Select Committee on Trade and Industry Written Evidence

APPENDIX 6

Memorandum by Michael Gillard

  I write this submission as a freelance journalist with considerable experience investigating British Petroleum's worldwide operations for national newspapers and television.

  Working with others I have been responsible for two World In Action documentaries exposing BP's human rights and environmental record in Colombia. We have also written investigative articles on BP in Alaska and Colombia for The Guardian, when I was part of the investigation unit.

  In Alaska, we exposed evidence from senior whistleblowers that key safety equipment was malfunctioning and could cause a major spill or explosion that escalated because of inadequate emergency response plans. The main allegations of the whistleblowers—BP safety inspectors—centred on inadequate monitoring and preventive maintenance of vital safety systems and equipment, and the falsification of reports to show compliance with quality assurance programmes.

  The articles helped US Congressmen and state and federal regulators force BP to carry out a widespread audit of its TAPS pipeline and oil terminals. It confirmed employee concerns.

  In the UK, the parliamentary International Development Select Committee picked up on another front-page article that had uncovered plans by BP's security department to supply arms and training to the brutal Colombian armed forces protecting a major new pipeline. We also exposed a spying operation against the local communities living near the Ocensa pipeline, which BP operates on behalf of an international consortium. MPs subsequently grilled BP executives at a hearing in 1999. The oil company blamed rogue elements in its security department.[79]

  Most recently, on 15 February, 2004, I wrote an article with The Sunday Times Insight team on BP's operations in Azerbaijan and Georgia, where the oil company is operator for an international consortium constructing an underground pipeline from the Caspian Sea to the Mediterranean.[80]

  The Baku-Tblisi-Ceyhan pipeline crosses archaeological sites, nature reserves and 1,500 watercourses, from small streams in the Caucasus Mountains to rivers and canals. Almost all the route cuts through an active earthquake zone. Long-standing regional conflicts with Armenian and Kurdish separatists add to the instability.

  Unfortunately, again internal warnings were ignored, this time claiming the BTC pipeline is being built with a major design fault, which, if uncorrected, corrosion experts warn will effectively see thousands of environmental "time bombs" buried along the 1,000-mile route.

  Ordinarily, this would not be a matter for this committee but for the involvement of the ECGD and the comments you have invited on its performance.

  The committee may be aware that this year on 3 February the BP-led consortium secured a $2.6 billion loan agreement with development banks, export credit agencies and commercial banks for the construction of the BTC pipeline by the first quarter of 2005.

  It should be remembered that BP did not want to build this pipeline. It was regarded internally as uneconomical and wrought with reputation risk. The main impetus was the US governments of Clinton and now President Bush, following BP's merger with American oil giant, Amoco in 1998.

  The BTC pipeline is the biggest and most challenging BP has ever built. It is being constructed under the intense gaze of international campaigners demanding the oil giant lives up to its ethical commitments that followed the major environmental and human rights scandals in Alaska and Colombia.

  Neither of those pipelines have the geopolitical significance of the BTC one—the first east-west energy corridor that will reduce Anglo-American reliance on the Middle East and OPEC.

  In Whitehall, the pipeline is seen as vital to Britain's energy policy as North Sea oil reserves dwindle. It is also key to what analysts call the "New Great Game", a geopolitical struggle between Presidents Bush and Putin for control of the 15 new republics that emerged from the former Soviet Union in 1991. Washington sees Azerbaijan and Georgia as bulwarks against Russia and Iran. The Pentagon is training security forces in both former Soviet republics to defend the offshore platforms, oil installations and pipeline.

  BP chief executive Lord Browne made it clear to the Financial Times in November 1998 that there would be no BTC pipeline without "free public money". BP, who recently announced record profits of £15 billion, and its BTC partners are putting in $1 billion of their own money. This means 70% of construction costs will be met by loans from the Lenders Group.

  It is undisputed that without the support of the World Bank, the European Bank of Reconstruction and Development, political risk insurers and export credit agencies, among them the US, French, Italian and Japanese governments, the commercial banks, which includes the Royal Bank of Scotland, owners of NatWest, would not have got involved in this mega-project.

  In December 2002, Baroness Symons assured British NGOs by letter that ECGD cover "would only be given if the department was satisfied that the relevant environmental, social and human rights impacts had been properly addressed and the financial and project risks were acceptable."

  A year later, the EBRD described the benefits of the BTC project to Azerbaijan and Georgia in these terms: "The application of the highest international environmental and technical standards, highest health and safety standards."

  On 17 December 2003 Mike O'Brien, the minister responsible for the ECGD, informed parliament he was agreeing with his department's recommendations to provide $150 million cover. He said the decision was made after "a rigorous assessment of the risks associated with the project and a thorough review of the environmental, social and human rights impacts".

  So it was that on 3 February, 2004 at the official signing ceremony in Baku, the capital of Azerbaijan, the ECGD agreed to provide cover of $106 million, approximately £58 million.

  Before financial closure and up until the eve of publication on 14 February, we were in discussion with the ECGD. It, along with other members of the Lenders Group, assured The Sunday Times that its cover came with unprecedented levels of environmental and social monitoring during the construction phase and afterwards.

  In the ECGD decision document we noticed it had referred to being in receipt of "allegations concerning the quality control systems in use for construction of the pipeline." We asked for further details and received this response:

"During the early construction phase of the pipeline in June 2003, a contractor working for BTC experienced some difficulties in producing welds of sufficient quality. The company informed the lender group—the banks, export credit agencies, including ourselves and the IFC and EBRD—of these problems (which) aren't uncommon in pipeline construction.

"They confirmed all the faults would be eliminated prior to the pipeline being laid underground. A few months later we received allegations through an NGO (Corner House) that the welding quality and the contractor's quality control systems were not adequate. ECGD passed this information on to the company, which indicated that it was in fact the same problem they'd brought to our attention earlier.

"BTC reconfirmed that the Quality Control systems for the construction of the pipeline were sufficiently rigorous and all poor quality welds were being cut out and re-welded. Something to bear in mind, the lenders group, which includes ECGD, do have independent engineers and we are also checking on the Quality Control systems as part of their regular monitoring visit to the sites. Also, when the pipeline is actually physically completed it will be tested both for the company itself but also for the lender to check that it is fit for purpose.

"The documentation explicitly requires BTC to inform us of incidents that are likely to have significant environmental or social impacts. The fact that a number of welds had failed their tests and were going to be cut out and re-welded would not be classed as something significant, as this is part of normal construction process of pipelines. Any leaks during the operation would need to be reported to us however. But oil isn't expected to be flowing through the pipeline until 2005, so we are some time away."

  The ECGD press office also told us it would expect BP to make it aware of any issues that might affect the integrity of the pipeline. And confirmed that the welding problems in June 2003 were the only issues reported by BP before financial closure on 3 February. The spokeswoman reassured The Sunday Times it had been carrying out "extensive due diligence for over two years before financial closure."

  ECGD also told us that at financial closure BP/BTC was "required to warrant that they have provided all relevant information including that which relates to environmental and social issues." Drawing down on the now agreed loan would be prevented, ECGD said, if there was a default or "other events occur that have a material adverse effect on the project."

  All these unequivocal statements came after the ECGD press officer had made over a week of extensive inquiries of her colleagues at our insistence.

  The Sunday Times investigation had by this time already established the following facts, none of which BP made known to ECGD nor were they picked up by its much-vaunted independent monitoring systems:

  In May 2002, two BP managers in Baku—Paul Stretch, Technical Manager, and Rod Hensman, Senior Project Engineer—were concerned about proposals coming from London for coating the field joints of the BTC pipeline. These are the welds on each 11.5 metre section of the steel pipeline. There are approximately 150,000 field joints on the pipeline.

  Stretch had been in the Caspian region for 10 years. He was responsible for ensuring all engineering aspects were satisfactory to BP.

  He and Hensman had received a report from a Mr Trevor Osborne, BP's materials consultant for the BTC project. BP had put together a Design Group based at the West London offices of the US construction giant Bechtel. In June 2001, Bechtel had won a US$150 million contract to provide procurement and detailed engineering for the Azerbaijan and Georgia sections of the pipeline. They contracted John Brown Engineering, who in turn contracted Osborne.

  Such was their concern about Osborne's report, Stretch and Hensman commissioned a leading pipeline integrity consultant called Derek Mortimore to review it. He was already working for BP on its offshore operations in Baku.

  The two BP managers were concerned Osborne's report appeared to favour the use of a liquid epoxy paint—manufactured by a Canadian company called Speciality Polymer Coatings (SPC)—for coating the field joint. Mortimore immediately felt this would be a serious mistake.

  Each section of steel pipe has a plastic "jacket" to protect it from external corrosion. After the sections are welded together on site, the field joint is coated with an anti-corrosion paint. To be effective, the field joint coating must bond permanently with the steel and the plastic. It must also be hardy but flexible to earth movements when buried. Liquid epoxy paint was known to have good adhesion to steel and poor adhesion to plastic.

  Mortimore advised Hensman to find a proven alternative as he predicted the Canadian coating, SPC 2888, wouldn't bond to the plastic and could crack in the winter leaving the steel pipe dangerously exposed to corrosion.

  In effect, he argued that each field joint would be a buried environmental time bomb waiting to explode as one million barrels of hot, high-pressured oil coursed through the pipeline daily.

