on analysis of damage data for UK pipelines
provide the means to test design intentions and determine the effect of
engineering changes (e.g. wall thickness of pipe, depth of burial, diameter,
protection measures, inspection methods and frequencies, design factor etc.)

Page 1 of 27

Issue: Final v1.0
1.3 Key Advantages

The database is designed to reflect the ways in which the UKOPA operators design,
build, operate, inspect and maintain their pipeline systems. Although the pipeline and
failure data are extensive, there are pipeline groups (e.g. large diameter, recently
constructed pipelines) on which no failures have occurred; however, it is
unreasonable to assume that the failure frequency for these pipelines is zero.
Similarly, further pipeline groups exist for which the historical failure data are not
statistically significant.
Unlike its Europe-wide EGIG* counterpart, this UKOPA database contains extensive
data on pipeline failures and on part-wall damage known as fault data, allowing
prediction of failure frequencies for pipelines for which inadequate failure data exist.
Using Structural Reliability Analysis techniques it is possible to determine the range
of defect dimensions that will cause a specific pipeline to fail; analysis of the
statistical distributions of actual defect dimensions from the part-wall defect data
allows the probability of a critical defect to be determined and failure frequencies for
any credible failure mechanism to be calculated.
This approach has been used extensively and successfully by contributing
companies in pipeline uprating projects and assessing failure rates for quantified risk
assessments.
*European Gas Pipeline Incident Data Group (Gas loss incidents in gas transmission
pipelines operating above 15 bar).

Page 2 of 27

Issue: Final v1.0
2 Product System Data

2.1 Exposure
The total length of Major Accident Hazard Pipelines* in operation at the end of 2010
for all participating companies (National Grid, Scotia Gas Networks, Wales & West
Utilities, Northern Gas Networks, BP, Shell UK Limited [now Essar UK Ltd], Shell
EPE, Ineos, Sabic and E-ON UK) was 22,370 km. The total exposure in the period
1952 to the end of 2010 was 785,385 km.yr; the development of this exposure is
illustrated in Figure 1.
Exposure of Pipeline before first recorded incident in 1962 = 3740 km.yr (included in
exposure and incident frequency calculations)
Above Ground Pipelines are included in totals.
Figure 1
Development of Pipeline Exposure
800000
700000
Expo sure km.yr
600000
500000
400000
300000
200000
100000
0
1950
1960
1970
1980
1990
2000
2010
Year
*For definition of Major Accident Hazard Pipelines (MAHPs) - see UK statutory

legislation - The Pipelines Safety Regulations 1996 [PSR96], for the full definition
for natural gas the classification is above 8 bar absolute.

Page 3 of 27

Issue: Final v1.0
2.2 Transported Products

The lengths of pipeline in operation at the end of 2010, by transported product, are
(in km):
Table 1
Natural Gas (Dry)
Ethylene
Natural Gas Liquids
Crude Oil (Spiked)
Ethane
21,053
1,153
225.8
212.6
38.1
Propylene
Condensate
Propane
Butane
TOTAL
36.3
24.0
19.5
19.5
22,370
Note:- The database includes 550 km of decommissioned pipeline, 412 km that used
to transport natural gas, 92.4 km that used to transport ethylene, 36.3 km that used
to transport carbon monoxide, 4.8 km that used to transport propane and 4.8 km that
used to transport butane.
3 Product Loss Incident Data

A product loss incident is defined in the context of this report as:
an unintentional loss of product from the pipeline

within the public domain and outside the fences of installations
excluding associated equipment (e.g. valves, compressors) or parts other
than the pipeline itself
A total of 184 product loss incidents were recorded over the period between 1962
and 2010 compared with 179 product loss incidents documented in the report covering
the period to 2008. No product loss incidents were recorded prior to 1962. An annual
breakdown of incidents is illustrated in Figure 2.

