Skip to main content

Full text of "Pipeline performance in Alberta, 1990-2005"

See other formats


Is 



m 



Report 2007-A 



Pipeline Performance in Alberta, 1990-2005 

April 2007 



Alberta Energy and Utilities Beard 



ALBERTA ENERGY AND UTILITIES BOARD 

Report 2007-A: Pipeline Performance in Alberta, 1990-2005 



April 2007 



Published by 

Alberta Energy and Utilities Board 
640 - 5 Avenue SW 
Calgary, Alberta 
T2P 3G4 

Telephone: (403)297-8311 
Fax: (403) 297-7040 
E-mail: eub.infoservices@eub.ca 
Web site: www.eub.ca 



Contents 



1 Introduction 1 

2 Data Collection Parameters 2 

3 Conclusions 5 

3.1 Inventory Highlights 5 

3.2 Pipeline Incident and Performance Highlights 5 

4 Data Analysis 6 

Figures 1-29 7-63 

5 Future Initiatives 64 

6 Other Information 65 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • i 



Digitized 


bv 1 


the Internet Arch 






i 


in 2015 







https://archive.org/details/pipelineperformaOOalbe 



1 Introduction 



The Alberta Energy and Utilities Board (EUB) is responsible for the 
regulation of energy projects in Alberta. As energy industry-related 
pipelines are included in the EUB mandate, the EUB issues licences for 
pipeline construction and maintains an extensive database of licence 
information going back to the beginning of the energy industry in 
Alberta. Additional responsibilities of the EUB include inspection and 
surveillance of pipeline construction and operation; the EUB maintains 
records of these activities as well. 

At the end of 2005 there were over 377 000 kilometres (km) of energy- 
related pipelines in Alberta. This report provides a summary and analysis 
of that pipeline inventory, as well as an evaluation of the types and 
frequency of incidents and failures that occurred in relation to those 
pipelines during the period 1990-2005 inclusive. 

Pipelines under EUB jurisdiction include all pipelines except 

• low-pressure gas distribution network pipelines operating at 700 
kilopascals (kPa) or less (unless they are operated in connection with 
an EUB-licensed facility), 

water pipelines (unless they are operated in connection with an EUB- 
licensed facility), 
sewage pipelines, 

• pipelines under the jurisdiction of the National Energy Board, 

• pipelines situated wholly within a refinery, processing plant, 
marketing plant, or manufacturing plant, 

• pipelines carrying fuel from a tank and situated wholly within a 
consumer's property, and 

• pressure piping systems under the jurisdiction of the Safety Codes 
Act. 

Note that these exemptions are listed here in brief; any official 
interpretation should be taken from Sections l(l)(t) and 2 of the Alberta 
Pipeline Act. The Alberta Pipeline Act and Pipeline Regulation are the 
legislative standards applicable to the regulation of pipelines in Alberta; 
copies can be obtained from the offices of the Alberta Queen's Printer or 
downloaded from the EUB Web site www.eub.ca. 



The EUB and its precursor, the Energy Resources Conservation Board 
(ERCB), previously published pipeline performance data. Earlier 
publications were 

• An Analysis of Pipeline Performance in Alberta {Report D83-G, 
1983) 

• Pipeline Performance in Alberta (Report 91 -G, 199 1 ) 

• Pipeline Performance in Alberta (Report 98-G, 1998) 

Those reports provided summations of the amount of pipeline in Alberta, 
followed by analysis of pipeline incidents and of the failure performance 
of various types of pipeline infrastructure. This 2007 report follows the 
same general concept, but also includes a more in-depth assessment of 
the details of the pipeline infrastructure currently in place and an analysis 
regarding spill frequencies and spill volumes, along with 42 figures to 
depict this information. 

Inquiries regarding this report may be directed to the EUB Operations 
Group at (403) 297-8432. 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 1 



2 Data Collection Parameters 



To collect and compile data for this report, the EUB contracted the 
consulting firm of Visible Data, Inc., of Calgary, Alberta, which provides 
specialized data analysis services to the oil and gas industry. The EUB 
provided carefully defined data collection parameters and objectives to 
Visible Data, which then developed the search queries and compiled the 
resultant data. The various charts and graphs of the compiled data were 
developed by the EUB Communications Group. 

The evaluation period used in this report is from January 1, 1990, 
through December 3 1 , 2005. Data were obtained from two EUB 
databases, the Pipeline Attribute database and the Environmental 
Incidents database (now replaced by the Field Information System). Data 
were downloaded on March 1, 2006, in order to allow for data gathering 
and input to be completed for incidents and applications that occurred at 
the end of 2005. Both of the databases are live systems, meaning that 
they are constantly changing as licensing data are amended and updated 
and environmental incidents are updated. For example, pipelines are 
constantly being added, discontinued, abandoned, or returned to service. 
Thus the inventories are subject to daily variation. Incident data are also 
updated as investigation work is completed and new information 
obtained. 

The Pipeline Attribute database includes the basic licensing and physical 
information on all operating, discontinued, abandoned, and permitted 
pipelines. Permitted pipelines were included in the tally for the 2005 
year, as once the licence is issued the EUB does not know when 
construction is complete or when the pipeline is actually put into service. 
It has been necessary to assume that pipelines were commissioned in the 
year they were licensed. The total of pipeline recorded in the EUB 
database to the end of 2005 is 377 248 km, made up of 235 707 
individually identified pipeline segments. This includes 2363 km of 
pipeline (1131 segments) that have no dates specified. For the purposes 
of this report, these have been assumed to be old construction occurring 
prior to 1990, and therefore they are included in the tally of pipeline 
installed prior to 1990. 



2 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



We note that the data for cumulative length per year and length per 
product class, as collected for this report, do not agree consistently with 
length data collected in prior statistical studies. The reasons for this are 
complex, as data collection methods and computer platforms have 
changed considerably since data collection was originally begun in 1975. 
The most likely causes of discrepancies are licence dates that were 
changed at the time of future licence amendments and the actual 
relicensing of existing pipeline to different product categories or 
operating status. 

In gathering the length data for this report, a method was developed that 
examined the existing licensing data to estimate the construction year. 
The EUB Pipeline Attribute database contains four different date fields, 
which complicates determining the most likely construction date. For this 
purpose, the earliest date on record was generally used, but where 
recorded dates showed inconsistencies or overlap in the year, original 
construction date selection was made based on the following declining 
priority: Original Permit Approval Date, Permit Expiry Date, Permit 
Approval Date, and Last Occurrence Year. This was believed to provide 
the most likely construction date for each pipeline. Each individual 
pipeline segment was evaluated separately. 

In comparison to the previously published length data, discrepancies in 
the reported lengths become smaller moving towards recent time, and 
figures published in previous statistical reports since about 2001 are in 
reasonable agreement with the data collected for this report. The net 
realization of all this is that, despite best effort, the amount of pipeline 
attributed to a specific construction year is an estimate and becomes less 
certain as we look farther back into past years. 

Many EUB pipelines are licensed to carry multiple products, either due 
to production stream changes during the year (e.g., enhanced oil recovery 
schemes) or batch transmission of product. For the purposes of this 
report and the previous report in 1998, multiple substances are prioritized 
in accordance to the following declining order of priority: sour gas (SG), 
high vapour pressure product (HVP), crude oil (CO), oil effluent (OE), 
low vapour pressure product (LV), natural gas (NG), fuel gas (FG), salt 
(produced) water (SW), miscellaneous liquids (ML), miscellaneous gases 
(MG), fresh water (FW). (Note that in some instances in this report and 



in the figures oil effluent is sometimes called by its more colloquial 
name, multiphase.) Because of the categorization system, a batch 
products pipeline licensed to carry HVP, CO, and LV products will be 
captured as HVP in the summary of inventory. Based on length, the 
number of multiple product pipelines accounts for 5102 km, or 1.35% of 
the entire Alberta inventory; thus the statistical significance of the 
classification uncertainty is small. Of note is the observation that because 
most of these multiple substance pipelines are batch transmission 
pipelines carrying segregated product, most of them will have been 
captured in the HVP or CO substance categories. 

Pipeline diameter data show that for the smaller pipeline sizes, there 
were many recorded variations slightly off the nominal values. This is 
likely due to data entry errors, although pipes of older vintage did 
sometimes vary. It was therefore necessary to broaden the diameter 
ranges to capture similar-sized pipeline into nominal classes. For 
instance, a wide variety of pipe dimensions in the 1 to 2.5 inch ranges 
were found; these have all been categorized as 60.3 millimetre (mm) (2") 
pipe for simplicity. Similarly, 3.5" pipe has been included with 88.9 mm 
(3") pipe, 4.5" pipe has been included with the 1 14.3 mm (4") pipe, and 
5" pipe has been included with 168.3 mm (6") pipe. 

Material classes have been grouped together as well to simplify data 
analysis. In this report, materials are listed as steel, aluminum, 
composite, fibreglass, polyethylene, and other. Fibreglass denotes the 
traditional rigid stick-type fibreglass, and composite denotes the newer 
spoolable composite pipes. The "other" category includes asbestos 
cement, ductile cast iron, cellulose acetate butyrate, polybutylene, 
polyvinyl chloride (PVC), and unknown. There are fewer than 850 km of 
these "other" materials in aggregate. 

Any pipeline failure or hit upon a pipeline requires reporting to the EUB. 
Records of those pipeline incidents include a cause of failure. To 
simplify charting, these have been grouped into eleven classes of similar 
nature, as described in the following table. 