  On 27 July 2002, Hensman sent a list of questions to Osborne about the use of SPC 2888. In one question Hensman said: "Track record should be one of the key factors in selection of any system. We should not be a testing ground for new materials, especially when we are designing for a life of up to 40 years. We need confidence in the long term integrity of the field joint coating . . .  the selection of the field joint coating is one of the most important technical decisions on the project."

  However, by this time Osborne was already organising a performance test at the Advantica Laboratory in Loughborough. SPC and two other products, one from the UK and the other from Germany, were tested for a range of physical and environmental factors like curing, adhesion, flexibility and impact.

  A preliminary test had already raised a question mark over SPC's adhesion to plastic. But the highly confidential Advantica test report nevertheless gave SPC the edge over its two competitors. Although significantly it remarked that, "no single coating material consistently performed well in all tests".

  In September 2002, at a crucial meeting in Bechtel's offices, senior BP managers approved the use of SPC for the field joints. However, the minutes described the decision as "a step change in present industry practice, both technically and commercially."

  Clearly BP recognised the paint had no track record on the type of pipeline it was using in Azerbaijan and Georgia, but went ahead anyway.

  Osborne then took the highly unusual step of nominating SPC 2888 in the official BP specification as the only paint to be applied when coating the field joints. In other words, there would be no fall back position if it failed. Industry experts consulted by The Sunday Times say single sourcing is so rare because it removes any liability from the contractors.

  A senior BP insider also questioned how Osborne got away with this single sourcing when he said in normal circumstances BP's procurement manager would have been all over him.

  Stretch was now even more alarmed at Osborne's specification and asked Mortimore to review all the paperwork. He produced a seven-page report dated 11 November 2002. It was passed to the BP Design Group in London, sparking an exchange of emails with David Winter, BP's health, safety and technical compliance manager, who worked closely with Osborne in the Design Group and supported his approach.

  Mortimore's conclusions were damning:

—  "It was clearly a serious mistake for BP to nominate one material only" and "not normal practice." He warned this could invite a legal challenge for restriction of trade under European law.

—  "We (BP) are completely out on a limb, we cannot identify any pipeline owner who uses (SPC 2888) on (plastic coated) field joints anywhere in the world."

—  "Clearly the use of (SPC 2888) is going to lead to a serious problem, particularly during the colder months, with curing the applied paint."

—  "Cost for repairs could be astronomical" in the events of cracks in the coating, he warned.

—  Mortimore's report also said Osborne's specification was "under-developed and incomplete" and "utterly inappropriate to protect the pipeline for its estimated design life."

—  "The potential for claims against the company is open-ended." He recommended immediate action and asked rhetorically whether BP had considered "the insurance implications."

  At the time BP received this report the oil company was in an intense propaganda war with international NGOs who were lobbying the Lenders Group to withdraw its financial backing, especially development banks and export credit agencies like the ECGD.

  The Mortimore report also noticed the paint had not been tested on a field joint in simulated weather conditions. This is remarkable because arguably there was even more reason than usual for BP to ensure a rigorous pre-qualification testing regime as SPC 2888 was a single source product recognised in the oil company's own internal memo as a "step change in industry practise". In other words, it was experimental engineering.

  The testing regime should have been exemplary and above reproach. Unfortunately, our investigation suggests this is not the case.

  Firstly, two months before his November report, Mortimore provided BP with a dossier in August 2002 containing independent industry opinions supporting his conclusion that any liquid epoxy, not just SPC 2888, would not adhere to a plastic coated pipeline, especially in cold weather conditions and with an expected guaranteed lifetime of 40 years.

  The dossier was about one inch thick. It contained a sworn declaration by Mortimore that he had no interests in any products concerned. The dossier contained declarations from international pipeline coating companies, corrosion consultants and scientists saying the type of paint specified was not appropriate for the BTC pipeline.

  The field joint coating market is very small. The Sunday Times spoke to a number of other specialists in this field who manufacture liquid epoxy paint. One leading manufacturer told me they didn't even tender for the estimated £5 million BTC contract because they knew liquid epoxy didn't bond to a plastic coated pipeline.

  Secondly, the Mortimore dossier included the results of a second test by Advantica on SPC 2888. Remarkably, it contradicted the earlier one commissioned by Osborne and showed that SPC 2888 exhibited poor performance. In other words, BP was told well in advance that two tests by the same laboratory showed different results for the performance of SPC 2888, yet still it was the preferred option.

  A suspicious competitor called ShawCor, also from Canada, had commissioned this second test from the same lab just weeks after the first test.

  ShawCor makes liquid epoxy. They too didn't believe it was appropriate for coating the field joints on the BTC pipeline. Instead, they offered BP a conventional and proven alternative called a Shrink Sleeve. This is wrapped around the field joint. Heat is then applied so the shrink sleeve moulds around the field joint. It has proven adhesion to plastic.

  They felt BP, through Osborne, were not interested in using a shrink sleeve at all. That is why they commissioned the second test to see how their product compared with SPC 2888 and the two other paints.

  Thirdly, we established that Osborne's specification for the field joint coating—the instructions to the contractor on how to apply SPC 2888—differed significantly from the test regime carried out at Advantica.

  Our investigation showed that in the first Advantica test Osborne commissioned, the performance of SPC 2888 was aided by the use of a wash primer. But no wash primer was included in the specification he subsequently wrote.

  The serious anomalies surrounding the testing regime of SPC 2888 had also become a cause of major concern to its British competitor E.Wood. They believed the selection process had actually been "rigged" to favour SPC 2888.

  Chris McDonnell, managing director of E.Wood Ltd, told me: "Our view is that individuals within BP were trying to cover up a decision which was clearly flawed and which we believe was influenced by an individual who was closely associated with our main competitor (SPC) . . . BP have never moved an inch."

  McDonnell said such was his firm's sense of "injustice" they were willing to risk their long relationship with BP. At the time E.Wood had other contracts with BP. McDonnell said he was used to winning and losing multi-million pound contracts but always on a level playing field.

  Consequently, in October 2002 he wrote to Jim Mooney, the senior BP manager who had approved SPC's use in the crucial meeting at Bechtel a month earlier. He also wrote to BP's compliance manager David Winter.

  McDonnell's letters suggested the first Advantica test had been manipulated. He also alleged there was "a very close association" between Osborne and SPC.[81]

  In November 2002, Jim Mooney launched an internal inquiry into E.Wood's allegations of procurement fraud and manipulation of the selection process.

  Two BP employees, Frank Ravenscroft and Tim Charters carried out the so-called "independent audit". They worked from the same building in Middlesex where the Pipeline Transportation Team of BP's Upstream Technology Group was based. This was the group that had signed off SPC 2888 as the right product to use on the field joint.

  Those who were aware of the "independent" audit included Mooney and BP compliance manager, David Winter.

  At this stage E.Wood didn't believe BP chief executive John Browne and his board knew what was going on. So McDonnell hired a lobbyist to alert more senior BP figures that his firm was considering legal action for restriction of trade, in which allegations of kickbacks would almost certainly surface.

  The two BP auditors eventually contacted Derek Mortimore on 21 November. He told them about the dossier of evidence he had left with Stretch and advised them to collect it on their trip to Baku. This they did.

  Mortimore also informed the BP auditors of the association between the SPC agent Mike Bird and Osborne. He also suggested they contact an independent corrosion engineer called David Norman, who could provide further information.

  Norman is the former head of corrosion at British Gas, where he worked for 25 years. Like Mortimore, he is now a highly regarded independent corrosion consultant who works all over the world for oil and gas companies with failing pipelines.

  Norman confirms that Bird approached him as SPC's agent in September 2001 to sell their products in India. At the time he associated Bird with Osborne's group of companies.

  Norman could also have provided BP auditors with expert evidence that the selection process was seriously flawed and that it was totally inappropriate to use any liquid epoxy paint on plastic coated pipeline.

  Norman is one of a handful of corrosion experts who when he was at British Gas developed the industry standard for testing pipeline products before they were used in the field. This standard is known as CW6.

"I've never seen a single sourced large diameter pipeline material anywhere in the world I've worked on in 40 years," says Norman.

  BP's auditors, he says, never contacted him although they had his name. This year I contacted Ravenscroft and Charters at BP, where they still work, and asked them to comment on this. They declined.

  Before working for SPC, Bird was a director of the well-established British field joint coatings firm Winn & Coles. However they confirm he was sacked in 1994, after 30 years, for highly unethical financial conduct—passing commercially sensitive information to one of Winn & Coles' competitors in South Africa. Mr Bird did not contest the dismissal, the company said and they do not accept the contention that Bird resigned, which is how he has presented the matter to Companies House.

  Trevor Osborne was contracted to BP as its materials consultant through his company Deepwater Corrosion Services (UK) Ltd. Osborne marketed this and his other companies under the corporate umbrella of the Deepwater Corrosion Services International Group or DCSI Group.

  Certainly by July 2001, while Bird was working for SPC, he, Osborne and a Mr Eddie Field began a joint venture to set up and market the DCSI Group. Field was a former director of an Osborne company. The flyer for the DCSI Group says, "quality guaranteed by director-owners". Bird is listed as the sales director. Osborne is the managing director, his boss.

  On 12 December 2002 the BP auditors produced a report. Before publication, The Sunday Times formally asked BP for a copy of the report and to discuss its conclusions. They declined. Apparently the two auditors completely exonerated Osborne and their employer.