Page 4 of 27

Issue: Final v1.0
Figure 2
Annual Number of Product Loss Incidents
Number of Incidents
16
14
12
10
8
6
4
2
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Year
Differences between 2008 and 2010 product loss statistics

Five product loss incidents were recorded in the last two years, two due to minor
external corrosion leaks, two leaks due to external interference, and one classified
as other
The cumulative number of incidents over the period 1962 to 2010 is shown in Figure
3.
Figure 3
Total Number of Product Lo ss Incidents (Cumulative)
180
160
140
120
100
80
60
40
20
0
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Number of Incidents
200
Year

Page 5 of 27

Issue: Final v1.0
3.1 Incident Ignition

There were 9 out of 184 (4.9%) product loss incidents that resulted in ignition. Table
2 below provides more detail:
Table 2 Incidents that Resulted in Ignition
Affected Component
Pipe
Pipe
Cause Of Fault
Seam Weld Defect
Ground Movement
Pipe
Pipe
Pipe
Pipe
Pipe
Bend
Bend
Girth Weld Defect

Unknown
Pipe Defect
Unknown
Lightning Strike
Internal Corrosion
Pipe Defect
Hole Diameter Class

0-6 mm
Full Bore and Above
(18 Diameter Pipe)
6-20 mm
6-20 mm
0 6 mm
40 110 mm
0-6 mm
0-6 mm
6-20 mm
3.2 Incident Frequency

3.2.1 Trends over the Past 5, 20 and 48 Years
The incident frequency over eight consecutive 5-year periods up to the end of 2010
is shown in Table 3.
Table 3
Period
Number of
Incidents
Total Exposure
[km.yr]
1966 - 1970
1971 1975
1976 - 1980
1981 - 1985
1986 - 1990
1991 - 1995
1996 - 2000
2001 - 2005
2006 2010
21
25
27
39
33
9
11
3
10
33,306
63,035
77,627
87,166
93,202
99,233
103,121
108,742
107,691

Frequency
[Incidents per 1000
km.yr]
0.631
0.397
0.348
0.447
0.354
0.091
0.107
0.028
0.093
Page 6 of 27

Issue: Final v1.0
The overall incident frequency by hole size over the period 1962 - 2010 is shown in
Table 4.
Table 4
Hole Size Class
Number of
Incidents
Full Bore* and Above

7
110mm Full Bore*
3
40mm 110mm
7
20mm 40mm
23
6mm 20mm
29
0 6mm
113
Unknown
2
Total
184
* Full Bore diameter of pipeline
Frequency
[Incidents per
1000 km.yr]
0.009
0.004
0.009
0.029
0.037
0.144
0.005
0.234
The total exposure for the last 20 years 1991-2010 is 418,717 km.years and the
resulting incident frequency is shown in Table 5.
Table 5
Frequency
Hole Size Class
Number of
Incidents
[Incidents per
1000 km.yr]
Exposure
Full Bore* and Above
110mm Full Bore*
40mm 110mm
20mm 40mm
6mm 20mm
0 6mm
Unknown
Total
1991-2010
0
1
1
6
4
21
0
33
418787
0.000
0.002
0.002
0.014
0.010
0.050
0.000
0.079
The failure frequency over the last 20 years is therefore 0.079 incidents per 1000
km.years and for the last 5 years (2006-2010) is 0.093 incidents per 1000 km.yr.
These compare with the failure frequency during the period 1962-2010 of 0.234
incidents per year per 1000 km.yr. An overview of the development of this failure
frequency over the period 1962 to 2010 is shown in Figure 3.
In order to see the results over recent periods, the moving average for each year is
calculated with reference to the incidents from the previous 5 years (2006-2010,
2005-2009, 2004-2008 etc.).

Page 7 of 27

Issue: Final v1.0
Figure 3
Development of Overall Incident Frequency
Frequency per 1000 km.yr
1.4
1.2
Overall Average up to Year

Moving 5-year Average
1
0.8
0.6
0.4
0.2
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Year
3.2.2 Confidence Intervals

Confidence intervals take uncertainty into account. The greater the exposure, the
smaller the confidence interval which shows that uncertainty decreases as more
operating experience is gained. To calculate the confidence intervals, the population
is assumed to have a known distribution.
Failure events generally follow a random distribution so it is assumed that a Poisson
distribution can be applied. The 95% confidence intervals for the overall average
failure frequency is shown in Figure 4 and for the 5-year average in Figure 5.