EUB legislation requires pipeline operators to report any pipeline 
incident that results in a loss of pipeline product, regardless of volume, or 
any incident where a pipeline was struck. This requirement is applicable 
regardless of the status of the pipeline; thus even hits on discontinued or 



Causes of Pipeline Failure by Class 



Report cause class 


Cause of failure in raw data 


Construction damage 


• Construction damage (improperly applied or 




damaged coatings, inadequate support, faulty 




alignment, bending) 


Damage by others 


• Damage by others (third-party excavation or 




interference) 


Earth movement 


• Earth movement (watercourse change, slope 




movement, heaves, subsidence) 


External corrosion 


• Corrosion, external 




• Mechanical pipe damage (dents, scrapes, 




gouges leading to corrosion) 


Internal corrosion 


• Corrosion, internal 


Joint failure 


• Mechanical joint failure (gasket or o-ring 




failure, internal joint coating failure, 




mechanical couplings) 




• Miscellaneous joint failure (butt fusion, 




interference joints, fibreglass bonded or 




threaded joints, explosive welding) 


Overpressure 


• Overpressure failure 


Pipe 


• Pipe failure (pipe body failure due to stress 




corrosion cracking [SCC], hydrogen-induced 




cracking [HIC], cracking, fatigue, laminations) 


Valve/fitting 


• Valve failure (seal blowouts, pig trap failures) 




• Installation failure (at compressor, pump, 




meter station) 


Weld 


• Girth weld failure (not by corrosion) 




• Seam rupture (electrical resistance weld 




[ERW] or other seam weld failure) 




• Other weld failure (weldolets, thermowells) 


Other 


• Installation failure (compressor, pump, or 




meter station) 




• Operator error 




• Unknown (pipe cannot be exposed/examined) 




• Miscellaneous (erosion, vandalism, lightning, 




flooding) 



abandoned pipelines must be reported. Spills occurring within facilities 
(such as satellites, batteries, or plants) are not considered to be part of the 
pipeline system and so would not be included in the pipeline incident 
data. 

Pressure test failures were not included in the operating incident data 
evaluation, as test failures do not occur under actual operating 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 3 



conditions. An analysis of pressure test failures has been conducted 
separately, and that information is also contained in this report. The 
number of pipeline incidents recorded in the reporting period was 
12 848, comprising 1 1 433 leaks, 758 ruptures, and 657 hits that did not 
result in product loss. The number of test failures for that period, not 
included in the 12 848, was 1256. 

In the environmental incident reporting database, there are only six 
available substance codes to choose from when an incident is recorded. 
They are CO, NG, OE, SG, WA (FW and SW), and OT (other). Only one 
substance code is recorded in the incident database: it is the first of the 
alphabetical series. Thus a failure on a CO/LVP/HVP pipeline when 
carrying HVP product would be recorded as a CO pipeline failure, 
whereas this same pipeline would have been counted as an HVP pipeline 
in the attribute database. A failure occurring on an HVP-only pipeline 
would be recorded as occurring in the "other" category. This mismatch 
between the two independent databases, though unfortunate, is not 
expected to result in significant misinterpretation, as the number of 
multiple substance pipelines is only 1.35%, based on length. 

Figures reporting data from pipeline spills indicate that there have been 
more spills than actual incidents. This is not an error, as some incidents 
release more than one substance (i.e., oil, water, and gas), each of which 
is recorded separately for spill measurement purposes. Spilled fluids are 
categorized into five classes: sour gas (contains greater than 10 moles 
hydrogen sulphide (H 2 S) per kilomole of natural gas), hydrocarbon gas, 
hydrocarbon liquid, water, and other. Bitumen, condensate, and HVP are 
included with the hydrocarbon liquids. Acid gas is included with sour 
gas. 

A measure of overall annual pipeline performance is calculated by 
dividing the number of failures recorded for each calendar year by the 
total kilometres of operating and permitted pipeline on record at calendar 
year-end. Discontinued and abandoned pipelines are not included in this 
calculation. There is of course some inaccuracy in the calculated values 
due to the consistently changing nature of the pipeline database, as 
previously discussed. 



4 * EU6 Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



The EUB database includes 51 16 km of pipeline licensed by the National 
Energy Board (NEB). Those pipelines are included in the total inventory 
as they do physically exist in Alberta. However, the EUB is not involved 
in the regulation of those pipelines and thus the EUB incident database 
has no record of a failure related to any of these licences. The specifics 
of the NEB pipeline inventory in Alberta is as follows: 



Substance 


Length (km) 


Crude oil 


1090 


Fuel gas 


12 


Natural gas 


2137 


Sour gas 


41 


Oil effluent 


5 


LVP 


55 


HVP 


1776 



NEB pipelines are those that cross a provincial or national border. In the 
case of the 5 1 16 km reported here, 90% of that length is the transmission 
pipelines of major pipeline operators. 

While the inclusion of this pipeline in the overall pipeline inventory will 
have a minor influence on the failure frequency numbers, the only 
substances for which this would be significant are natural gas, crude oil, 
and "other" (including LVP and HVP.) If the additional NEB mileage 
were removed, the average failure frequency of these three substances 
(for the year 2005 as an example) would increase from 1.61 to 1.74 for 
crude oil, from 1.55 to 1.57 for natural gas, and from 0.98 to 1.03 for 
"other." In respect to Figure 28, the difference would not be discernible. 



3 Conclusions 



3.1. Inventory Highlights 

Alberta's pipeline inventory continues to experience steady growth. As 
of year-end 2005, Alberta totalled 377 248 km of pipeline. For the 15 
years ending in 2005, the growth of pipeline infrastructure averaged 
6.2% per year, with the highest two being 8.6% growth in 1997 and 7.7% 
growth recorded in 2005. Natural gas pipelines make up the largest 
portion of the inventory, 62%. 

Operating pipelines (including new pipelines still licensed as 
"permitted") constitute 89.7% of the entire inventory, followed by 
abandoned pipeline, at 6.5%, and discontinued pipeline, at 3.8%. The 
great majority of all pipeline is constructed of steel, at 89.7%. The next 
largest material category is polyethylene, at 5.7%, followed by 
aluminum, at 2.2%, and fibreglass, at 1.9%. In terms of internal 
corrosion protection, 4.8% of pipeline contains some sort of corrosion 
barrier inside, and almost half of water pipelines contain some sort of 
internal corrosion barrier. Some types of pipeline, such as polyethylene, 
and the composite materials are inherently corrosion resistant. In 
comparison to EUB Report 98-G, this report shows significant growth in 
the amount of non-metallic materials being used for pipeline in recent 
years, as steel now makes up about 90% of the provincial total in 
comparison to the previous 94%. The use of advanced composite and 
polymer materials has great potential to reduce the number of corrosion- 
related pipeline failures. 

3.2. Pipeline Incident and Performance Highlights 

During the period 1990 to 2005, there were 12 848 pipeline incidents 
reported to the EUB (not including test failures). Of these, 657 were hits 
with no release, leaving 12 191 resulting in a pipeline release. Of all 
releases, 93.8% were leaks, and the other 6.2% were ruptures. A leak is 
defined as a situation where a pipeline may be losing product but 
continuing to operate. A rupture is a situation where the pipeline has 
been compromised to the point where it cannot continue to operate. The 
clear decline in the number of pipeline ruptures over the last eight years 
is a positive trend. The number of pipeline failures related to each 
product classification has been relatively steady, although 2005 shows an 



increase in the number of pipeline failures occurring in natural gas 
pipelines. 

Internal corrosion continues to be the most prevalent cause of pipeline 
failure during the reporting period, representing 57.7% of all releases, 
followed by external corrosion at 12.0%. The combined total of 69.7% is 
a little higher than the total for corrosion presented in Report 98-G, 
which indicated 51.2% and 13.3% respectively, for a combined total of 
64.5%. 

Of all pipeline releases, 90.5% occurred on pipelines 60.3 mm (2") to 
168.3 mm (6") in diameter, the most common pipelines used in Alberta's 
oil and gas gathering fields. Similarly, it was determined that 96% of all 
reported pipeline spills were of less than 100 m 3 of liquid or 100 000 m 3 
of gas, again reflecting the prevalence of small-diameter pipelines. Of the 
reported spills or releases, 29.8% of recorded releases were hydrocarbon 
liquid, 23.6% hydrocarbon gas, 44.4% water, 1.9% sour gas, and 0.3% 
other materials. 

When reviewing the average frequency of pipeline incidents, it is found 
that while the last ten years have shown a steady, gradual decrease in the 
number of incidents occurring per 1000 km of installed pipeline, 
reaching a low of 2.2 in 2004, the year 2005 saw a slight upward bump, 
mostly attributable to a higher number of natural gas pipeline failures, 
bringing the incident rate for 2005 to 2.4 incidents per 1000 km. Some of 
the product classes have quite low, steady failure frequency rates, 
suggesting that it may be difficult in some cases to realize much further 
improvement with incremental effort. However, the EUB is concerned 
about the increasing number of "other" failure causes being reported. 
"Other" includes installation failures (at a compressor, pump, or meter 
station,) operator error, miscellaneous (erosion, vandalism, lightning, 
flooding), and unknown (pipe cannot be exposed for investigation.) In 
years past only a few failures have been attributed to these causes, and 
the EUB is concerned that the recent increase in number of "other" 
failures may be serving to mask other data trends related to cause. 