  We were able to establish however that Chris McDonnell of E.Wood was not satisfied with BP's internal inquiry and entered into further correspondence with David Winter in January 2003. Winter maintained E Wood had not been treated unfairly and in a letter dismissed any suggestion of an undue influence by Osborne in the selection process, saying it was a team decision involving BP, Bechtel, John Brown Engineering and Advantica. He also said SPC was chosen only after full internal discussions within BP and that the oil company had made "the correct product selection."

  McDonnell had evidence that SPC was given a copy of the full and confidential first Advantica test results and was using it to drum up further business.

  Winter responded claiming he had sought and received a written assurance from SPC that they had not seen a copy of the full Advantica report. He asked E Wood to provide evidence of its allegations or the matter would be closed.

  McDonnell responded on 28 February 2003. He sent Winter a breakdown of Osborne's business interests and claimed there was a "very close association" with SPC sales agent Mike Bird. He also sent a letter from a Middle East company, Anticorrosion Protective Systems, confirming that in October 2002 the managing director received a presentation from Bird in Abu Dhabi during which he was shown the results from what was referred to as the final Advantica report.

  McDonnell was also intrigued at how SPC 2888 paint cost double its competitors per tin. He told Winter: "Your letter merely raises more questions than its answers and leads us to believe that the outcome of the whole evaluation programme was already decided before the tests began . . . The good name of BP is being dragged down by this selection process which has been a poor reflection on the reputation and integrity of one of the world's leading companies." Winter's boss, Richard Halligey, BP's commercial compliance manager, had been copied in on the correspondence.

  McDonnell says Winter never replied to his letter. He told me: "I still feel what happened (with the field joint coating bid) was totally corrupt and totally wrong. One of the weaknesses in BP is if there is a problem they close ranks." He says he would like to see an independent investigation.

  Meanwhile, Derek Mortimore's services were abruptly dispensed with in January 2003. Two months later he sought a meeting in London with Winter to press his case one final time. Instead the compliance manager agreed to pay out his contract.

  By July 2003, BP was ready to lay what it promised the people of Azerbaijan and Georgia, their governments and the Lenders Group was a state-of-the-art pipeline "built to the highest international standards."

  A wide strip of land 1,000 miles long was being cleared and levelled across the Caucasus Mountains and three countries. Sections of pipe were positioned end-to-end waiting to be welded, coated and then lowered into a one-metre trench for burial.

  Work progressed in Azerbaijan and Georgia at a reasonable pace to take advantage of the summer sun. But by September, BP was said by sources on the ground to be experiencing problems applying SPC 2888 to the field joint.[82]

  By mid-November, the greenish-grey SPC 2888 coating was cracking on the field joints, just as Mortimore and BP managers had warned it would. Work stopped for the next six-eight weeks on the Georgian and Azerbaijan sections of the pipeline as BP, Osborne and SPC frantically worked out what they were going to do.

  Unaware, one month later trade minister Mike O'Brien told the Commons the pipeline was safe, environmentally sound and millions of pounds of taxpayers' money should be risked underwriting it.

  BP is trying to blame the contractors for having applied the SPC 2888 wrongly, whilst maintaining that the paint was correctly specified and is the right product for the job. The contractors told me they didn't accept it was an application problem.

  The British firm that applied the paint in Azerbaijan and Georgia is Pipeline Induction Heating. PIH hired David Norman in December 2003 to examine the situation. He made a site visit to the Tsvalka region in Georgia where the first cracked field joints were found.

  BP, he says, must recognise that Osborne's specification was flawed because they have since modified it to try and find the solution to the cracking paint.

  This has now been temporarily achieved by adding extra heat to the field joint on application of the paint to ensure it dries in cold temperatures.

  However, Norman and other corrosion experts point out that cracking is an indication of a serious problem with the paint's chemistry and a symptom of its inappropriateness for a plastic coated pipeline that it can never protect.

  BP convened several emergency and confidential meetings in London, Tblisi and Baku between December 2003 and January 2004. Again the Lenders Group was unaware of this.

  BP compliance manager David Winter was at a confidential meeting at the Hilton Hotel on 5 January where he warned those present not to speak to the press, say sources. Winter refused to comment when I contacted him.

  We had also picked up claims from an insider that BP personnel in Baku concerned about the issues raised by Mortimore had their work phones and emails monitored as part of a leak inquiry.[83]

  By Christmas 2003 an estimated 15,000 field joints had already been coated and buried in Azerbaijan and Georgia and cannot now be regarded as safe. Exact figures are still hard to come by because BP refuses to come clean about the scale of the problem.

  Part of the reason for this secrecy is the enormous cost of putting it right and who is going to pay those costs that could escalate to hundreds of millions of dollars.

  Since the first cracks appeared BP has been privately locked in a crucial contractual battle in Georgia and Azerbaijan with its contractors—Spie/Capag/Petrofac/Amec and CCIC respectively—over the failure of the field joint coating.

  The contractors in both countries stopped work from mid-November to early February safe in the knowledge that the "standing time" following the failure of the SPC 2888 paint would have to be paid by BP. This cost alone is said to be $40 million.

  The contractors are normally to blame for a failure of this sort. But Osborne's extraordinary specification, nominating SPC 2888 as the single source product, has transferred liability onto BP.

  This is a critical battle between BP and SPC on one side and its contractors on another. Whoever loses the battle for responsibility for the cracking will ultimately be liable for the standing costs, but more crucially, also liable for the remedial repairs and any legal action the host governments may take for loss of revenue because of these delays. That figure is estimated to lie between $300 million and $1 billion.

  David Woodward is president of BP Azerbaijan. He has been in the region at the helm for five years. Before that he was in charge of the Prudhoe Bay oilfield in Alaska.

  On 15 January 2004 the BTC board met in London. Press reports suggested the key issue was preparing for the fast approaching signing ceremony of 27 credit and loan agreements for the project on 3 February.

  Since mid-January BP was briefing the media that delays in construction were down to "bad weather" and the Turkish section of the pipeline, where BP has no operational role. In late January, Woodward accompanied the Azeri deputy prime minister on a trip to Turkey to discuss the delays. The impression given to reporters was that the Azerbaijan and Georgian sections are on schedule. There was no public reference by Woodward to the major row over the field joint coatings non-performance in Azerbaijan and Georgia, and whether it should all be dug up and recoated with another product.[84]

  I visited Azerbaijan and Georgia in this period shortly before the financial closure ceremony at the presidential palace in Baku.

  BP managers like Stretch appeared too frightened to speak out. Another said the whole matter was "very sensitive." Stretch reported my approach to BP. The oil company was aware of my presence because I had also approached them. They declined to facilitate a site visit to where the pipeline was being welded and coated claiming bad weather conditions and denying there was any work stoppage on the pipeline.

  Back in the UK, just before publication, BP, Trevor Osborne and Mike Bird of SPC declined to answer any detailed questions. In fact, Osborne and Bird referred all questions to BP.

  The Sunday Times investigation had shown that before financial closure of the $2.6 billion loan agreement BP did not tell the ECGD about:

—  BP managers and Derek Mortimore's concerns over the use of a liquid epoxy on a critical part of the pipeline.

—  The Mortimore dossier and report in August and November 2002.

—  The second Advantica test casting doubt on the choice of SPC 2888.

—  The allegations by E. Wood of procurement fraud and rigging of the selection process.

—  The internal audit and its report.

—  The cracking of the field joint coating.

—  The six-eight week stoppage in the run up to the financial closure ceremony in Baku.

  When I spoke again to the ECGD press office just before publication at first a spokesman tried to claim they were aware of the allegations. He withdrew this suggestion and accepted they had not been told.

  Consequently, it appeared that BP had misled the ECGD and the ECGD's own monitoring systems had failed. ECGD officials then misled their minister, Mike O'Brien, who in turn unwittingly misled parliament.

  We asked whether the ECGD would now investigate its failure to discover the problems with the pipeline, which on any reading had a "material adverse affect" on the BTC project.

  Remarkably, we received a statement from the ECGD that attempted to minimise the field joint coating problem and suggest the agency had "satisfied all its due diligence procedures in coming to a decision to provide cover for the project."

  The statement also stated the problem was being dealt with by BP "as part of an ongoing testing regime which is designed to ensure that the pipeline passes all required tests before it is certified operationally fit."[85]

  After the article was published, an additional ECGD statement to the NGOs was sent to me. It said:

"We do not agree that The Sunday Times article provides evidence of shortcomings in ECGD's diligence with regard to the BTC pipeline."

"The article's central allegation is that the BTC Pipeline Co failed to disclose an assessment that "discovered `serious flaws' in the pipeline's design, which would make it highly likely to leak."

"We do not dispute that BTC Co did not inform ECGD of problems with the field joint coating material: this was because its testing and use is regarded by both parties as a routine part of the pipeline construction process. This was not a problem that has had a materially adverse effect on the project, and as such neither BTC Co nor BP was under any obligation to report this to us. This was reflected in the brief duration of the halt in work on the pipeline, which remains on target for completion in 2005 and on budget.

"BTC Co and BP have now provided us with more information about the coating. They have assured us that any problems that have arisen with the field joint coating are being dealt with routinely in the course of work. These problems were detected by the ongoing inspection regime in place on the pipeline—evidence that the monitoring regime is doing its job.