Page 8 of 27

Issue: Final v1.0
Figure 4
Overall Average with 95% Confidence Intervals
Frequency per 1000 km.years
1.400
1.200
1.000
0.800
0.600
0.400
0.200
0.000
Year
Figure 4 shows that the overall frequency for the whole period is 0.234 per 1000
km.years +/- 0.035.
Figure 5
Frequency pe r 1000 km.years
5-Yearly Average with 95% Confidence Intervals

1.4
1.2
1
0.8
0.6
0.4
0.2
2010
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
1972
1970
1968
1966
1964
1962
Year
Figure 5 shows that the 5-year average failure frequency for 2006-2010 is 0.093 per
1000 km.years +/- 0.059.

Page 9 of 27

Issue: Final v1.0
3.3 Incident Frequency by Cause

The development of product loss incident frequency by cause is shown in Figure 6.
Figure 6
Development of Incident Frequency by Cause
0.5
External Corrosion
Ground Movement
Girth Weld Defect
Pipe Defect
Unknown
0.45
0.4
External Interference
Internal Corrosion
Oth er
Seam Weld Defect
0.35
0.3
0.25
0.2
0.15
0.1
0.05
2010
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
1972
1970
1968
1966
1964
1962
Year
Table 6 Product Loss Incidents by Cause

Product Loss
Cause
Girth Weld Defect
External Interference
Internal Corrosion
External Corrosion
Unknown
Other
Pipe Defect
Ground Movement
Seam Weld Defect
Total
No. of
Incidents
34
40
2
37
7
41
13
7
3
184

Other Cause
Incidents
Internal cracking due to wet town gas
30
Pipe-Fitting Welds
4
Leaking Clamps
3
Lightning
1
Soil stress
1
Threaded Joint
1
Electric Cable Arc Strike
1
Total
41
Page 10 of 27

Issue: Final v1.0
Figure 7 shows the product loss incident frequency by cause over the period 19622010 compared with the frequency over only the last 5 years (2006-2010).
Figure 7
Historical and Recent Failure Frequencies
0.06
1962 to 2010
2006 to 2010
0.05
0.04
0.03
0.02
0.01
0
t
e
r
n
n
n
ct
ct
ct
w
en
io
io
he
fe
nc
fe
fe
s
t
s
o
e
e
e
e
m
n
O
o
D
ro
D
D
er
rr
ve
nk
e
rf
d
d
or
o
o
l
l
e
p
U
i
e
M
e
lC
lC
P
nt
W
W
lI
nd
na
na
h
a
r
m
r
t
u
n
ir
te
te
ea
ro
er
G
In
S
G
Ex
xt
E
Cause

Page 11 of 27

Issue: Final v1.0
An overview of the product loss incident frequency by cause and size of leak in the
period 1962 to 2010 is shown in Figure 8.
Figure 8
Prod uct Loss Incidents by Cause and Equivalent Hole Diameter
0.07
0 - 6 mm
6 - 20 mm
20 - 40 mm
40 - 110 mm
0.06
110 mm - Full Bo re
Frequency per 1000km.yr
0.05
Full Bore and Above
0.04
0.03
0.02
0.01
0.00
na
er
t
x
i
os
r
or
lC
on
te
Ex
al
rn
fe
er
t
In
n
re
ce
h
rt
el
W
ec
ef
D
d
un
o
r
t
en
em
ov
M
al
rn
e
t
In
n
io
s
ro
or
uc
tr
s
on
er
at
M
n/
it o
l
ia
e
th
O
no
nk
r/U
wn
Cause
Construction/Material = Seam Weld Defect + Pipe Defect + Pipe Mill Defect +