There were 1311 pipeline releases resulting from pressure testing and 
requalification testing. Of these, 85.8% were plain water; the remainder 
were various fluids. 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 5 



4 Data Analysis 

The following section contains 42 figures, along with brief 
interpretations and commentary, prepared by senior members of the 
Compliance and Operations Branch, Pipeline Section. 

Figure 1 shows that as of December 31, 2005, there were 377 248 km of 
EUB-licensed pipeline in Alberta. This inventory includes 235 707 
individually identified segments, with the simple average length of a 
pipeline segment being 1.60 km. This may seem counterintuitive when 
considering that pipelines are generally envisioned as long cross-country 
facilities, but in fact it reflects the reality that in Alberta the majority of 
pipelines are indeed short segments, 1 or 2 km in length, leading from 
individual producing wells to gathering facilities for treatment or 
processing. Larger group pipelines then carry the commingled 
production away, but these are much fewer than the small-diameter flow 
lines. The largest portion of the total inventory is natural gas pipeline, 
comprising about 62% of the inventory. Under EUB licensing protocol, 
natural gas pipelines may not contain more than 10 moles of H 2 S per 
kilomole of natural gas, equivalent to 1%, or 10 000 parts per million. 
Pipelines having gas containing greater than 10 moles of H 2 S per 
kilomole of natural gas are licensed as sour gas pipelines. 

Pipeline classification generally follows the designation of a producing 
well. Thus gas pipeline is associated with a gas well. There may also be 
hydrocarbon liquids and water associated with the gas well. Where the 
composition of produced fluids causes the well to be classified as an oil 
well, the associated pipeline is licensed as a multiphase pipeline. Crude 
oil pipelines are generally considered to be carrying treated product 
where initial processing has removed gas and water. Water pipelines are 
those used for water supply or collection and for water injection. Sour 
gas pipelines are those where the H 2 S content exceeds 10 moles of H 2 S 
per kilomole of natural gas. They could also contain water and 
hydrocarbon liquids. Pipelines classified as "other" carry HVP products, 
such as ethane, propane, butane, ethylene, and mixes of produced natural 
gas liquids, as well as LVP products, such as fuel oil, motor fuel, and 
condensate. Additional pipelines contained in the "other" classification 
include hydrogen, carbon dioxide, nitrogen, ammonia, polymer, sulphur, 
etc. 



6 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



For the 15 years ending in 2005, the annual increase in pipeline 
infrastructure averaged 6.2% per year, with the highest two being 8.6% 
growth in 1997 and 7.7% growth recorded in 2005. 



Average Length of Pipeline Segments 

All pipelines current to December 31 , 2005; includes abandoned, discontinued, 
operating, and permitted lines 





Auorano lonnth nor 
rWcidyc iciiyui pel 

segment (km) 


Tntal lonnth 
1 Uldl Icllylll 

(km) 


Mi imhor nf 

segments 


Crude oil 


3.74 


18019 


4812 


Natural gas 


1.73 


235592 


136023 


Sour gas 


2.45 


20168 


8244 


Fresh water 


1.47 


6445 


4397 


Salt/produced water 


1.09 


14403 


13212 


Oil well effluent 


0.88 


50977 


57790 


Fuel gas 


1.71 


12839 


7493 


HVP 


6.77 


11880 


1754 


LVP 


6.33 


5432 


858 


Miscellaneous gases 


0.98 


756 


775 


Miscellaneous liquids 


2.12 


739 


349 


Totals 




377248 


235707 


Average length 


1.60 







Figure 1. Length of pipelines in Alberta by year and substance 

All pipelines current to December 31 , 2005 
400 000 



% OF ENTIRE INVENTORY El 






1990 


1991 


1992 


1993 


1994 


1995 


1996 


1997 


1998 


1999 


2000 


2001 


2002 


2003 


2004 


2005 


Crude oil 


10 284 


10 829 


11 280 


11 749 


12106 


12 606 


12 853 


14 442 


15 340 


16137 


16423 


16 703 


17 213 


17 568 


17 698 


18019 


Natural gas 


86 501 
4 948 
10 926 


90 060 
5164 
11416 


93 472 
5 325 
11850 


99 981 
5 821 
12 394 


108 517 
6114 
13113 


115102 
6 769 
13812 


122 505 
7 747 
14421 


132 192 
8 848 


141 010 
10 442 
15 825 


149 712 
11740 
16 346 


161 522 
12 957 
16 941 


175 396 
14 374 
17 750 


186 227 
15 545 
18 364 


200 250 
16 709 
19 235 


215 903 

18 562 

19 984 


235 592 
20168 
20 847 


Water 


15164 




25 552 


26 543 


27 598 


29 275 


31 102 


32 880 


34 655 


36 844 


38 383 


39 470 


41 075 


43 785 


45 309 


46 911 


48 594 


50 977 


Other 


13 761 


14 084 


14 446 


15 042 


15 502 


16 280 


17154 


19 886 


22 576 


23 846 


25189 


26 534 


27 267 


28 320 


29 444 


31 645 


Total (km) 


151 973 


158 095 


163 970 


174 262 


186454 


197 449 


209 335 


227 376 


243 577 


257252 


274 108" 


294 542 


309 925 


328 993 


350 186 


377 248 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 7 



Figure 2 shows that as of the end of December 2005, about 89.7% 
(338 448 km) of the Alberta pipeline inventory was listed as operating 
(or permitted, as permitted and operating are combined, as permitted 
pipeline automatically converts to operating after one year), 3.8% 
(14 371 km) was listed as discontinued, and 6.5% (24 430 km) as 
abandoned. These numbers fluctuate, since pipeline status is being 
changed daily as companies modify or terminate projects. Changes to the 
Pipeline Regulation in 2005 have also led to additional amendments to 
pipeline status, as the regulation required licensees to properly 
discontinue and abandon pipelines that had been existing in an unused 
state or to include pipeline that is not flowing but licensed as operating 
within the licensee's corrosion mitigation program. 



8 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 2. Length of pipelines in Alberta by status and substance 

All pipelines current to December 31, 2005 




i.5%) 





OPERATING (including Permitted) 


DISCONTINUED 


ABANDONED 




%of 






%of 




%of 




%of 


Total 


entire 




km 


product type 


km 


product type 


km 


product type 


km 


inventory 


Crude oil 


14 902 


82.7 


1511 


8.4 


1606 


8.9 


18019 


4.8 


Natural aas 


223 921 


95.1 


4082 


1.7 


7588 


3.2 


235 592 


62.5 


Sour qas 


18 120 89.8 


1067 


5.3 


982 


4.9 


20168 


5.3 




14 463 


69.4 


1502 


7.2 


4882 


23.4 


20 847 


5.5 




38 536 


75.7 


43 562 


8.6 


8058 


15.8 


50 977 


13.5 


Other 


28 479 


90.0 


1853 


5.9 


1313 


4.2 


31 645 


8.4 


Total 


338 448 


89.7 


14 371 


3.8 


24 430 


6.5 


377 248 


100.0 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 9 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 3 shows two concepts: how much pipeline of each material class 
exists and what services those materials are used in. The second concept 
shows the overall use of each material in proportion to each other and the 
whole. 

The six small pie graphs show the six primary pipe material classes: 
steel, aluminum, fibreglass, composite, polyethylene, and other. Each of 
these graphs shows in what service that pipeline material is used; for 
example, the great majority of aluminum pipeline is used in natural gas 
service. The predominant use of fibreglass pipeline is in multiphase 
service. 

The large pie shows the proportion of each material in use overall. The 
predominant pipeline material is steel, which has been used in almost 
90% of the total pipeline infrastructure, compared to 94% stated in 
Report 98-G. Thus there has been meaningful growth in the use of 
polymeric and composite pipeline as an alternative to steel. With 
improved technology allowing higher working pressures for both 
polymeric and composite pipeline and the inherent corrosion resistance 
of polymers, industry is choosing to use these products more frequently. 



Figure 3. Installed pipelines by pipe material 

All pipelines current to December 31 , 2005 



and substance 




Composite Polyethylene Other Material totals 

1095 km 21417 km 839 km 377 248 km 



PIPE MATERIAL 



SUBSTANCE 


Steel 




Alum 


tnum 


i Fibre 


glass 


I Composite 


Polyethylene 


Otr 


er 


Total 


%of 


CARRIED 


km 


% 


km 


% 


km 


% 


km 


% 


km 


% 


km 


% 


km 


Inventory 


Crude oil 


17 949 


99.6 


39 


0.2 


22 


0.1 


1 


<0.1 






8 


<0.1 


18019 


4.8 


Natural gas 


209 437 


88.9 


7431 


3.1 


200 


<0.1 


611 


0.3 


17 871 


7.6 


41 


<0.1 


235 592 


62.4 


Sour gas 


20135 


99.8 


24 0.1 














10 


<0.1 


20168 


5.3 




16 921 


81.2 


199 


0.9 


2093 


10.0 


151 


0.7 


896 


4.3 


588 


2.8 


20 847 


5.5 


45 471 


89.2 


50 


0.1 


4939 


9.7 


316 


0.6 


63 


0.1 


138 


0.3 


50 977 


13.5 


[Other 


28 549 


90.2 


395 


1.3 


45 


0.1 


15 


<0.1 


2 586 


8.2 


54 


0.2 


31 645 


8.4 


Total 


338 461 


89.7 


8139 


2.2 


7298 


1.9 


1095 


0.3 


21 417 


5.7 


839 


0.2 


377 248 


100.0 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 11 



Figures 4a-4g show how much pipeline is installed for each product class 
and what sizes of pipe are used. Figure 4a shows how the large majority 
of pipeline in Alberta is small diameter. Figure 4b, on natural gas, shows, 
for example, that pipe of 2" to 6" nominal size makes up a total of 
194 791 km of the entire 235 592 km of natural gas pipeline. The length 
of other sizes of pipe can also be determined from the chart. To clarify 
graphing, some of the sizes have been combined into one colour band. 