"The article also made allegations about the procurement process that led to the selection of that coating. BTC Co made their choice as a result of an extensive technical exercise, which concluded that the SPC coating was the most suitable. The choice of coating was scrutinised and approved by Parsons Energy & Chemicals, the independent engineering company operating on behalf of the project lenders. BP has also shared the outcome of its internal investigation into the procurement of the field joint coating material with ECGD, which concluded that any allegation of impropriety was unfounded."

"We remain satisfied that all our due diligence procedures were followed before coming to a decision to provide cover for the BTC pipeline project. At present no monies have been drawn down under the financing arrangements established for the BTC pipeline. The drawing down of monies in respect of which ECGD has agreed to provide cover prior to the pipeline's completion is subject to ongoing scrutiny."

  It strikes me that the ECGD's only defence to its due diligence failings is to ally itself with BP and suggest the problems with the BTC pipeline exposed by The Sunday Times were "routine" and the oil company was therefore under no obligation to report them before financial closure.

  Such a stance does not bear scrutiny and begs the question, why would BP voluntarily refer to the ECGD "routine" problems concerning the welding in Turkey but not those that led it to carry out an internal audit into procurement fraud and a work stoppage in Azerbaijan and Georgia?

  There is nothing routine about the problems with the field joint coating. And if the ECGD monitoring system was truly independent and functioning properly it would have known that the selection of the coating is one of the most important technical decisions on the project. Certainly BP managers thought so—that is why they employed Derek Mortimore to review Osborne's specification.

  It is also very telling that the ECGD is quite willing to accept BP's unsupported assurances, yet is unwilling to attend a briefing by Mortimore and other corrosion experts with a combined one hundred years of experience.

  It also appears from its new statement that the ECGD is relying on Parsons Energy & Chemicals' review to explain why the field joint coating is safe. On 25 February , I spoke to the manager of the Pipeline Group at Parsons responsible for overseeing the monitoring work for the Lenders Group.

  The ECGD said Parsons had "scrutinised" the choice of the field joint coatings. This does not accord with what Parsons told me:

—  The Parson contract was "to review the design of the facilities, specifications, plans, execution plans and preparatory work prior to starting construction." The contract started in 2002 and finished in January 2004.

—  Parsons did no on site inspection for the Lenders Group after The Sunday Times published its story.

—  The reviewer was never permanently based in Azerbaijan or Georgia. He was not a corrosion expert.

—  The review was a paper exercise. It was not a quality assurance exercise.

—  The field joint coating specification was one of over 100 other specifications that Parsons reviewed.

—  The review didn't guarantee the integrity of the field joint specification for the pipeline's estimated design life of 40 years.

—  Parsons carried out no independent investigation into the coating or the selection process controlled by Osborne.

—  Nor did Parsons look at how any specified products were applied in the field.

—  Parsons determined SPC 2888 was an industry standard coating, but couldn't name one major pipeline where it had been used on plastic.

—  The Parsons reviewer was in Azerbaijan between late 2002 and early 2003. But BP did not disclose to him the Mortimore report nor did he see any background material behind each specification or internal BP reports.

—  BP/BTC imposed on Parsons a confidentiality agreement "not to divulge information" to third parties. BTC paid Parsons for the work it did for the Lenders Group. BP is also a client of Parsons.

  It would be an abuse of the English language to suggest this review was a thorough scrutiny of the field joint coating selection process. This is no criticism of Parsons. They say that was not what the Lenders Group commissioned them to do. In effect it was a box ticking exercise, with BP picking up the bill.

  Given the major problems identified by The Sunday Times investigation, it is my firm belief that before £58 million pounds of taxpayers' money is risked any further there should be several full and independent investigations.

  One investigation should look at the science underpinning the claims of BP and the ECGD. This is urgently required because information I have very recently received suggests BP is planning to continue using the problematic Canadian paint. In fact I was told BP will shortly start digging up some of the pipeline already buried in Georgia and recoat the field joints with SPC 2888.

  Corrosion experts I have consulted are appalled at this strategy and say this is like worrying what colour the Titanic is. The only way forward, they agree, is an independent review of the entire field joint coating system for the pipeline. This scientific review must be done by independent laboratories in circumstances not controlled by BP or any of its agents.

  A second parallel investigation should determine how the ECGD got it so wrong and in what way BP misled the government to secure the $2.6 billion loan and credit facility.

  Before publication in the Sunday Times we spoke with you, Mr O'Neill, as committee chairman. You agreed to raise the matters with the ECGD and said the committee would examine whether BP has fulfilled its obligations under the terms of the loan and if they had done so with sufficient candour.

  Derek Mortimore and David Norman have asked me to inform you that they and others would be willing to appear before your committee to provide evidence that the £58 million of taxpayers' money is at great risk if the BTC project is allowed to continue on its current path.

  As a professional journalist for some time I have been struck by the depth of concern felt by those industry insiders who have spoken to me about the BTC project.

  These are not people I have found to be ideologically opposed to the extractive industries or who have allied themselves with campaigners opposed to the BTC pipeline. On the contrary they want to see the pipeline built, but properly.

  A good pipeliner is a good environmentalist because he understands that the field joint coating is the first line of defence against corrosion and subsequent environmental and human disaster. When they are not helping to build them, they spend their lives putting right failed pipelines. I have learned that the most frequent cause of ruptures is a failed coating.

  Equally important issues surround the probity of the selection process. These should not be allowed to rest with the so-called "independent" audit BP refuses to disclose.

  The former chairman of Shell, Mark Moody-Stuart, once said, "We are moving from a trust me world to a show me world." The ECGD, however, is satisfied with BP's word that the internal audit, which it wasn't aware of until the Sunday Times mentioned it, exonerated the oil company. Mantra like and without seeing it the ECGD now promulgates the view this was an "independent" audit and therefore no further action is required.

  In June 2003, the Corner House criticised the ECGD for "turning a blind eye" to corruption allegations in those mega project it underwrote overseas. The ECGD responded that it had just signed a memorandum of understanding with various agencies about referring such allegations.

  In the public interest, the selection of SPC 2888 as the single source product for the BTC field joint coating should now be referred to the Serious Fraud Office or another appropriate agency for a full investigation.

Michael Gillard

15 March 2004

Second memorandum by Michael Gillard

  1.  The WorleyParsons desktop study (WP report) is the document that the Lenders Group of government export credit agencies, development and commercial banks is already pointing to as proof the BTC pipeline is safe and environmentally responsible. The ECGD, World Bank and European Bank of Reconstruction and Development also hope this document will re-establish their reputation as a genuine and effective watchdog. For BP, as operator and main shareholder in the BTC pipeline, the WP report is being used to re-establish its credibility with host governments, especially Georgia.[86]

  2.  In his covering letter to the Committee dated 19 July 2004, John Weiss of the ECGD claims the WP report "deals effectively" with the field joint coating issues raised by veteran pipeline integrity consultant Derek Mortimore, the BP Baku managers he worked for, and The Sunday Times investigation.

  3.  Mr Weiss goes on to say the WP report "indicates the choice of coating was carried out after extensive consideration of its suitability, and by appropriate experts. It shows that the coating was subjected to rigorous testing, and has a track record in similar conditions in a significant number of other pipelines."

  4.  This submission will seek to persuade the Committee that the WP report does nothing of the sort. In fact, it makes the case for an immediate and independent audit of the BTC pipeline before it is completely buried with a major design flaw.

  5.  Mr Weiss also concludes "on the basis of independent expert advice received, ECGD does not consider the field joint coatings to be a serious ongoing environmental issue."

  6.  This submission will explain why there is no cause for reassurance. The £60 million of taxpayers' money pledged by the ECGD is still at serious risk and BP's attempts at self-regulation, to which the government gave free reign, have failed and are hopelessly marred by conflicts of interest, corruption allegations and evidence of a cover-up.

THE TIMING

  7.  The Committee chairman called for a report on 11 May 2004. By this time the Lenders Group had already commissioned WorleyParsons on 17 March as a result of The Sunday Times' Insight team expose of the field joint scandal on 15 February.

  8.  WP sent its un-redacted report to the Lenders Group before May. Nevertheless, it wasn't passed to the Select Committee until late July. By then, Parliament had risen for the summer recess and the opportunity to consider the report's contents in oral committee hearings was lost until after the autumn. Meanwhile, BP has been forging ahead with improper haste in Georgia and Azerbaijan to complete a pipeline that has a major design fault. BP managers and contractors receive bonuses for finishing on schedule.

  9.  I understand from the Committee there was some "haggling" between it and the ECGD, with the export credit agency claiming that "commercial confidentiality" prevented it from disclosing certain information. In fact, WorleyParsons confirms that during this time they were also having "discussions" with the Lenders Group and BP over who not to "embarrass".

THE REDACTION

  10.  ECGD has confirmed to me that the redaction process was done with BP, although the government department says it has no details of how it was carried out.

  11.  Jim Powers, the head of WorleyParsons pipelines division, oversaw the desktop study as part of the "construction monitoring" for the Lenders Group. In an interview on 7 September 2004 he told me WorleyParsons was asked to do "modifications" and that the report went back and forth between his team, BP and the Lenders Group.

  12.  He said: "I'm sure that there were some things that were taken out. Probably names that might have been embarrassing to people or any confidential information...(The Lenders Group) asked for input on what might be embarrassing to individuals." He didn't name them.