Damage During Original Construction

Page 12 of 27

Issue: Final v1.0
3.4 Girth Weld Defects

Figure 9 shows that 34 leaks due to girth weld defects were recorded in pipelines
constructed before 1980, 31 of which were before 1970. No further leaks due to this
cause have been observed since 1980.
Figure 9
Nu mber of Girth Weld Defects by Year of Construction
and Equivalent Hole Diameter
30
0 - 6 mm
Number Recorded
6 - 20 mm
25
20 - 40 mm
40 - 110 mm
110 mm - Full Bore
20
Full Bore and Above
15
10
0
Pre-1970
1971-1980
1981-1990
1991-2000
2001-2010
Year of Construction
The reduction in the number of girth weld defects in pipelines constructed after 1970
is associated with the improvements in field weld inspection and quality control
procedures, and the increasing capability of in line inspection tools to detect girth
weld anomalies.

Page 13 of 27

Issue: Final v1.0
3.5 External Interference

External interference is one of the main causes of product loss incident data with 40
recorded failures attributable to this cause.
3.5.1 External Interference by Diameter Class

Figure 10 shows the product loss incident frequencies associated with external
interference by diameter class and by hole size.
Figure 10
Product Loss Incidents Caused by External Interference
Frequency by Pipe Diameter and Equivalent Hole Diameter
0.20
0 - 6 mm
6 - 20 mm
0.18
0.16
20 - 40 mm
40 - 110 mm
0.14
110 mm - Full Bore

Full Bore and Above
0.12
0.10
0.08
0.06
0.04
0.02
0.00
0-4
5 - 10
12 - 16
18 - 22
24 - 28
30 - 34
36 - 48
Diameter Class (inches)
Table 7 Exposure by Diameter Class

Diameter
inches
Exposure
km.yr
Incidents
Frequency
/1000km.yr
0-4
5-10
39239
161610
5
19
0.127
0.118
12-16
18-22
131519
115165
9
3
0.068
0.026
24-28
127103
0.024
30-34
36-48
37942
172806
1
0
0.026
0.000
Total
785350
40
0.051

Page 14 of 27

Issue: Final v1.0
3.5.2 External Interference by Measured Wall Thickness Class

The relationship between product loss incidents caused by third party interference
and wall thickness is shown in Figure 11.
Figure 11
Frequency by Wall Thickness and Equivalent Hole Diameter
0.30
0 - 6 mm
6 - 20 mm
20 - 40 mm
40 - 110 mm
0.25
0.20
110 mm - Full Bore

Full Bore and Above
0.15
0.10
0.05
0.00
<5
5 - 10
10 - 15
>15
Wall Thickness Class (mm)
Note: Largest wall thickness for loss of product incident caused by external
interference to date is 12.7mm.
Table 8 Exposure by Wall Thickness Class
Wall
Thickness
mm
<5
Exposure
km.yr
Incidents
Frequency
/1000 km.yr
52576
12
0.228
5-10
378151
24
0.063
10-15
297274
0.013
>15
57384
0.000
Total
785385
40
0.051

Page 15 of 27

Issue: Final v1.0
3.5.3 External Interference by Area Classification

Figure 12
Frequency by Area Classification and Equivalent Hole Diameter
0.20
0 - 6 mm
0.18
6 - 20 mm
20 - 40 mm
40 - 110 mm
0.16
0.14
110 mm - Full Bore

Full Bore and Above
0.12
0.10
0.08
0.06
0.04
0.02
0.00
Rural
Suburban + Semi-Rural
Urban
Area Classification
Table 9 Exposure by Area Classification in km.yr

Area Classification
Rural
Suburban
Urban
Total
Exposure
Incidents
km.yr
711658
72722
1005
785385
29
11
0
40
Frequency
/1000
km.yr
0.041
0.151
0.000
0.051
Note:
Rural = population density < 2.5 persons per hectare
Suburban = population density > 2.5 persons per hectare and which may be
extensively developed with residential properties, and includes data classed as semirural
Urban = Central areas of towns or cities with a high population density