The tables in Figures 4c through 4g provide the data for each of the other 
product classes. 

A common trend for natural gas, oil effluent, water, and sour gas 
pipelines is that pipe sizes used are typically small. Larger pipe sizes 
become more predominant in the crude oil and "other" product 
categories, as they both contain a significant amount of transmission 
pipeline for treated or refined product. There is a significant amount of 
very large pipeline inventory shown in the natural gas category, which of 
course reflects the major natural gas transmission pipeline systems in 
Alberta. 



12 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 4a. Total installed pipelines by pipe size 

All pipelines current to December 31 , 2005 



400 000 



350 000 




300 000 



Sr 250 000 



200 000 



150 000 



100 000 



50 000 




114.3 mm (4" 
98170 




PIPELINE SIZE 



mm (inches) 


km 


273.1 (10) 


! 11 793 


323.9 


(12) 


11 583 


355.6 


(14) 


475 






7 684 






606 
3 808 
586 


610.0 


(24) 


4 830 


660.0 


(26) 


61 


711.0 


(28) 




762.0 


(30) 


2 686 


813.0 


(32) 


8 


864.0 


(34) 


781 


914.0 


(36) 


2 075 


1 067.0 


(42) 


1989 


1 219.0 


(48) 


714 


1 524.0 


(60) 


1 



Total all sizes 377 249 



All 



Figure 4b. Installed pipelines by pipe size and substance (natural gas) 

All pipelines current to December 31 , 2005 



250 000 



60.3 mm (2") 
24 726 



200 000 




Natural gas 



Figure 4c. Installed pipelines by pipe size and substance (crude oil) 

All pipelines current to December 31 , 2005 



20 000 



60.3 mm (2") 
367 



9 mm (3") 
1629 



15 000 



114.3 mm (4" 
3 317 



10 000 



PIPELINE SIZE 

mm (inches) 




km 
1 590 



7 

1 023 
435 

m 



129 



18 019 



660.0 (26) 

711.0 (28) 

762.0 (30) 

813.0 (32) 

864.0 (34) 

914.0 (36) 

1 067.0 (42) 

1 219.0 (48) 

1 524.0 (60) 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 13 



Figure 4d. Installed pipelines by pipe size and substance (sour gas) 

All pipelines current to December 31, 2005 



25 000 



60.3 mm (2") 
175 



20 000 



■ II ■ ■■ hum 

88.9 mm 



2 232 



(3") 



PIPELINE SIZE 



323.9 


(12) 


355.6 


(14) 



km 
924 
120 
295 
106 



Total all sizes 20 168 



114.3 mm (4") 
5 673 



168.3 mm (6") 
5 827 



219.1 mm (8") 
3 683 



273Tirim(iOTl 
1126 I 



14 * EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 4e. Installed pipelines by pipe size and substance (water) 

All pipelines current to December 31 , 2005 



25 000 




Figure 4f. Installed pipelines by pipe size and substance (multiphase) 

All pipelines current to December 31 , 2005 



60 000 



50 000 



60.3 mm (2") 
8 654 



40 000 

PIPELINE SIZE 
mm (inches) km 




U|gjM| Total all sizes 50 977 

0 

Multiphase 



Figure 4g. Installed pipelines by pipe size and substance (other) 

All pipelines current to December 31 , 2005 



40 000 



30 000 




| 20 000 

"53 

E 



.9 mm (3"' 
4 515 




168.3 mm (6" 
3 035 



10 000 




PIPELINE SIZE 



mm 


(inches) 


km 


323.9 


(12) 


3 010 


355.6 


(14) 


12 






1 504 






5 




580 


610.0 


(24) 


605 


660.0 


(26) 




711.0 


(28) 




762.0 


(30) 


604 


813.0 


(32) 


8 


864.0 


(34) 


345 


914.0 


(36) 




1 067.0 


(42) 




1 219.0 


(48) 


76 


1 524.0 


(60) 





Total all sizes 



31 645 



Other 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 15 



Figure 5 shows how much pipeline in Alberta is constructed of steel, as 
well as how much of those steel pipes have a fixed internal corrosion 
barrier installed. The large pie chart in Figure 5 shows the relative 
proportion of each substance class that is carried in the steel pipeline 
infrastructure. 

The categories of internal corrosion protection are none (bare pipe), thin 
film (baked-on polymer coatings), slip lined (loose-fit liners or pipe), 
polyvinyl chloride (liners), cement (bonded lining), expanded (tight-fit 
plastic liner), and other. Data show that 94.8% (320 977 km) of steel 
pipeline contains no internal corrosion barrier. The six small pie charts in 
Figure 5 show the amount of each corrosion prevention method used in 
each substance class. 

The most successful of the various types of corrosion barrier have been 
expanded plastic liners and freestanding plastic or composite pipe. In 
some systems, cement lining and thin-film lining have also worked 
successfully, but these systems must be installed and operated carefully 
for sustained performance. 

Although the majority of pipe is shown as having no internal corrosion 
barrier, this does not suggest that this pipe does not have internal 
corrosion prevention. The most common method used to protect steel 
pipelines against corrosion is the application of filming corrosion 
inhibitors (chemicals), accompanied by regular pipeline pigging 
(cleaning) to remove water and contaminants. This method is widely 
used and can be very effective when properly implemented. 



16 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 5. Types of internal corrosion prevention installed in steel pipelines in Alberta's pipeline inventory 

All pipelines current to December 31 , 2005 




Water Multiphase Other 

16 921 km 45 471 km 28 549 km 



CORROSION PREVENTION 



SUBSTANCE 
CARRIED 


Cement 


km 


% 


None 
km 


% 




% 




Slip lined 
km % 


km 


% 


Total 
km 


%of 
Inventory 


Crude oil 


1 <0.1 


46 


0.3 


17 801 


99.2 


20 


0.1 






49 


0.3 


33 


0.2 


17 949 


5.3 


Natural qas 


74 


<0.1 


266 


0.1 


205 976 


98.4 


1 849 


0.9 






264 


0.1 


1006 


0.5 


209 437 


61.9 


Sourqas 


2 


<0.1 


163 


0.8 


19 893 


99.0 


7 


<0.1 






59 


0.3 


11 


0.1 


20135 


6.0 




1428 


8.4 


3113 


18.4 


7624 


45.0 


2198 


13.0 


40 


2.0 


483 


2.9 


2 034 


12.0 


16 921 


5.0 




178 


0,4 


1713 


3.8 


41430 


91.1 


704 


1.6 


7 


<0.1 


644 


1.4 


794 


1.8 


45 471 


13.4 


Other 


1 


<0.1 


37 


0.1 


28 254 


99.0 


158 


0.5 






30 


0.1 


68 


0.2 


28 549 


8.4 


Total 


1684 


0.5 


5 339 


1.6 


320 977 


94.8 


4935 


1.5 


47 


<0.1 


1530 


0.4 


3 948 


1.2 


338 461 


100.0 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 17 



The data presented in Figure 6 show that the number of recorded pipeline 
incidents has been fairly stable over the reporting period, typically 
around 800 annually. These data exclude test failures, as they do not 
occur under normal pipeline operating conditions. As pipeline 
infrastructure has increased about 6.2% per year over the last 15 years, it 
is encouraging to see that the number of incidents has not similarly 
increased. Looking at incidents per installed length, pipeline incident 
frequency has actually decreased over the last several years. This 
conclusion is demonstrated clearly in Figures 28 and 29. 

There has been a noticeable decrease in the number of pipeline ruptures 
over the last 15 years. This suggests that the pipeline industry is 
successfully monitoring pipelines for problems, anomalies, and situations 
that might lead to catastrophic failure. There have been steady advances 
in technologies for internal pipeline inspection and in predictive 
modelling that assists in the interpretation of when pipeline corrosion or 
other defects might deteriorate into unsafe conditions. The data suggest 
that these tools are being applied successfully. The small pie chart in 
Figure 6 shows that of all incidents, 89% resulted in leaks, 5.9% resulted 
in ruptures, and 5.1% were hits with no product loss. However, those 
numbers are average values for the entire period of the report, and it is 
obvious that the frequency of rupture has been considerably less than 
5.9% in the last few years; in fact, in 2005 ruptures represented only 
1.2% of all incidents. 

The last three years of the reporting period show an increase in the 
number of hits on buried pipeline. This is disappointing, but not 
unexpected in view of the very robust oil and gas development in the last 
few years. Increased field activity increases the chance of unplanned 
pipeline contact, and the increasing presence of a "green" or 
inexperienced junior workforce may be impacting this negatively. In 
2005 the Pipeline Regulation strengthened some of the requirements 
regarding safe ground disturbance practices, inspection, and training, and 
it is hoped that in subsequent years these enhancements will translate 
into a reduction in the annual number of hits. 