  13.  ECGD tell me they have no plans to make the un-redacted version available to Parliament. I asked BP what role any of the following employees had in drafting or redacting the WP report:

—  Trevor Osborne (BP materials consultant)

—  David Winter (BP Compliance manager)

—  Richard Halligey (BP's head of compliance)

—  David Fairhurst (BP Corrosion chief)

—  Jim Mooney (BP Upstream Technology Group)

—  Kevin Muller (BP Global SCM Integrity Assurance Manager)

—  Jaime Potes (BP London Project Manager)

  All these men figure in the events leading up to the selection of SPC 2888 as the field joint coating in October 2002 and/or the remedial strategy after cracked field joints were discovered in November 2003. BP refused to answer this or other simple informational questions.[87]

THE INDEPENDENCE

  14.  Last March, Foreign minister Mike O'Brien claimed in a letter to Corner House that Parsons (as they were known before the merger with Worley) had "scrutinised and approved" the selection of SPC 2888 before the financial closure ceremony in February 2004, when the BP-led consortium received US $2.6 billion in multilateral loans.

  15.  I refer the Committee to my earlier submission where, in summary, Mr Powers confirmed that the Parsons review supported Mr Osborne's field joint coating specification but involved no independent assessment of the coating he selected for BP or the quality of that selection process.[88]

  16.  More recently, Mr Powers confirmed to me that the same project manager involved in the first Parsons review was also involved in the second one. He refused to name him for fear that any subsequent interview would seek to elicit "discrepancies".

  17.  It is clear that before writing this latest report WP had a position to protect on the field joint coating issue. That position was identical to their client's—the Lenders Group—and to BP/BTC Co, who has paid for both WP reports.

  18.  The latest report by WP cannot therefore be considered in any way "independent", as Mr Weiss repeatedly claims in his covering letter to the Committee.

THE DISCLAIMER

  19.  The illusion of independence becomes even more apparent when one considers the extraordinary disclaimer WorleyParsons has inserted on page 2 of its report.

  20.  Mr Weiss claims in his covering letter that although his department is reliant on BP for information, ECGD also relies on their "independent consultants" WorleyParsons, to "verify" that information. Yet here we have WP making it clear they have "not independently verified that (the information from BP) is comprehensive, complete, accurate or up to date."

  21.  Four of the most alarming failures in the WP report are:

—  Not interviewing those people they nevertheless choose to rubbish solely on the word of BP and SPC.

—  Not interviewing those who could challenge the technical arguments from BP and SPC.

—  Not obtaining all documents in the possession of BP and SPC that would cast doubt on their assertions.

—  Not doing a site visit before publishing the report.

  22.  I return to my recent interview with Jim Powers of WorleyParsons. He said the "instructions" from his client, the Lenders Group, was to only carry out a desktop study. This only involved, he said, asking BP and SPC for all "pertinent" documents concerning the field joint coating issue raised by The Sunday Times. He was thus reliant on the integrity of BP and SPC to disclose all the relevant documents. "You can't ask for things you don't know exists," he maintained.

  23.  Mr Powers went on to say: "We did what we were instructed to do. And you can call (the Lenders Group) to find out why they didn't go into it further or want it gone into further."

  24.  This statement provides an extraordinary insight into the approach taken by the Lenders Group after its much vaunted due diligence procedures were exposed by The Sunday Times. They went on to limit the investigation of the field joint coating issue to a simple desktop study wholly reliant on one partial source of information, which the Lenders Group deliberately chose not to independently verify.

  25.  Why would ECGD and its partners in the Lenders Group impose such a limited remit on this vital investigation, notwithstanding that WorleyParsons was the wrong company to do it? What did the Lenders Group have to lose from a full and truly independent investigation? Did the ECGD not want to hear evidence from independent sources that would expose the inadequacy of their due diligence when gambling with taxpayers' money?

THE EXECUTIVE SUMMARY

  26.  WorleyParsons' key finding is that the evaluation, testing and selection of SPC 2888 was "very thorough". With these two words, BP, SPC, Mr Osborne and the ECGD are let off the hook.

  27.  Of course the PR value of such a finding for BP and the government is plain to see. But in reality it is worthless. WorleyParsons is making a finding of fact, when in the same breath it tells us nothing this finding is based on has been independently verified as true.

  28.  The finding relies on eight tests provided by BP and SPC. On the face of it this looks convincing. But on closer examination of when these tests were carried out, it emerges that the selection process controlled by Mr Osborne and overseen by senior BP managers was not just seriously flawed but arguably reckless.

  29.  In the course of my investigation I have consulted widely with paint coating companies in the UK and the US who supply the oil industry; corrosion and coatings consultants; pipeline integrity specialists, field joint coaters and major oil companies.

  30.  The consensus concerning best practice for the evaluation, testing and selection of field joint coatings is unanimous, logical and very simple. The client (in this case BP) has the responsibility to specify material(s) that have been fully tested to ensure full anti-corrosion protection in the conditions the pipeline will operate in for its entire design life (40 years). Once these tests have been carried out and the client is satisfied, only then is the specification or instructions to the applicator, issued.

  31.  It has already been established from BP's own reports that the use of a liquid epoxy field joint coating like SPC 2888 on a polyethylene (plastic) coated large diameter pipeline like BTC is "a step change" in industry practice. In other words, experimental engineering.

  32.  Extra care in the selection process was therefore required. Some industry sources have suggested that full lab tests followed by full field trials on a pipeline in simulated weather conditions and then a further lab test on a cut-out section of the coated pipeline would constitute best practice before specification. This is even more necessary when the committee considers that Mr Osborne and BP nominated SPC 2888 in their October 2002 specification to the exclusion of all alternatives. In other words, the specialist contractor applying the coating to the field joint was forced to use SPC 2888 and had no opportunity to rely on a fully-tested fallback material if the chosen coating failed, which of course it did.

  33.  The key point is this: All but one of the eight tests were carried out long after Mr Osborne had specified SPC 2888 as the single source field joint coating (October 2002).

  34.  Three of the eight tests were carried out by SPC on its own product. These looked at its appropriateness when applied to plastic coated pipeline. Significantly, the Canadian paint company only did these test after Mr Osborne had awarded it the contract (October 2002); after the field joint coating application had already begun (August 2003); and just a few weeks before their coating cracked (November 2003).

  35.  A number of issues arise:

—  Why didn't Mr Osborne ensure these tests were done before he selected SPC 2888 as the single source field joint coating?

—  Why did BP compliance manager David Winter and Corrosion chief David Fairhurst allow Mr Osborne to warrant a product as fit for purpose and therefore environmentally responsible before all these crucial tests had been completed and proved it so?

—  How was this massive failure of BP's quality control and assurance programmes possible, especially after Derek Mortimore had warned the oil company between July 2002 and March 2003?

THE MISSING REPORTS

  36.  Besides WP's failure to verify information given to it by BP and SPC, there has been a repeat failure by BP to disclose key reports that would have forced WP at the very least to investigate the selection process more effectively.

  37.  In my earlier submission I explained how Jim Powers confirmed that for the original pre-loan closing Parsons review in March 2003, BP had failed to disclose any of Derek Mortimore's reports on the field joint coating; the correspondence between BP Baku (who had contracted Mr Mortimore) and the BP Design Group in London (who were pushing SPC 2888); and the internal audit report into procurement fraud allegations.

  38.  When I spoke recently with Mr Powers he said: "It's likely that all the reports Mr Mortimore produced were not seen by us. I don't know how many reports he submitted." The Lenders Group only asked WP to review the one report quoted in the Sunday Times article, Mr Powers revealed.

  39.  One of the eight studies relied on by WP is a lab test carried out by Advantica in July 2002. It is on this test that BP and Mr Osborne have based their entire selection of SPC.

  40.  In my earlier submission I explained that it was this test that led one of the unsuccessful competitors, the British coatings firm E Wood, to allege Mr Osborne had "rigged" it.

  41.  You will be aware that Mr Chris McDonnell of E. Wood Ltd is making a direct allegation that Mr Osborne "slanted" the ranking and weighting system in favour of SPC 2888. BP was forced to conduct a secret internal audit following Mr McDonnell's procurement fraud allegations. The oil company refuses to release the report to The Sunday Times, but claims it fully exonerates Mr Osborne and others. However in my earlier submission I also explained how key witnesses were not interviewed and evidence was not collected or acted on by the internal BP auditors.

  42.  I also revealed there is a clear conflict of interest between Mr Osborne and the Canadian firm Speciality Polymer Coatings, makers of SPC 2888, through their British agent Mike Bird, who at the relevant times was setting up a materials supply and corrosion consulting company with Mr Osborne. He doesn't dispute this. In his defence Mr Osborne says the joint venture never got off the ground. But it did get to the point of preparing and distributing promotion literature. The fact still remains that he and Mr Bird, who had been sacked from his previous company for disreputable conduct involving a competitor, were planning a business partnership in the run up to the selection of SPC 2888.

  43.  In my earlier submission I also explained that ShawCor, another unsuccessful participant in the Advantica test, was so concerned about the selection process that just a few weeks later it commissioned the same lab to retest the materials. On this occasion SPC 2888 failed.

  44.  Once again, BP has not disclosed to WorleyParsons these Advantica test results, notwithstanding that Mr Mortimore ensured a full copy was available to the internal auditors. Mr Powers of WorleyParsons told me: "If we didn't mention them in the report we weren't aware of them."