Page 16 of 27

Issue: Final v1.0
3.6 External Corrosion

3.6.1 External Corrosion by Wall Thickness Class
Figure 13
Product Loss Incidents Caused by External Corro sion
Frequency by Wall Thickness and Equivalent Hole Diameter
0.40
0 - 6 mm
0.35
6 - 20 mm
20 - 40 mm
0.30
40 - 110 mm
110 mm - Full Bore
Full Bore and Above
0.25
0.20
0.15
0.10
0.05
0.00
<5
5 - 10
10 - 15
>15
Wall T hickness Class (mm)
Table 10 Exposure by Wall Thickness Class

Wall
Thickness
mm
Exposure
km.yr
Incidents
Frequency
/1000 km.yr
<5
52576
20
0.380
5-10
378151
17
0.045
10-15
297274
0.000
>15
57382
0.000
Total
785385
37
0.047

Page 17 of 27

Issue: Final v1.0
3.6.2 External Corrosion by Year of Construction

Figure 14
Frequency by Year of Construction and Equivalent Hole Diameter
0.07
0 - 6 mm
6 - 20 mm
0.06
20 - 40 mm
40 - 110 mm
0.05
110 mm - Full Bore

Full Bore and Above
0.04
0.03
0.02
0.01
0.00
Pre-1975
1976-1985
1986-1995
1996-2005
2006-2010
Table 11 Exposure by Year of Construction

Construction Exposure
Year
km.yr
Incidents
Frequency
/1000 km.yr
Pre-1975
583858
36
0.062
1976-1985
142457
0.007
1986-1995
41115
0.000
1996-2005
15610
0.000
2000-2004
2345
0.000
Total
785385
37
0.047
The reduction in the number of incidents due to external corrosion for pipelines
constructed after 1976 is predominantly associated with the introduction of in line
inspection, which together with appropriate defect acceptance criteria, means that
metal loss defect are detected and repaired before developing to through wall.

Page 18 of 27

Issue: Final v1.0
3.6.3 External Corrosion by External Coating Type
Figure 15
Frequency by External Coating Type and Equivalent Hole Diameter
0.20
0 - 6 mm
6 - 20 mm
0.18
0.16
20 - 40 mm
40 - 110 mm
0.14
0.12
110 mm - Full Bore

Full Bore and Above
0.10
0.08
0.06
0.04
0.02
0.00
m
itu
B
r
Ta
l
oa
C
en
ly
Po
h
et
e
en
l
y
E
FB
er
th
/U
w
no
k
n
External Coating Type
Table 12 Exposure by External Coating Type
External Coating
Exposure
km.yr
Incidents
Frequency
/1000 km.yr
Bitumen
29090
0.103
Coal Tar
569393
25
0.044
Polyethylene
75292
0.047
FBE
74829
0.000
Other/Unknown
36780
0.163
Total
785385
37
0.047

Page 19 of 27

Issue: Final v1.0
3.6.4 External Corrosion by Type of Backfill

Figure 16
Frequency by Backfill Type and Eq uivalent Hole Diameter
40
0 - 6 mm
6 - 20 mm
35
20 - 40 mm
% of total Incidents
30
40 - 110 mm
110 mm - F ull Bore
Full Bore and Above
25
20
15
10
5
0
Clay
Gravel
Heavy Soil
Peat
Sand
Other
Backfill Type

Page 20 of 27

Issue: Final v1.0
3.7 Pipeline Failure Classified as Other

Pipeline failure rates due to causes other than those defined as
External interference
Corrosion
Material and construction
Ground movement (or other environmental load)
are generally classified as Other.

The UKOPA product loss data contains the following incidents under this category:Table 13 Pipeline Failures Classified as Other
Other Cause
Internal cracking due to wet town gas
Pipe-Fitting Welds
Leaking Clamps
Lightning
Soil stress
Threaded Joint
Electric Cable Arc Strike
Total
Incidents
30
4
3
1
1
1
1
41
The UKOPA product loss data indicates that Other causes account for
approximately 28% of the total failure rate.
88% (36 out of 41) of the incidents recorded in this category relate to pipelines
constructed before 1970, and are not relevant to pipelines designed, constructed and
operated in accordance with current pipeline standards.