A simple analysis was done to see whether pipelines of certain product 
classes experienced more hits than others. By dividing the percentage of 
incidents caused by "damage by others" for each product class by the 



18 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



percentage of inventory that each class represented, a simple ratio was 
developed. For example, if a certain product pipeline represented 5% of 
the total pipeline inventory but experienced 10% of the number of 
"damage by other" incidents, the ratio was 2. The results are as follows: 



Substance 


Percentage of all 
incidents due to 
damage by others 


Percentage of total 
inventory 


Ratio 


Crude oil 


7.27 


4.78 


1.52 


Multiphase 


31.18 


13.51 


2.31 


Natural gas 


45.80 


62.45 


.73 


Sour gas 


2.45 


5.35 


.46 


Water 


7.84 


5.53 


1.42 


Other 


5.47 


8.39 


.65 



This indicates that the more hazardous product classes, such as sour gas, 
natural gas, and "other" (which often contains HVP or LVP products), 
are experiencing fewer hits than the potentially lower risk product 
classes. This could suggest that shortcuts are being taken around certain 
types of pipeline. The relatively low ratio for natural gas pipelines, which 
are the most prolific substance class, suggests that mere abundance of 
pipeline does not necessarily translate into a higher number of hits. 
Industry should examine these data and consider why the differences 
exist. On a positive note, industry does seem to be exhibiting better care 
than "average," that is, a ratio of 1, when excavating or working around 
pipelines carrying the most hazardous products. 



Figure 6. Pipeline incidents, by type of incident per year 

All pipeline incidents from January 1, 1990, to December 31, 2005 (includes all leaks, ruptures, and hits [did not result in product loss]) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 19 



Figure 7 shows how the raw number of incidents for each product class 
has changed from year to year. A decrease in the number of water 
pipeline failures from 1990 is apparent, though the annual failure number 
seems to be quite static for the last several years. The only other 
significant trend shows that the number of natural gas pipeline failures 
declined following an increase in 1999 and 2000, but has recently 
increased again. 



20 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 7. Pipeline incidents, by product per year 

All pipeline incidents from January 1, 1990, to December 31, 2005 (includes all hits, leaks, and ruptures) 



1000 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 21 



Figure 8a. Total number of releases by cause per year 

All pipeline releases from January 1, 1990, to December 31, 2005 (leaks and ruptures only) 




22 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



During all the years the EUB has tracked failure 
data, the predominant cause of pipeline failure 
has been internal corrosion, followed by external 
corrosion. This is not surprising, as most of 
Alberta's pipeline infrastructure is used for the 
production of raw oil and gas, which by nature 
can be highly corrosive. As pipes are mainly 
steel and are buried underground, corrosion of 
the external surfaces is also possible. Through 
regulatory direction and targeted surveillance 
and inspection, the EUB has drawn attention to 
these issues and encouraged industry to improve 
its operating practices. Figure 8a shows that for 
the years 2000 through 2005, these efforts may 
be having an impact, as the raw number of 
internal and external corrosion events has 
declined. However, offsetting this decline seems 
to be a pronounced increase in the number of 
"other" failures over the last three years. This 
raises the question of whether the cause of some 
of these "other" failures has been properly 
determined. 

Figure 8b provides an averaged representation of 
the failure causes for leaks and ruptures over the 
reporting period. Internal and external corrosion 
constituted 69.7% of all failures, an increase 
over the 63% reported in Report 98-G. Given 
that over recent years the total number of 
pipeline releases has been relatively steady, this 
suggests that industry is being more successful 
at reducing failures of the other various types 
than it is at reducing failures due to corrosion. 
However, on a positive note, the overall failure 
frequency has been declining during this same 
time. 



Figure 8b. Pipeline releases, by cause for all years combined 

All pipeline releases from January 1, 1990, to December 31, 2005 (leaks and ruptures only) 




Total number of releases 12 191 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 23 



Figure 9 shows that the great majority of pipeline incidents occur on 
small diameter pipeline, which reflects the actual infrastructure present in 
Alberta and the corrosive nature of the products carried in those small 
diameter pipelines. Averaged over the analysis period, 90.5% of all 
pipeline incidents occurred on pipe of 168.3 mm (6") diameter and 
smaller. 



24 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 9. Pipeline incidents, by pipe size 

All pipeline incidents from January 1 , 1990, to December 31 , 2005 (includes all hits, leaks, and ruptures) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 25 



Figure 10a shows that 66.2 % of water pipeline failures are caused by corrosion. Progress has been made in reducing the number of water pipeline failures, 
as shown in Figure 10b, although the raw number is relatively steady in recent years. Almost half of water pipelines do have internal corrosion protection, 
as shown in Figure 5. 

Figure 10a. Water pipeline incidents, by cause for all years combined 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 




Total number of incidents 3332 (100%) 



26 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 10b. Water pipeline incidents, by cause per year 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 



350 




CAUSE 


1990 


1991 


1992 


1993 


1994 


1995 


1996 


1997 


1998 


1999 


2000 


2001 


2002 


2003 


2004 


2005 


Internal corrosion 


251 


239 


178 


156 


140 


140 


149 


124 


110 


85 


81 


68 


84 


65 


64 


73 


External corrosion 


7 


9 


23 


11 


15 


8 


11 


6 


13 


15 


9 


9 


17 


17 


20 


11 


Damage by others 


2 




4 


2 


7 


8 


5 


5 


7 


7 


6 


11 


12 


4 


9 


7 


Weld 


8 


12 


15 


10 


11 


9 


13 


13 


8 


7 


9 


11 


5 


7 


4 


8 


Construction damage 


13 


5 


14 


13 


8 


7 


11 


12 


9 


14 


12 


23 


8 


13 


10 


16 




2 
6 


1 

7 


1 
4 


1 
6 


2 
6 


1 
4 


1 
8 


1 
5 


i i 
9 5 


9 


2 
12 


2 
8 


13 


1 

16 


8 




13 


8 


13 


10 


10 


7 


10 


8 


17 


15 


13 


14 


29 


11 


9 


2 


Earth movement 


9 


1 




1 


7 


1 


1 


7 


3 


3 


3 


3 


2 


3 


2 


3 




3 


2 


3 


2 


2 


4 


6 


4 


6 


3 


4 


3 


2 


1 


10 


8 




24 


20 


24 


21 


16 


18 


11 


15 


14 


11 


7 


11 


6 


12 


13 


22 


Total number incidents 338 
% of total 10.2 


304 
9.1 


279 
8.4 


233 
7.0 


224 

6.7 


207 
6.2 


226 
6.8 


200 
6.0 


197 
5.9 


167 
5.0 


153 
4.6 


167 
5.0 


175 
5.3 


146 
4.4 


158 
4.7 


158 
4.7 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 27 



Figure 1 la shows that 69.2% of multiphase pipeline failures are caused by corrosion. It can be hard to mitigate corrosion in multiphase pipelines, as 
variations in water content make it difficult to select and maintain effective inhibitor films. There appears to be a relatively high number of external 
corrosion failures as well. This is likely due to exterior coating damage, as some of these systems operate at higher temperatures where coating systems 
tend to deteriorate over time. Figure 1 lb shows that in recent years operators seem to be reducing the number of corrosion failures. 



Figure 11a. Multiphase pipeline incidents, by cause for all years combined 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 




28 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 11b. Multiphase pipeline incidents, by cause per year 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 
350 



300 



250 



200 



150 



100 



50 




■ 





1990 


1991 


1992 


1993 


1994 


1995 


1996 


1997 


1998 


1999 


2000 


2001 


2002 


2003 


2004 


2005 


CAUSE 


































Internal corrosion 


127 


119 


150 


140 


167 


142 


142 


173 


201 


200 


176 


168 


163 


145 


150 


158 


External corrosion 


62 


59 


53 


52 


52 


38 


54 


21 


38 


46 


38 


29 


61 


50 


53 


46 


Damage by others 


8 


12 


18 


17 


43 


38 


33 


41 


31 


13 


21 


16 


12 


22 


27 


30 


Weld 

Construction damage 


10 
6 


11 
16 


4 

7 


6 

3 


1 
9 


2 
17 


5 

18 


6 
15 


8 
15 


4 
4 


4 
16 


10 
16 


4 
11 


2 
27 


5 
13 


4 
17 


Overpressure 


8 


12 


8 


14 


13 


7 


15 


11 


15 


4 


6 


8 


10 


4 


1 


14 




3 


2 


6 


6 


2 


10 


9 


6 


4 


3 


7 


7 


8 


16 


17 


7 




1 


6 


6 


4 


7 


15 


24 


9 


7 


12 


17 


7 


15 


16 


10 


8 


Earth movement 


3 


6 


1 


3 


7 


8 


10 


5 


2 


4 


8 


4 


8 


10 


5 


10 


Valve/fitting 


6 


2 




1 


5 


3 


6 


3 


2 


1 




3 


1 


3 


10 


9 


Other 


6 


9 


9 


15 


9 


5 


18 


11 


12 


19 


11 


10 


10 


18 


18 


19 


Total number incidents 240 


254 


262 


261 


315 


285 


334 


301 


335 


310 


304 


278 


303 


313 


309 


322 


% of total 


5.1 


5.4 


5.5 


5.5 


6.7 


6.0 


7.1 


6.4 


7.1 


6.6 


6.4 


5.9 


6.4 


6.6 


6.5 


6.8 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 29 



Figure 12a. Crude oil pipeline incidents, by cause for all years combined 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 



Weld 18(4.4%) 



Figure 12a shows that crude oil pipelines experience a 
very small number of failures and that corrosion is 
responsible for a smaller proportion of crude oil 
pipeline failures, 37.7%, than of raw production fluids 
pipeline failures. This is understandable, as oil in 
major shipping pipelines will have had water removed 
before being shipped and should be less corrosive. 
Proportionally, the number of failures due to damage 
by others seems high for crude oil pipelines. This may 
be simply the result of the reduced number of 
corrosion failures, which tends to make all other 
failure causes seem greater in proportion. 