THE ADVANTICA TESTS

  45.  The "confidential" Advantica test (5426) in July 2002 is the only one BP can point to as evidence that it carried out any evaluation before selecting SPC 2888 as the field joint coating. The WP report says this Advantica test was "rigorous".

  46.  Advantica is the former British Gas laboratory. It is agreed that Trevor Osborne was responsible for choosing the products that would be tested and also setting the parameters of those tests and then weighting and ranking the test results.

  47.  Mr Osborne selected three liquid coatings. They were SPC 2888 (a urethane modified liquid epoxy made by SPC from Canada); Copon 165 (a urethane made by E. Wood Ltd from the UK) and Protegal 3210 (a tar modified urethane made by Goldschmidt from Germany).

  48.  Mr Osborne also included another type of field joint coating called a shrink sleeve. Unlike the liquid coatings, which are sprayed or hand applied with roller to the field joint, the shrink sleeve is wrapped around it and then moulded on by heat. There are many shrink sleeve types. Mr Osborne however chose Canusa, manufactured by ShawCor, also from Canada.

  49.  BP approved Mr Osborne's test programme, which was carried out by Ian Thompson and Colin Murdoch from Advantica. Ten performance tests were conducted on samples of each of the four materials pre-applied to a plastic coated pipeline exactly like the one intended to be used for the BTC project.

  50.  The tests included cure, adhesion, hardness, impact, flexibility, abrasion and water absorption. The lab made an alarming finding strangely missing from the WP report. The Advantica scientists wrote that, "No single coating material consistently performed well in all tests."

  51.  In the ranking and weighting table devised by the projects team certain key test results were omitted. For equally unclear reasons the ShawCor shrink sleeve was also excluded from the ranking table.

  52.  SPC 2888 subsequently came first over Copon and then Protegal. The Advantica test results said spray-applied SPC 2888 outperformed the others on penetration/indentation, gouging and abrasion, but says it performed "poorly" on impact and water absorption. The hand applied SPC 2888 also tested poorly on flexibility.[89]

  53.  Although the Canusa shrink sleeve was tested with the liquid coatings, Mr Osborne excluded it from the ranking table. ShawCor suspected this was because it would have been shown to outperform SPC 2888. Consequently, that same month, July 2002, ShawCor commissioned the same Advantica scientists to test and rank SPC 2888, Protegal and the Canusa shrink sleeve for adhesion and flexibility to plastic coated pipeline. The results were remarkable and worryingly at variance with the first Advantica test.

  54.  During the Advantica test (5405) for ShawCor, SPC 2888 was shown to exhibit the worst adhesion of the liquid coatings after it had been dried (cured) and when it had been soaked in hot water for 28 days. Advantica failed SPC 2888 under the UK industry CW6 standard. Under the Canadian CSA standard, Advantica ranked SPC 2888 with a 5, a "FAIL" defined as "the coating can be removed as a single piece."

  55.  During the Advantica test (5435) for ShawCor, SPC 2888 was also shown to exhibit very poor flexibility on plastic coated pipeline. It cracked at just 1.0% strain, when the minimum pass under the CW6 standard is 3.0%. It also delaminated from the plastic coating.

  56.  What conclusions are we to draw from this information? At the very least it puts a significant question mark over the Advantica test which BP, Mr Osborne, SPC and WorleyParsons are now relying on as proof that the chosen coating was properly selected and is the best one for the BTC pipeline. At its highest, this information proves that SPC 2888 was scientifically shown to be an inappropriate coating for the BTC field joints as early as July 2002. BP's internal auditors were made aware of the alternative Advantica tests in November 2002. But by then SPC 2888 had been selected by Mr Osborne.

  57.  BP ignored these results and continues to do so after the coating cracked in Georgia and Azerbaijan a year later.[90]

THE PEER ASSIST

  58.  Weeks after the Advantica tests, on 5 September 2002 a Peer Assist Review by BP in London signed off SPC 2888 during a meeting in the BTC Project Office at Bechtel. Those in the Peer Assist team were Jim Mooney (BP Upstream Technology Group construction consultant) and Kevin Muller (BP Global SCM Integrity Assurance Manager) with attendance from the BP projects team consisting of Ian Parker, BP's design manager, Jaime Potes, Carl Slater, Steve Evans, David Willis and Trevor Osborne.

  59.  The minutes of the Peer Assist meeting totally undermine repeated claims by minister Mike O'Brien and the ECGD made after the field joint scandal was exposed that SPC 2888 has a proven track record on plastic coated (PE) pipelines, especially those operating in cold climates.

  60.  Mr Weiss said in his covering letter that SPC "has a track record in similar conditions in a significant number of other pipelines." Yet the BP Peer Assist report on several occasions made the point that SPC as a field joint coating system with plastic coated pipeline was accepted to be "a step change in present industry practice, both technically and commercially". It also says "the selected field joint system has previously only been proven on FBE (Fusion Bonded Epoxy) systems." Later on in the same minutes it said: "Whilst this clearly looks like an attractive field joint coating option, it is relatively unproven on three-layer PE systems."

  61.  Furthermore, Ian Parker, who was part of the Peer Assist meeting one year later was quoted in the BP Frontiers magazine saying: "As far as we know the first time such a system has been employed." Even the WP report concedes, "SPC coating material may not have been used on projects with three-later PE coated line pipe".

  62.  For these very reasons the Peer Assist team made two sensible and precautionary recommendations:

—  "(a) full size field trial of SPC 2888 to ensure repeatable performance under field condition", and

—  further field-testing on alternatives to SPC 2888 as a "fall back if problems are experienced."

  63.  One month later in October 2002 Mr Osborne/BP issued the controversial specification nominating SPC 2888 as the only field joint coating. Neither of the two tests required by the Peer Assist team had been carried out.

  64.  The WP report refers to a "field demonstration of the application of SPC field joint coating material." British company Pipeline Induction Heating carried this out in January 2003 in Algeria. The WP report suggests, presumably on BP's word, that this test was in conformance with the Peer Assist requirements. This is not the case.

  65.  If WorleyParsons had spoken to PIH it would have been told the Algeria test was not a full size field trial but a demonstration of how PIH's specialised equipment would apply SPC 2888 in the field. They had been awarded the contract to coat the field joints in all three countries.

  66.  The reality is that no proper field test was carried out until December 2003, after SPC 2888 had failed as a field joint coating and cracked in Georgia and Azerbaijan. Only then did Mr Osborne oversee a field test by PIH on selected field joints.

  67.  As to the second precautionary test the Peer Assist team required, I have spoken with various sources on the BTC project and can find no evidence that BP or Mr Osborne ever conducted field-testing on alternatives to SPC 2888.

  68.  The only alternatives to SPC 2888 which Advantica tested were in July 2002. But the Peer Assist team was already aware of that when they asked for further tests as a "contingency plan if the performance (of SPC 2888) is not achieved." This never happened so when the field joints cracked in November 2003 there were no "fall back" coatings materials. This is precisely because Mr Osborne had nominated SPC 2888 as the single source coating.

  69.  Furthermore, I am informed that Raychem, a major field joint coating manufacturer, was asked in 2002 to make a presentation to BP Baku on the viability of shrink sleeves over liquid epoxies on plastic coated pipeline. Raychem presented to Technical Manager Paul Stretch and senior project engineer Rod Hensman, both from BP Baku, who had already employed Derek Mortimore to look at Mr Osborne's specification because of their concerns about SPC 2888.

  70.  Additionally, Mr Mortimore had organised a testing regime in Dunkirk, France where four companies would demonstrate alternative field joint coatings for plastic coated pipe. One of the four firms was Eupec, who were responsible for the "Blue Stream joint" for a gas pipeline from the Black Sea. But when the BP Design Group in London (composed of Mr Osborne, Mr Winter and others) found out about this trial they ordered BP Baku to cancel it, says Mortimore.

  71.  A senior source in Raychem told me the main BTC contractors had complained to them they were being forced to use SPC 2888 and felt it wouldn't work. Raychem then spoke to BP corrosion chief, David Fairhurst, who informed them shrink sleeves were not being considered as an alternative to liquid coatings.

  72.  The WP report makes no mention of any tests by BP on alternatives to SPC 2888. And BP has not provided any evidence it complied with this second safety requirement of its own Peer Assist team.

THE SELECTION PROCESS (ADHESION)

  73.  The committee will have heard a lot of technical terms in this inquiry. So for clarity I have included a picture (Appendix 1) of the BTC pipeline that I took earlier this year in Azerbaijan.

  74.  The photo shows the field joint after welding but before the coating has been applied. It shows the three different surfaces SPC 2888 must stick to in order to provide full anti-corrosion protection over forty years.

  75.  The black material on either side is the plastic coating (or polyethylene jacket) of the main pipeline. The plastic is made up of three layers, which is why it is sometimes referred to as "three layer, high density polyethylene". These three layers are (from the outside in) plastic, adhesive and fusion-bonded epoxy (FBE), the green material in the photo, which lies on top of the bare steel. The brown material is the bare steel.

  76.  When each 11.5 metre section of the pipe comes from the factory the plastic jacket should be cut back a little at each end to reveal some bare steel for welding and a 50 mm wide ring of FBE (called a FBE toe).