Page 21 of 27

Issue: Final v1.0
3.8 Pipeline Failures Caused by Internal Cracking

A significant proportion of the failures classified as Other (30 out of 41 = 73%) were
caused by internal cracking (stress corrosion cracking [SCC]) in pipelines which had
seen wet towns gas (pre-natural gas) service. 90% of these failures (27 out of 30)
were in pipelines constructed before 1970.
Figure 17
Number of Failures caused by Internal SCC by Year of Construction
and Equivalent Hole Diameter
30
0 - 6 mm
6 - 20 mm
Number Recorded
25
20 - 40 mm
40 - 110 mm
110 mm - Full Bore
20
Full Bore and Above

15
10
0
Pre-1970
1971-1980
1981-1990
1991-2000
2001-2010

Page 22 of 27

Issue: Final v1.0
3.9 Detection of Pipeline Failures

Figure 18
Detection of Produ ct Loss Incidents by Eq uivalent Hole Diameter
40
0 - 6 mm
6 - 20 mm
% of total Incidents
35
30
20 - 40 mm
40 - 110 mm
25
110 mm - Full Bore

20
Full Bore and Above
15
10
5
0
P
s/
ip
C
ea
on
rs
G
un
ro
l
ro
t
a
r
ne
w
o
nd
a
L
LI
O
ak
Le
n
t io
c
e
et
D
e
ic
ol
n
or
ct
ow
a
n
P
k
tr
n
on
U
/
r
C
he
te
t
i
S
O
lic
ub
Detection Method
Note: Leak detection and On-Line Inspection (OLI) are not applicable to all pipelines.

Page 23 of 27

Issue: Final v1.0
4 Fault Data
4.1 Pipeline Damage Data
A Fault is a feature that has been confirmed by field investigation, excavation and
measurement. Any features that are inferred by other measurements such an
intelligent pig in line inspections, CIPS, etc. and have not been verified in the field
are not included in the UKOPA database. However pipeline defects comprising of
coating damage or grinding marks confirmed by field inspection are included.
The total number Faults recorded at the end of 2010 was 3080. The main causes of
the Faults are shown in Figure 19.
Figure 19
Fault Cause Classification
40
% of total Faults
35
30
25
20
15
10
5
0
t
e
t
e
n
ct
ct
ec
w
fe
ag
en
fe
ag
f
o
e
e
e
m
m
n
ro
D
D
D
am
e
ve
e
Da
or
D
nk
d
d
rf
on
l
l
o
i
e
l
p
U
C
n
e
e
/
s
il
M
l
Pi
nt
W
W
er
tio
M
ro
d
lI
r
h
na
c
h
n
e
t
r
m
a
t
o
u
e
p
n
ru
O
ir
ea
lC
Pi
xt
ro
G
er
st
S
t
a
E
G
n
o
rn
Ex
C
Fault Cause
te
n
I
e
nc
e
r
on
si
C
C
S
/

Page 24 of 27

Issue: Final v1.0
4.2 Part-Wall Defect Data

One of the main benefits of collecting Fault data is to record of the size of part-wall
defects which are measured and recorded in the database. Many faults have several
defects and as a result the database contained 5087 defects at the end of 2010.
Classification of defect data is shown in Figure 20.
Figure 20
Defect Type Classification
50
45
% of t ota l Defec ts
40
35
30
25
20
15
10
5
0
on
si
o
r
or
c
fe
de
ts
s
nt
e
D
s
ge
u
o
W
d
el
ts
ec
f
e
D
ec
ef
D
ill
M
ts
c
ra
C
ks
er
th
w
no
k
n
/U
Defect Type