Pipeline failures caused by damage by others are 
reviewed in Figures 16 and 17. 

Figure 12b shows an increase in the number of failures 
in the "other" category, which seems to be a trend for 
some other products as well and which may indicate 
that accurate cause of failure is not being conclusively 
determined. 




^^onstruction damage 20 (4.9%) 
BSk Overpressure 6(1.5%) 



Pipe 18(4.4%) 
Joint 11 (2.7%) 
Earth movement 12(2.9%) 



Valve/fitting 33 (8.0%) 



Total number of incidents 411 (100%) 



30 * EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 12b. Crude oil pipeline incidents, by cause per year 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 
40 

35 

30 



1990 

CAUSE 


1991 


1992 


1993 


1994 


1995 


1996 


1997 


1998 


1999 


2000 


2001 


2002 


2003 


2004 


2005 


Internal corrosion 


j 10 


8 


1 


3 


9 


7 


10 


8 


4 




7 


10 


4 


5 


6 


2 


External corrosion 


4 


4 


4 


3 


7 


4 


2 


2 


3 


8 


4 


2 


2 


1 
1 


1 
3 
1 
1 


3 
6 

1 


Damage by others 


5 


1 


3 


3 


11 


7 


7 


6 


10 


6 

1 
1 


8 


8 
2 


4 
1 


Weld 

Construction damage 


I 2 
1 


1 
1 


1 


5 
2 




4 


1 
1 


5 


2 

3 


Overpressure 




1 
2 


3 


2 


1 
3 




2 

1 


1 








2 


1 


1 








1 


3 




Earth movement 


; 1 


3 




1 

1 


3 


1 


2 

1 


1 


1 

2 


1 

1 


1 






1 
1 




1 




I 1 


1 


3 


2 


2 


1 


4 


1 


1 


2 


4 


3 


1 




4 


3 







7 


2 


5 


2 


1 




1 


2 




2 


2 


3 


4 


6 


8 



Total number incidents 29 29 17 27 38 25 31 25 28 28 26 29 16 14 25 24 
% of total 7.1 7.1 4.1 6.6 9.2 6.1 7.5 6.1 6.8 6.8 6.3 7.1 3.9 3.4 6.1 5.8 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 31 



Figure 13a. Sour gas pipeline incidents, by cause for ail years combined 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 



Figure 13a shows that internal corrosion continues 
to be the primary cause of sour gas pipeline 
failures, as expected due to the very corrosive 
nature of sour gas production. The EUB conducts 
a high level of technical review and surveillance 
on sour gas pipeline and continually looks for 
opportunities to improve regulations directed at 
the operation of sour gas pipelines. In the 2000- 
2003 period, the EUB drew industry attention to 
an increasing number of weld failures in sour 
service, and it appears that some improvement has 
already resulted. 

Figure 13b also suggests improvements in 
corrosion control measures. 

However, the EUB is concerned about the number 
of "other" failures being recorded in the last three 
years, which could suggest that an accurate cause 
of failure has not always been determined. 




Total number of incidents 357 (100%) 



32 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 13b. Sour gas pipeline incidents, by cause per year 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 33 



Natural gas pipelines include dry sales gas, as well as wet or dry 
produced gas, and may contain H 2 S at levels up to and including 10 
moles H 2 S per kilomole of natural gas. There are a large number of 
small -diameter gas lines, as shown in Figure 4a, and a great number of 
these carry corrosive wet raw gas, the primary cause of a significant 
number of natural gas pipeline failures. As shown in Figure 14a, internal 
corrosion causes 57.4% of natural gas pipeline failures. Most of the 
failures occur on pipelines 6" and smaller, as shown in Figure 9. 

Analysis of the data shows that of the 235 592 km of natural gas 
pipeline, about 11% contains some H 2 S within the 10 mol/kmol criterion. 
This ratio has been fairly consistent over the span of this report. A 
separate subcategory of gas pipeline, called fuel gas pipeline, is included 
within the natural gas pipeline category for the purposes of this report. 
Analysis shows that about 1% of those pipelines had some H 2 S content 
within the 10 mol/kmol criterion. 

Figure 14b shows a pronounced spike in the number of natural gas 
pipeline failures for the year 2005. Although corrosion failure numbers 
increased slightly in 2004 and more in 2005, the numbers are still much 
lower than those for 1999 and 2000. The failure numbers had been 
reduced in the years after 2000 due to a concerted industry effort to 



34 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



address failures occurring in wet shallow-gas pipelines in southeastern 
Alberta. 

When analyzing the recent increase, two interesting facts came to light. 
The first is that of all the 3826 incidents on natural gas pipeline during 
this reporting period, 51.4% occurred in the Medicine Hat Field Centre 
area. This includes the large area of densely packed low-pressure sweet 
gas production around the Suffield/Medicine Hat area. Much of that low- 
volume gas production produces formation waters, resulting in difficult 
conditions for corrosion mitigation. The second fact is that when looking 
at other regions of the province, it was evident that natural gas pipeline 
failure numbers had increased in every area of the province, and no 
particular regional differences could be identified. Evaluation confirmed 
that the majority of failures continued to occur primarily on the small 
gathering system pipelines of 60.3 mm (2") to 168.3 mm (6") diameters. 
There were a few more recorded incidents on large-diameter natural gas 
pipeline, but analysis found that a number of these were scheduled test 
program failures and the remainder were attributable to a random 
distribution of causes. The EUB intends to further examine the increased 
number of natural gas pipeline failures to determine what efforts might 
be made to reverse the trend. 



Figure 14a. Natural gas pipeline incidents, by cause for ail years combined 

January 1 , 1990, to December 31 , 2005 (includes hits, leaks, and ruptures) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 35 



Figure 14b. Natural gas pipeline incidents, by cause per year 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 



t sEi."i=Bii!i?iHl 



CAUSE 


1990 


1991 


1992 


1993 


1994 


1995 


1996 


1997 


1998 


1999 


2000 


2001 


2002 


2003 


2004 


2005 


Internal corrosion 


62 


68 


61 


114 


136 


101 


119 


145 


154 


273 


246 


182 


142 


125 


118 


152 




External corrosion 


15 


22 


43 


22 


24 


17 


22 


16 


17 


20 


28 


32 


22 


15 


38 


44 




Damage by others 


25 


15 


21 


12 


53 


39 


38 


40 


40 


34 


51 


48 


33 


26 


31 


55 




Weld 


18 


6 


13 


9 


14 


9 


2 


14 


5 


3 


16 


5 


5 


1 


7 


15 




Construction damage 


9 
2 
8 


5 


3 
2 
7 


4 

1 
1 


3 


9 

1 
5 


4 

1 
1 


7 
1 
1 


10 


9 


8 


9 

2 
5 


8 

1 
3 


1 


7 


13 
1 
4 




Overpressure 
Pipe 


6 


1 


6 


2 


4 


6 


11 






5 


5 


2 


1 


4 


1 


1 


1 


1 


5 


3 


1 


5 


3 


5 


8 




Earth movement 


6 


2 


2 




1 


3 


3 


6 


4 


1 


1 


5 


3 


2 


1 


6 






7 


1 


2 


2 


1 


2 


2 


4 


6 


3 


5 


6 


3 


7 


3 


18 




Other 


9 


5 


7 


11 


11 


5 


5 


12 


3 


6 


10 


5 


1 


20 


15 


31 




Total number incidents 166 
% of total 4.3 


135 
3.5 


163 
4.3 


177 
4.6 


248 
6.5 


192 
5.0 


198 
5.2 


247 
6.5 


246 
6.4 


356 
9.3 


372 
9.7 


300 
7.8 


226 
5.9 


217 

5.7 


236 
6.2 


347 
9.1 



36 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figures 15a and 15b show that the causes of failure in the "other 
product" category are varied, due to an overall low number of failures 
and the relative lack of corrosion-related failures. As most "other" 
pipelines carry refined or processed substances, corrosion is infrequent. 
Third-party incidents make up a greater proportion of the failure causes, 
due to the reduced number of corrosion failures. Pipelines in this 
category include a significant number of LVP and HVP product pipelines 
traversing highly populated urban and suburban areas. Constant vigilance 
is exercised by operators in those areas to ensure that rights-of-way are 
patrolled frequently and that public in the area is aware of the presence 
of the pipelines. This could partly explain why pipelines in this product 
class exhibit the second-lowest ratio of third-party hits, as shown in the 
simple analysis in the commentary of Figure 6. The higher proportion of 
valve and fitting incidents and the recent increase in the number of those 
incidents are likely the result of recent increased operator awareness and 
monitoring of fugitive emissions in response to greenhouse gas 
conservation efforts. 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 37 



Figure 15a. All "other product" pipeline incidents, by cause for all years combined 

January 1, 1990, to December 31, 2005 (includes hits, leaks, and ruptures) 



Weld 5 (2.6%) 




.0%) 



Internal corrosion 8(4.1%) 
Total number of incidents 196 (100%) 



38 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 39 



Figure 16 shows the annual number of pipeline hits and whether those 
hits resulted in a leak or rupture of the pipeline. Prior to 1994, the EUB 
did not record pipeline hits that did not result in product loss. Since 
beginning to record such hits, early data showed that about one out of 
every two pipeline hits resulted in product loss and, worse, that many 
pipeline ruptures resulted, potentially creating an immediately dangerous 
situation. Fortunately, over the last several years the number of product 
releases caused by third-party damage has decreased, particularly the 
number of ruptures. This is very positive, as it implies a safer workplace 
for the excavation contractors and reduced risk to the public. 