  77.  The SPC 2888 field joint coating is being applied by spray using special equipment that rotates around the pipe. It is unanimously accepted in the pipeline world that liquid epoxies like SPC 2888 have good adhesion to steel and FBE. Equally, it is widely accepted that liquid epoxies have very poor adhesion to plastic and therefore carry risks that can be removed by using another proven type of field joint coating system.

  78.  The most critical point is the overlap area where the SPC 2888 must bond to the black plastic on either side of the weld. If it doesn't, and the SPC 2888 disbonds, then you get water ingress between the two that will eventually travel towards the bare steel and start the corrosion process.

  79.  The ability of the SPC 2888 coating to stick to plastic is therefore the most crucial test that any proper evaluation process should prove before selection, let alone before making it a single source nomination. It is truly alarming that the WP report confirms Advantica only carried out this test in January 2003 after Mr Osborne and BP had awarded SPC the contract.

  80.  The committee might ask whether the earlier Advantica test in July 2002 (referred to in Points 45-57) looked at adhesion when it compared SPC 2888 to the two other liquid coatings before selection? The answer is yes but BP is not readily revealing the conclusions. Why not? Because they showed there was a big "question" over SPC 2888's long-term adhesion to plastic.

  81.  I have seen a BP document prepared soon after this Advantica test in July 2002. I believe it supports Mr Mortimore's contention that there was a "push" to choose SPC 2888, despite recognised doubts about its ability to do the job. The document also appears to show that the ranking and weighting system adopted by Mr Osborne was wholly unscientific, and again demonstrated a bias for SPC 2888.

  82.  The document states:

—  The results for adhesion to plastic, curing and water absorption were "omitted" from the Advantica ranking and weighting table that gave SPC 2888 the edge.

—  The adhesion of SPC 2888 to plastic "diminishes rapidly"/"dramatically reduced" during temperature changes. This was confirmed by SPC's own follow-up tests.

  83.  The document ends by saying: "There is a question regarding the SPC product and its adhesion to (plastic), which needs to be resolved." However, the WP report confirms that this question was not revisited until a further Advantica test in January 2003, six months later. By which time SPC had been awarded the contract and had been written in to the specification of October 2002.

  84.  The Osborne specification, authorised by BP, is a remarkable document. On the issue of adhesion it allows the contractor and on-site inspectors of the field joint to pass what in normal circumstances would be considered a massive failure in performance. On page 13 of the specification it says, where the SPC 2888 is found to "peel in large pieces adhesively" from the plastic this should be considered "a pass". Corrosion consultants I have spoken to, other than Mr Mortimore, are aghast at this and say if the coating peels so easily then it is failing to adhere and cannot therefore be considered fit for purpose for one minute let alone 40 years.

  85.  The adhesion test Advantica did for ShawCor in July 2003 "failed" SPC 2888 on this very point and after a 28-day hot soak. (See Point 54) All of which might explain why the adhesion test Advantica carried out for BP in January 2003 contains no glowing endorsement of SPC 2888. The WP report chose merely to extract one sentence from it, which states a known proposition: that SPC 2888 will adhere better if the plastic surface is slightly roughened. This is by no means a scientific endorsement that it will stick permanently for the estimated design life of 40 years. Subsequent events have shown it hasn't. I have photographs of the SPC 2888 coating coming off in chunks leaving the bare steel exposed.

  86.  The WP report refers to a Technical and Risk Assurance Process (TRAP) carried out in January and May 2003 on the field joint coating. It makes the very interesting concession that the SPC 2888 coating has "limited adhesion" and "slow curing" when applied to plastic coated pipeline. What is extraordinary about this statement is that it was also being made seven months after the SPC coating had been selected.

THE SELECTION PROCESS (CURE)

  87.  The committee will by now be aware that SPC 2888 coating was discovered to have cracked on the field joint in November 2003. BP claims this was a cure problem, which has now been solved. Cure here means how to dry the coating on the field joint in changing ambient temperatures, which go from warm (when the coating is applied during the day) to cold (as it tries to dry overnight). This fluctuation in temperatures is known as thermal cycling.

  88.  The curing of the SPC 2888 is aided by heating the field joint immediately before and after the coating has been applied. It is universally recognised by the industry that it is the responsibility of the client (BP) to ensure it has developed the appropriate curing regime for the selected material. This is especially the case when nominating a liquid epoxy like SPC 2888, which had well known problems of curing in cold temperatures. BP would also have known that Georgia in particular gets very cold in winter both above and below ground, especially in the Caucasus Mountains.

  89.  Yet BP admits in the WP report that it did not know how to cure (dry) the SPC coating when it issued a specification for a field joint that was supposed to last 40 years. Furthermore, BP, Mr Osborne and SPC only developed a curing regime for SPC in December 2003, some 17 months after the spec was issued and one month after the cracking had been discovered.

  90.  The WP report confirms that BP started coating the field joints with SPC 2888 in August 2003 without having any idea how to cure the paint in cold weather. The on-set of winter was just weeks away. Industry experts I have consulted say this is plain reckless. And finding out how to eventually make the paint dry has nothing to do with making the field joint work for forty years. BP projects were aware of a lab report by TransCanada, an energy company responsible for 38,000 kms of pipeline in bitterly cold areas of North America. In the late nineties they had tested the viability of using liquid epoxies on plastic coated pipe. It failed EVERY test. The TransCanada report concluded that the two materials were "not compatible". Furthermore, it disputed the canard BP and the ECGD have now come up with: that having sorted out how to dry the paint ipso facto there is no problem and the field joint is safe and fit for purpose. Far from it. TransCanda's leading scientist wrote: "The coating may be applied and (dry) to look acceptable, but once (buried) and operation of the pipeline is started, the coating will fail (on the overlap area onto the plastic coating)."

  91.  SPC only did a cold weather test on 26 August 2003 and a cure test in September 2003. These tests are designed respectively to see how the field joint coating behaves in cold weather conditions, and to see what sort of heating regime is required when applying the coating to the field joint to ensure it dries correctly and is therefore not liable to crack and expose the steel to corrosion when buried in moist soil for 40 years.

  92.  These SPC tests were done without BP present. Similarly, PIH were not invited to witness this test, despite the fact it was already applying SPC 2888 to the field joints in Azerbaijan and Georgia. The results of these SPC tests have not been made public. And the WP report gives no indication of how SPC performed. What we do know is that the coating cracked on site weeks later in November 2003.

  93.  Derek Mortimore had pointed out back in November 2002 the dangerous lack of cure information in the BP specification. Exactly one year later the coating cracked.

  94.  If, as it appears, BP was determined to use SPC 2888 come what may then it was incumbent on them to ensure that Mr Osborne and his team selected it based on the highest standard of pre-qualification testing. The WP report, when stripped way, proves Derek Mortimore's warning that the selection process was at the very least "seriously flawed, underdeveloped and incomplete".

THE CRACKING

  95.  The WP report says BP became aware of the cracking problems sometime in November 2003. BP declined to tell me exactly when. The date is important because between 4 and 11 November the International Finance Corporation and the European Bank for Reconstruction and Development approved US $250 million loan each to the BTC project. The cracking had occurred approximately three months before the crucial formal financial closure ceremony for the full US $2.6 billion multilateral loan, marking the end of the BTC consortium's spending. From that point on the Lender Group would finance construction with loans underwritten by the taxpayer.

  96.  On discovery of the "many" cracks BP did some preliminary checks and determined it was down to "improper application at low ambient temperatures". In other words, BP put all the blame on PIH.

  97.  BP then "suspended" the field joint coating process and formed an "investigation team" made up of SPC, the BTC Projects Team and UTG in London. In other words, the very people whose failures of quality assurance and control had led to the cracking. Unsurprisingly, this investigation team reaffirmed the early decision that the cracking was an application problem. BP will not say when this decision was made.

  98.  Over the following eight weeks, when all field joint coating stopped, BP, Mr Osborne and SPC determined yet another new heating regime to cure the failed coating in cold weather. New corrective instructions were issued to PIH on 9 February 2004, one week after BP had received the multilateral US $2.6 billion loan.

  99.  On the basis of 87 field joints per kilometre of pipeline there are 150,000 field joints on the project in Azerbaijan, Georgia and Turkey. There are 38,450 field joints in Azerbaijan and 21,500 in Georgia.

  100.  By the time BP issued new instructions to the field joint coating applicators in these two countries, we are led to believe, there were 16,276 field joints coated with SPC 2888. The overwhelming majority (11,386) were said to be in Azerbaijan. Of those BP estimate 300 were problematic. Of the 4,890 joints in Georgia, BP estimate 1,260 were faulty, a staggering 26%. The reliability of these figures as a true assessment of the scale of the problem is dealt with in the next section. But taken at face value it is still a compelling indictment of the selection process.

  101.  BP failed to report any of this to the Lenders Group before the financial closure ceremony in front of the world's press on 3 February 2004. Incredibly, ECGD has supported BP's position and still maintains that none of the above represented a material adverse threat to the integrity of the BTC project.

THE REMEDIAL PLAN (HOW BIG A PROBLEM?)

  102.  The Lenders Group has allowed BP to assess the scale of a major engineering problems caused by its own lack of quality control and inspection. It has also allowed BP to oversee its own remedial work. Mr Weiss says in his covering letter that ECGD is currently "reassured" BP has identified the problems and has put in place "rigorous monitoring and corrective action systems." The WP report endorses this view, without having done a site visit.