Page 25 of 27

Issue: Final v1.0
4.3 Statistical Distributions of Defect Dimensions

Pipeline damage due to external interference occurs in the form of gouges, dents or
combinations of these. This type of damage is random in nature, and as operational
failure data are sparse, recognized engineering practice requires that a predictive
model is used to calculate leak and rupture failure frequencies for specific pipelines.
Predictive models such as those described in references (1,2,3) use dent-gouge
fracture mechanics models to predict the pipeline probability of failure, which is
dependent upon the pipeline geometry, material properties and operating pressure.
The UKOPA database includes reports of external interference incidents, including
the type of damage (dent, gouge and combinations of these), the size of the damage
and the number and location of the incidents. The external interference damage data
recorded up to and including 2010 in the UKOPA database has been analyzed to
determine the best fit Weibull distribution parameters for gouge length, gouge depth
and dent depth. The Weibull distribution parameters for the data are given in Table
14.
Table 14
Distribution
Parameters
Gouge Length
Gouge Depth
Dent Depth
Weibull Shape ()
0.575
0.666
1.028
Weibull Scale () mm
127.3
0.846
9.930
These parameters allow pipeline failure probabilities to be derived for external

interference events. An estimate of hit rate (i.e. frequency of damage incidents) is
also required to obtain pipeline frequencies to be calculated. Hit rate is dependent
on specific pipeline parameters including location (rural-suburban), depth of cover,
and frequency of external interference events for the pipeline population.
Note: Weibull distributions were identified as appropriate distributions in work carried
out to develop the FFREQ predictive model, which is recommended by UKOPA.
1 A Methodology for the prediction of Pipeline Failure Frequency Due to

External Interference. C Lyons, J V Haswell, P Hopkins, R Ellis, N
Jackson. IPC 2008-64375, 7th International Pipeline Conference, Calgary
2008.
2 Reduction Factors for Estimating the Probability of failure of Mechanical
Damage Due to External Interference. A Cosham, J V Haswell, N Jackson.
IPC 2008-64345, 7th International Pipeline Conference, Calgary 2008.

Page 26 of 27

Issue: Final v1.0
3 Modelling of Dent and Gouges, and the Effect on the Failure Probability of
Pipelines. P Seevam, C Lyons, P Hopkins, M Toft. IPC 2008-64061, 7th
International Pipeline Conference, Calgary 2008.
4 The Application of Risk Techniques to the Design and Operation of Pipelines. I
Corder. C502/016/95, Proceedings of International Conference on
Pressure Systems: Operation and Risk Management, Institution of
Mechanical Engineers, London, UK, p. 113-125. 1995.

Page 27 of 27

Title: UKOPA Pipeline Product Loss Incidents (1962-2010) : Report Number: UKOPA/11/0076 Issue: Final v1.0

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Title: UKOPA Pipeline Product Loss Incidents (1962-2010) : Report Number: UKOPA/11/0076 Issue: Final v1.0

Hochgeladen von

Copyright:

Verfügbare Formate

Report Number: UKOPA/11/0076

Issue: Final v1.0

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

UKOPA PIPELINE FAULT DATABASE

Pipeline Product Loss Incidents

Report Reference: UKOPA/11/0076

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

INCIDENT FREQUENCY BY CAUSE ...................................................10

3.5.1 External Interference by Diameter Class................................................ 14

External Corrosion by Wall Thickness Class .......................................... 17

3.7 PIPELINE FAILURE CLASSIFIED AS OTHER .....................................21

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

1.2 Purpose of the Database

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

1.3 Key Advantages

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

2 Product System Data

Expo sure km.yr

*For definition of Major Accident Hazard Pipelines (MAHPs) - see UK statutory

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

2.2 Transported Products

3 Product Loss Incident Data

an unintentional loss of product from the pipeline

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

Differences between 2008 and 2010 product loss statistics

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

3.1 Incident Ignition

Girth Weld Defect

Hole Diameter Class

3.2 Incident Frequency

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

Full Bore* and Above

Hole Size Class

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

Frequency per 1000 km.yr

Overall Average up to Year

3.2.2 Confidence Intervals

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

Frequency per 1000 km.years

Frequency pe r 1000 km.years

5-Yearly Average with 95% Confidence Intervals

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076

3.3 Incident Frequency by Cause

Frequency per 1000 km.yr

Table 6 Product Loss Incidents by Cause

Reproduction permitted with due acknowledgement

Report Number: UKOPA/11/0076