There is a correlation between the number of damage incidents and the 
robust level of industry activity: a jump in incidents occurred in 2004 and 
2005, reflecting current robust oil patch activity. As discussed for Figure 
6, growing field activity increases the chance of unplanned pipeline 
contact, and the increasing presence of an inexperienced junior 
workforce may also be contributing to this. 

In 2005 the Pipeline Regulation strengthened the requirements for safe 
ground disturbance practices, excavation inspection, and personnel 
training. It is hoped that in subsequent years these will translate into a 
reduction of the number of annual hits. 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 40 



Figure 16. Pipeline incidents due to damage by others per year 

All pipeline incidents from January 1 , 1990, to December 31 , 2005 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 41 



Figure 17 confirms that third-party damage occurs most frequently on 
pipelines between 60.3 mm (2") and 168.3 mm (6") diameter, which 
corresponds to the proportion of pipe sizes that make up the largest 
portion of the province's inventory. Figure 17 also shows that third-party 
damage occurs infrequently on large-diameter transmission pipelines, 
probably because their location is well known, they are well marked, and 
operators typically conduct frequent right-of-way surveillance. 



42 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 17. Damage by others, by pipe size for ail years combined 

All pipeline incidents from January 1, 1990, to December 31, 2005 



500 











454 






























































200 


































ncidents 


150 
100 




1 






























Number of i 

















1 


































1 






















50 


















































MM 




















0 


■ 












I 






■ 


wm 




_______ 


... 






SIZE OF PIPE mm 
(nominal inches) 

INCIDENTS 


Not 
specified 


60.3 
(2) 


88.9 
(3) 


114.3 
(4) 


168.3 
(6) 


219.1 
(8) 


273.1 
(10) 


323.9 
(12) 


355.6-406.4 
(14-16) 


508 
(20) 


559-610 
(22-24) 


762-914 
(26-36) 


Total 
number 


%of 
total 


I Hit (no Droduct loss) i 


12 


105 


222 


126 


79 


30 


15 


12 


8 


1 


2 


5 


617 


50.4 


Leak 


7 


52 


92 


40 


15 


4 


5 


1 






2 




218 


17.8 


Rupture 


17 


87 


140 


86 


40 


12 


4 


1 


1 




2 




390 


31.8 


Total number incidents 36 
% of total 2.9 


244 
19.9 


454 

37.1 


252 
20.6 


134 
10.9 


46 

3.8 


24 

2.0 


14 

1.1 


9 
0.7 


1 

0.1 


6 
0.5 


5 
0.4 


1225 


100.0 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 43 



Figure 18 shows the number and composition of the 16 004 releases from 
pipeline that occurred during the reporting period. There were more 
releases than incidents, which is to be expected because raw fluid 
production often consists of multiple substances that are recorded 
separately in the estimation of spill volumes. Spills from pressure tests 
are excluded from the analyses of Figures 1 8 to 25 and are evaluated 
separately in Figures 26 and 27. 



44 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 18. Number of pipeline releases, total for all years 

All pipeline releases from January 1, 1990, to December 31, 2005 (test failures are excluded) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 45 



Other than a general mild increase in numbers of releases coincident with 
the increasing size of the pipeline inventory, there are no unusual trends 
indicated by Figure 19. Although the length of installed pipeline has 
increased on average by 6.2% per year, the number of releases has 
remained fairly constant, suggesting a overall reduction in the frequency 
of spills based on installed pipeline length. 



46 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 19. Pipeline releases by substance released per year 

All pipeline releases from January 1 , 1 990, to December 31 , 2005 (test failures are excluded) 




1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Total %of 



PRODUCT 
Hydrocarbon liquid 
Hydrocarbon gas 


244 
168 




256 
168 


261 
179 




291 
169 












number 


total 


268 
138 


315 
228 


185 


0 I I 

234 


330 337 
231 338 


333 


291 


^yy 
252 


225 


251 


311 
388 


4769 
3778 


29.8 
23.6 


Other 


2 


4 


1 


2 




2 




1 






1 


1 


1 


9 


9 


12 


45 


0.3 


Sour gas 


14 


11 


14 


12 


8 


14 


18 


13 


21 


21 


30 


36 


21 


19 


29 


18 


299 


1.9 




511 


457 


461 


411 


444 


410 


470 


417 


460 


426 


445 


454 


469 


408 


420 


450 


7113 


44.4 



Total number releases 939 878 900 865 995 886 1022 976 1042 1122 1129 1087 1042 947 995 1179 16 004 

% of total 5.9 5.5 5.6 5.4 6.2 5.5 6.4 6.1 6.5 7.0 7.1 6.8 6.5 5.9 6.2 7.4 100.0 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 47 



The EUB requires that the volume of spilled substance be reported in the 
event of a pipeline release. The accuracy of these reports varies, as not 
all pipelines are equipped with metering and sometimes the starting time 
of an event is unknown. Best estimates are made based on production 
rates, pipeline capacities, metering, and measurement of spill areas. In 
the case of gas production, the gas disperses, making accurate 
measurement difficult. 

Release volumes are divided into four volume classes: less than 100 m 3 
of liquid, or 100 000 m 3 of gas; 100 to 1000 m 3 of liquid or 100 000 to 
1 000 000 m 3 of gas; 1000 to 10 000 m 3 of liquid or 1 000 000 to 
10 000 000 m 3 of gas, and greater than 10 000 m 3 of liquid or 10 000 000 
m 3 of gas. Figure 20 shows that of the 16 004 releases, 96.0% fall into 
the smallest category. Another 3.5% are between 100 and 1000. Releases 
greater than 1000 m 3 liquid or 1 000 000 m 3 gas constituted only 0.5% of 
the events. Of all sour gas releases, only six (2%) exceeded 100 000 m 3 . 



48 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 20. Number of pipeline releases, by substance and volume 

All pipeline releases from January 1 , 1990, to December 31 , 2005 (test failures are excluded) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 49 



Figure 21 shows the release breakdown for the smallest volume category 
of releases by substance and year. In this volume class, the water, liquid 
hydrocarbon, and gaseous hydrocarbon releases are each responsible for 
approximately a third of the recorded releases. There is also a small 
number of sour gas and "other" releases. The releases in this volume 
class account for 96% of the number of all releases. 



50 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 21. Pipeline releases <100 m 3 (liquids) or <100 10 3 m 3 (gas), by substance and year 

All pipeline releases from January 1, 1990, to December 31, 2005 (test failures are excluded) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 51 



Figure 22 shows that releases in the second volume class account for 
3.5% of the total number. The higher proportion of water pipeline 
releases of medium volume suggests that either water pipeline failures 
are not readily detected or the flow rates are high to create these spills. 
The Canadian Standards Association is considering whether 
instrumented leak detection requirements should be applied to produced 
water pipelines, and it is possible that future editions of CSA Z662 may 
include this requirement. A positive trend is that the number of these 
medium-size releases seems to be gradually declining. 



52 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 22. Pipeline releases 100 - 1000 m 3 (liquids) or 100 • 1000 10 3 m 3 (gas), by substance and year 

All pipeline releases from January 1 , 1 990, to December 31 , 2005 (test failures are excluded) 




Figure 23 shows that large-volume releases account for about 0.5% of all 
releases. Most of the failures in the earlier part of the reporting period 
occurred in water pipelines, but those release numbers dropped 
noticeably in recent years. There were some significant large natural gas 
pipeline releases in the 2001-2002 period, but the most recent years show 
a marked improvement. 



54 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 23. Pipeline releases 1000- 10 000 m 3 (liquids) or 1000 - 10 000 10 3 m 3 (gas), by substance and year 

All pipeline releases from January 1, 1990, to December 31, 2005 (test failures are excluded) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 55 



Fortunately, very large releases are rare, as indicated in Figure 24. There have been a few isolated pipeline failures on sweet gas transmission pipelines that 
resulted in large volumes of released gas due to the large diameter and long length of the transmission line segments. The released gas will either disperse 
or burn if it is ignited. 

Figure 24. Pipeline releases >10 000 m 3 (liquids) or >10 000 10 3 m 3 (gas), by substance and year 

All pipeline releases from January 1, 1990, to December 31, 2005 (test failures are excluded) 



3 



2 




1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 

% of all 
Total releases by 

PRODUCT number number 



Hydrocarbon gas 


1 


1 1 


1 




4 


<0.1 










1 


1 


<0.1 



Total number releases 1 111 15 <0.1 



56 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 25 shows the proportions of each release volume class in relation to the total number of releases. 



