CONCAWE, the European Association for the environment, health and safety in refining and distribution, has collected data on spillages from European cross-country oil pipelines since 1971, and has published annual reports over this period.

CONCAWE gathers comprehensive information from the approximately 70 operating companies and joint-ventures that make up its membership. The information is published in a statistical form that provides an important insight into the industry and its increasing integrity over the decades. Outside the United States, there are no such publicly-available data for either oil or gas transmission pipelines, and so the CONCAWE annual reports are of significant value in establishing trends and providing historical analyses on which to base future decisions. The association is acknowledged for its assistance with this introduction to the report and brief discussion of its findings.

Assessing pipeline failure

The current report, entitled Performance of European cross-country oil pipelines: a statistical summary of reported spillages in 2007 and 1971, focuses on CONCAWE members’ pipeline network, which includes 150 pipeline systems with a total length of 34,721 km (at the time of the report), slightly less than the length of the network in the previous year (see Figure 1). The difference is partly due to corrections to the reported data, and because a few companies did not report and were therefore not included in the 2007 statistics. The reported volume transported in 2007 was 762 MMcm of crude oil and refined products, a figure that has been stable in recent years.

The nine spillage incidents reported in 2007 correspond to 0.26 spillages per 1,000 km of pipeline, just under the five-year average and well below the long-term running average of 0.55 per 1,000 km. The long-term running average has been steadily decreasing over the years from a high of 1.2 per 1,000 km in the mid 1970s. There were no reported fires, fatalities, or injuries connected with these spills, and the gross spillage volume was 988 cubic metres, equivalent to 1.3 parts per million of the total volume transported. This corresponds to 28 cubic metres per 1,000 km of pipeline compared to the long-term average of 56 cubic metres. CONCAWE reports that 53 per cent of the spilled volume was recovered or safely disposed of.

Most of the pipeline spillages were small, and just over 5 per cent were responsible for half of the gross volume spilled. In the past, the pipelines carrying hot oils such as fuel oil have suffered very severely in Europe from external corrosion. This has been due to design and construction problems, and most have now either been shut down or switched to cold service. The great majority of pipelines now carry unheated petroleum products and crude oil, and only 270 km of heated pipelines were still in service during 2007. There have been no reported spills from hot pipelines in the last five years of this survey.

Of the nine reported incidents in 2007, two were related to corrosion and seven were connected to third-party activities, either accidental or malicious.

Over the long term, third-party activity has been regularly responsible for the majority of spillage incidents, although this cause has progressively reduced over the years. Mechanical failure is statistically the second greatest cause of spillages, although not relevant in 2007, and corrosion (both internal and external) has generally come third in the league table of spillage causes. Interestingly, after considerable progress in reducing the frequency of mechanical failures in the 1970s and 1980s, the mid-1990s and beyond has seen an increase in mechanical failures.

Most of the European network of pipeline systems operated by CONCAWE’s members were constructed in the 1960s and 1970s. In 1971, at the time of the first report, around 70 per cent of the network was 10 years old or less; by 2007, only 5 per cent was 10 years old or less and 42 per cent was over 40 years old. So far this does not appear to have led to any increase in spillages. The industry’s concern to maintain this situation (or improve it) is evidenced by the fact that in 2007, 84 sections were inspected by at least one type of intelligent pig, representing total run length of around 9,300 km. However, most inspection programmes involved running more than one type of pig in the same section, so that the total actual length inspected was 5,281 km – 15 per cent of the network. CONCAWE points out that, overall, there is no evidence to show that the ageing of the pipeline system poses any increase in the level of risk. The development and introduction of new inspection tools and techniques will allow these pipelines to continue reliable operation for the foreseeable future, although performance statistics such as these will be necessary to confirm the achievement.

Intelligent inspection

CONCAWE has been collecting data on intelligent pig inspection activities for the last 17 years, including a one-off exercise to collect back data from the time intelligent pigs were first used around 1977. Separate records have been kept for inspections made by metal-loss pigs, crack-detection pigs, and geometry (or calliper) pigs. In 2007, as mentioned, the 84 sections that were inspected by at least one type of intelligent tool covered a 9,300km, split as seen in Table 1.

As shown in Figure 2, the use of intelligent pigs for internal inspection grew steadily up to 1994. They then levelled off for a period during stabilisation of the industry, and resumed an upward trend in 2003, with an average annual pig run total of almost 8,000 km, and an average actual length inspected of 4,890 km over the last five years.

Within the last ten years, a period considered as a reasonable cycle for this type of inspection, 443 out of a total of 699 active sections (63 per cent) have been inspected at least once by at least one type of pig, representing 76 per cent of the total length of the network. Of the balance of pipeline sections, there are certainly some – mainly the older ones – which were not designed to be pigged or, because of small sizes, small radius bends, or lack of suitable pig launchers or receivers, cannot be intelligent pigged. The relatively recent introduction of pigs to inspect 6 inch diameter pipelines and smaller, and the increasing number of tools available to inspect ‘unpiggable’ pipelines, means that these problems are diminishing in the network operated by CONCAWE’s members. The length of un-inspected pipelines is therefore certainly less than the above figure, and will continue to decrease in future years.