Figure 25. Number of pipeline releases, by volume 

All pipeline releases from January 1 , 1990, to December 31 , 2005 (test failures are excluded) 




RELEASE VOLUME, IVP (LIQUIDS) OR 1Q3 H/P (GAS) 

<100 100- 1000 1000- 10 000 >10 000 

Total 15 360 565 74 5 



Total % of 

number all releases 
16 004 100.0 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 57 



Figure 26. Pipeline release volumes from pressure test, m 3 (liquids) or 10 3 m 3 (gas) 

All pipelines from January 1, 1990, to December 31, 2005 (leaks and ruptures only) 



Pipeline releases that occur during 
pressure testing of a pipeline, whether 
for qualification of new pipeline or 
requalification of existing pipeline, 
have not been included in the previous 
figures. During the reporting period, 
1311 releases occurred during 
pressure testing. Figure 26 shows the 
proportions of each volume class in 




relation to the total number of <100 100-1000 1000- 10 000 >10 000 number all releases 

releases. Total 1279 27 4 1 1311 100.0 



58 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 27. Pipeline releases from pressure test, by substance 

All pipelines from January 1, 1990, to December 31, 2005 (leaks and ruptures only) 



During the reporting period, there were 
1311 substance releases from 1256 
pressure test failures. 76.8% of the 
releases were fresh water, and the 
remaining 23.2 % were a variety of other 
products. In Figure 27, substances 
combined into the natural gas category 
include fuel gas, gas production (raw), gas 
production (marketable), and natural gas. 
The produced water category includes 
process water, produced water, and 
corrosion inhibited water. 

Pressure tests are conducted for a number 
of reasons, such as proving out new 
construction, verifying integrity, 
requalifying for a higher operating 
pressure or change of substance, and 
identifying near-critical defects. 



Freshwater 1007(76.8%) 




Allother 16(1.2%) 
Air 18(1.4%) 

Methanol blends 40(3.1%) 



Natural gas 48 (3.7%; 



Crude oil 64 (4.9%) 



Produced water 118(9.0%) 



1256 actual events resulted in 1311 released volumes 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 59 



The data for Figure 28 were compiled by taking the cumulative length of 
all pipeline of each product class at each year-end and dividing that by 
the number of incidents recorded for that year for the product class. 
Frequency of incidents is reported as the number of incidents occurring 
per 1000 km of pipeline. 

Figure 28 shows that most pipeline substances exhibit a relatively low, 
and reasonably steady, incident frequency in the range of about 1 to 2 
incidents per 1000 km per year. The exceptions are water and multiphase 
pipelines. Water pipelines are very susceptible to corrosion unless coated 
with an internal corrosion barrier. Industry has showed significant 
progress in bringing down the water pipeline failure rate, particularly in 
the early years of this reporting period. It appears that more improvement 
is becoming more difficult to achieve, as the failure rate has levelled out. 

Multiphase pipelines are also very susceptible to corrosion, as they 
typically carry some water. If the amount of oil is sufficient to wet out 
the interior surface of pipe, the oil may be effective in protecting the steel 
from corrosion. However, when that oil film is disrupted or is 
insufficient to coat the pipe, corrosion can occur. Corrosion inhibition 
can be effective, but must be carefully administered to ensure that 



60 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



inhibition systems work effectively in the water and oil environments. 
Industry appears to be making gradual improvement in the performance 
of these multiphase pipelines as well, although the failure rate remains 
high, at 6.3 failures per 1000 km. 

The failure rates for crude oil, natural gas, sour gas, and other pipelines 
have been nearly steady over the reporting period. Of particular interest 
are the failure rates for sour gas pipelines, which are currently at an all- 
time low of 1.2 per 1000 km. This is very encouraging, as it 
demonstrates that operators of these critical pipelines are operating them 
very carefully. Indeed, the failure rate on these sour gas pipelines, which 
carry corrosive, untreated production, is bettered only by pipelines in the 
"other" category, which are primarily pipelines shipping clean 
hydrocarbon products. 

The only apparent disappointment is a minor increase in the failure rate 
of natural gas pipelines in 2005. Some of these pipelines do contain wet, 
corrosive raw natural gas, which contributes to the number of failures. 
More analysis will be conducted to look for problem areas that could be 
addressed further. 



Figure 28. Average frequency of pipeline incidents, by year 

All pipelines from January 1, 1990, to December 31 , 2005 (includes all hits, leaks and ruptures) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 61 



Data for Figure 29 were compiled by taking the cumulative length of 
operating and permitted pipeline at each year-end and dividing that by 
the number of incidents recorded for the year. 

Over the reporting period, pipeline failure frequency has dropped from 
5.2 to a low of 2.2 in 2004. In 2005 there was a slight rebound, to 2.4, 
due to an increased number of natural gas pipeline incidents. However, 
the overall trend shows a very successful performance story, in that 
industry has exhibited a steady, measurable reduction in the frequency of 
pipeline failures. Pipeline failures are very costly from the perspective of 
lost production and royalties, environmental damage and cleanup, 
increased greenhouse gas emissions, and loss of public goodwill. The 
EUB will continue to work with industry to continue this trend of 
improvement. Targets for a few years ahead will be to introduce and 
implement pipeline integrity management programs, bring more 
corrosion-resistant pipeline materials into common use, improve leak 
detection methods, and maintain relevant and meaningful regulations to 
assist in improving pipeline performance. 



62 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



Figure 29. Average frequency of pipeline incidents, by year 

All pipelines from January 1 , 1990, to December 31 , 2005, all products (includes all hits, leaks and ruptures) 




EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 63 



5 Future Initiatives 



The EUB is always looking for ways that pipeline performance may be 
enhanced. Over the last decade, the EUB has been a strong proponent of 
trying new corrosion-resistant composite and polymeric materials and 
has worked with industry to allow numerous installations of newly 
developed materials. Many of these have exhibited excellent 
performance and are now being used regularly. The EUB expects growth 
in this sector, especially as more new materials come to market. The 
EUB will, however, monitor performance of these materials carefully to 
ensure that the materials provide satisfactory performance. 

Senior technical staff from the EUB Pipeline Section participate in 
technical work groups for the Canadian Standards Association ( CSA ) 
Standard Z662: Oil and Gas Pipeline Systems, the main governing 
document for technical requirements related to pipeline design, 
construction, and operation in Canada. This ensures that experiences 
seen in Alberta are brought to the discussion table and that changes to the 
standard can be made if necessary. A new edition of CSA Standard Z662 
is expected in mid-2007, which will specify technical requirements for 
and acceptance of certain non-metallic and composite pipelines, 
improved construction standards for sour service pipelines, guidelines for 
pipeline integrity management programs, and various other 
improvements. 



The EUB recently completed a review of the requirements for HVP 
product pipelines, and recommendations from that work were included in 
a revision to the Pipeline Regulation completed in 2005. Many other 
changes were also made in that revision, such as improvements to the 
way pipelines are discontinued or abandoned, improvements to ground 
disturbance practices around pipelines, additional right-of-way 
inspection, and annual corrosion control evaluations. The EUB reviews 
and updates the Pipeline Regulation periodically as the need arises. 

On occasion the EUB conducts examination of specific pipeline issues 
and works with industry to see where improvements can be made. One 
such initiative was to examine different methods of initiating emergency 
shutdown valve closure to make sure that current methodology is still 
appropriate. The EUB has also been working on projects to re-evaluate 
existing requirements for emergency response planning around pipelines 
and make change as appropriate. Projects being considered include 
developing criteria for evaluating material compliance with sour service 
specifications and detailed processes for conducting assessments of 
integrity management programs to see that they meet regulatory 
requirements. 



64 • EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) 



6 Other Information 



Technical standards and regulatory information related to pipeline can be found in the following documents: 

Alberta Pipeline Act 
Alberta Pipeline Regulation 

American Society of Mechanical Engineers (ASME) B16.5 Pipe Flanges and Flanged Fittings 
ASME B3 1 .3 Process Piping 

Canadian Standards Association (CSA) Standard Z662: Oil and Gas Pipeline Systems 
CSA Standard Z245.1: Steel Pipe 
CSA Standard Z245.1 1: Steel Fittings 
CSA Standard Z245.12: Steel Flanges 
CSA Standard Z245.15: Steel Valves 

CSA Standard Z245.20: External Fusion Bond Epoxy Coating for Steel Pipe 

CSA Standard Z245.21 : External Polyethylene Coating for Pipe 

CSA Standard B 137 Series 2: Thermoplastic Pressure Piping Compendium 

EUB Directive 022: Use of Bimodal High-Density Polyethylene Pipe in Oil and Gas Service 
EUB Directive 026: Setback Requirements for Oil Effluent Pipelines 
EUB Directive 041: Adoption of CSA Z662-03, Annex N, as Mandatory 

EUB Directive 056: Energy Development Applications and Schedules (also contains a number of reference tools for pipeline applications) 

EUB Directive 066: Requirements and Procedures for Pipelines (pipeline inspection guide) 

EUB Directive 07 1 : Emergency Preparedness and Response Requirements for the Upstream Petroleum Industry 

EUB Guide 30: Guidelines for Safe Construction near Pipelines 

EUB Information Letter 2002-02: Strength and Leak Pressure Testing of Pipelines Using Gaseous Test Media 
EUB Provincial Surveillance and Compliance Summary 2005; also previous years' annual ST-57 Summaries 

NACE MR0175/ISO 15156: Materials for use in H 2 S-Containing Environments in Oil and Gas Production 



EUB Report 2007-A: Pipeline Performance in Alberta, 1990-2005 (April 2007) • 65 



N. 



- 



3 3286 53529685 5