  103.  I have evidence, contained in BP documents, suggesting there is a cover-up to downplay the scale and importance of the problems with the BTC pipeline, especially in Georgia. It appears vital checks on the pipeline are not being done for fear they will increase the repair programme and delay the project. Consequently, field joints that cannot be regarded as safe are being buried like environmental time bombs. The internal BP documents show the men who devised this secret plan are those who also approved the selection of the failed SPC 2888 coating.

  104.  Other internal documents also reveal a major split within BP. Senior Operations managers, who will shortly inherit the BTC pipeline from the BP Projects team, are questioning the integrity of the declared repair programme and have been slapped down.

  105.  According to the WP report, BP planned to "remove and recoat" all cracked field joints in Georgia, weather permitting, from July 2004 onwards. The WP report does not say this followed a limited number of inspections from late March to early June to establish the scope of the remedial work. Mr Osborne and his team say they inspected 1,000 field joints (between kilometres 147-175) in the Tabatskuri region of Georgia. Only coated field joints were inspected. Coated and buried field joints were not inspected. Nor were those sections of pipeline where the field joint had been coated but was lying in the trench waiting to be buried (backfilled[91]). BP will not say how many field joints are buried.

  106.  Based on this, Mr Osborne, the BTC pipeline corrosion control consultant, and Mr Fairhurst, BP's own senior materials and corrosion engineer, issued a report on 18 June 2004 outlining the entire repair strategy for all of Georgia. It is from this report that WorleyParsons and Foreign minister Mike O'Brien have based the figures given in point 100.

  107.  The committee should remember that the cracked and otherwise failing field joints are to be repaired and recoated using the same material, SPC 2888. BP refuses to accept there is a problem with the SPC coating, because to do so would leave the oil company, as BTC operator, open to major financial liability. Instead, the remedial strategy is designed to validate the unsustainable proposition that this is entirely an application problem.

  108.  Determining a remedial strategy does involve extrapolating from a sample inspection how many can be considered "fit for purpose" (safe) and how many should be repaired. Such a calculation depends on the inspectors not being influenced by any other factors except the long-term integrity of the pipeline.

  109.  However, in their 18 June report, Messrs Osborne and Fairhurst wrote the following alarming paragraph:

"Cross cut adhesion testing was limited, the decision to limit (it) was based on the acceptable outcome of previous testing and the desire to reduce repair frequency."

  110.  The St Andrews cross cut adhesion test is industry standard for determining whether the field joint coating is bonding to the plastic coated steel pipeline. It would strongly appear BP is trying to justify not carrying out full tests on the basis that previous adhesion testing was acceptable. This is unsupported by the evidence outlined above in this submission. Put simply, if the previous testing was professional why has the coating failed?

  111.  The more disturbing point though is the second reason BP gives for not doing full testing: "to reduce repair frequency." The irresistible conclusion one draws is this: BP is not testing unburied field joints because it doesn't want to identify more problems with the coating it selected and therefore delay the construction schedule with additional repairs. Also, as more data on the failed coating emerges, BP's liability increases. And the greater the evidence of why SPC 2888 failed, the less arguable it becomes to continue using it on BTC and on the parallel SCP gas pipeline, currently in the early stages of construction.[92]

  112.  It remains to be seen whether ECGD and the Lenders Group would endorse this secret remedial strategy, which is totally at odds with BP's own "Corrosion Control and Monitoring Philosophy"—a document also written by Mr Osborne. It states that to ensure the "high integrity" of the pipeline and field joint coating they must be "proven" and "compatible" with each other. Otherwise, the risk of "general and accelerated corrosion" occurs. A poor coating, the philosophy goes on, renders ineffective any cathodic protection, the second line of defence, which suppresses corrosion reaction on buried pipelines.

  113.  BP's secret corrosion control strategy helps explain why it told WP that cracks of less than 6 inches need not be repaired and can be buried.

  114.  The BP statement that "it is not necessary to repair all coating flaws, especially minor ones" has been roundly condemned by experts I've consulted, who feel it is part of the same attempt to "reduce repair frequency." Or put another way by Dr John Leeds, "out of sight, out of mind."

  115.  One of the main concerns of not treating every cracked joint is that it makes the pipeline more susceptible to the life-threatening process of stress corrosion cracking. For it to occur, a coating failure where the bare steel is exposed must have happened. Stress corrosion cracking will split open a buried pipeline like a tin can. The escaping high pressured, hot oil and gas can easily explode into a fireball.

  116.  When I visited various villages in Georgia and Azerbaijan there was concern about the pipeline being built too close to settlements. In Akhali Samgori, for example, village leaders told me their pleas to move the pipeline more than 500 metres from their homes was rejected. They worry about the gas and oil pipelines exploding either through technical failure or rebel attack.

  117.  I interviewed David Norman, a leading pipeline integrity consultant referred to in my earlier submission. He was contracted by PIH, the application firm, to assess the BP remedial strategy soon after the stoppage in November 2003. Mr Norman attended two private meetings in Tblisi, Georgia and later at Heathrow Airport between December 2003 and January 2004. In attendance were Mr Osborne, Mr Fairhurst, Mr Winter, SPC, the contractors and PIH. Mr Norman says he raised the issue of stress corrosion cracking and asked what mitigation measures were in place. He says BP told him it was not an issue. Mr Norman says the exchange was documented in PIH's minutes, which BP refused to agree.

  118.  BP has told WP that there has been no cracking since 9 February 2004, when they issued the corrective application specification to PIH. This does not accord with what my sources on the ground say. They tell me there was cracking in April, despite the new heating regime.

  119.  Returning to Mr Osborne's 18 June report. It reveals there are other serious problems with the pipeline's main plastic coating. The report says: "Many of the joints inspected were contaminated or showed signs of damage other than the cracking phenomenon." It then lists the following:

—  "A large proportion" of the pipes have the plastic jacket disbonding from the steel at the field joint interface.

—  Damage to the plastic coating on the overlap area where it is covered with SPC 2888.

—  Damage caused by the welding and field joint coating processes.

—  Damage caused where the plastic coating is resting on large rocks prior to burial.

—  "Moderate to severe rain damage" of the pipeline

  120.  According to David Norman, who has conducted site inspections for PIH, there are other problems that BP is not addressing. He says that a worrying number of 11.5 metre pipe coming from the factory does not have the 50 mm FBE toe. This is significant because BP claims the "toe" significantly aids the adhesion of SPC 2888 to the plastic coating.

  121.  Mr Norman also informs me that hundreds of field joints are being recoated without the FBE "toe". This, he says, is being blasted off when the faulty SPC 2888 coating is stripped from the faulty field joint.

  122.  A third problem which has apparently emerged is the quality of the chamfer—the point where the plastic coating is cut back on the pipeline to reveal the FBE toe and the bare steel. This is not smooth enough and has therefore created hundreds of stress points that will cause the SPC 2888 coating to crack in the ground when the pipeline is operational and therefore moving. This interface is the weakest spot in the field joint coating system and if it cracks will allow water to get underneath the SPC 2888 coating and between the plastic and the steel pipeline.

  123.  A number of issues emerge: Did WorleyParsons know about these additional problems with the BTC pipeline? And if so, why wasn't it in their report? Or did BP fail to mention this as well?

  124.  So far I've concentrated on the remedial strategy for the coated but unburied pipeline. Mr Weiss and the WP report say BP have engaged "an independent company" to do a technical survey (or DCVG test) to indicate flaws in the plastic and field joint coating of the buried sections of the pipeline. BP, writes Mr Weiss, has told ECGD that "any sections where material flaws are identified will be excavated and the problems corrected."

  125.  Ionic Consulting, part of JP Kenny, is the company carrying out the DCVG survey. Both companies already have a major contract with BP on the Shetland Islands. Ionic confirmed to me that Mr Osborne contracted them to do the DCVG survey. The Institute of Materials directory entry for Ionic lists Mr Osborne as one of their "key personnel". Ionic also confirmed that Mr Osborne is a consultant of theirs and is in pole position when they land a new contract.

  126.  Furthermore, Ionic confirms that the man who carried out the DCVG survey for Mr Osborne has recently left to join WorleyParsons. The DCVG survey is therefore incomplete and about 25% of the work, mainly in Azerbaijan, remains to be done, Ionic say.

  127.  The preliminary Ionic report dated 7 May 2004 suggests there are 23 "anomalies" that "could be classified as small where repair is not recommended as they are of low importance." Dr John Leeds, a world expert on DCVG surveys, will I understand be critiquing the Ionic preliminary survey. He is not alone in raising serious concerns about its adequacy. This after all is the mechanism through which BP will determine what field joints have to be dug up, stripped and recoated.

  128.  Documents I've seen show the BP Commissioning and Operations team in Baku have already expressed reservations to BP Projects about the attempt to play down these anomalies (holes/cracks in the coating). The documents also show BP Operations are worried about the continued use of SPC 2888 on the oil and gas pipelines.

  129.  BP Operations feel BP Projects did the DCVG survey to prove everything is fine, and not to appraise the true scale of the coating failing. They point out that the survey doesn't deal with the integrity of the buried field joints—the very point it was commissioned to address.

  130.  BP Operations believe all small defects in the coating are potentially significant. They therefore want to independently review the field joint coating issue but were over-ruled by BP Projects and its sponsors in London. BP compliance manager David Winter is one of the leading supporters of the view the DCVG survey by Ionic shows there is no cause for concern.

  131.  BP Operations are still unconvinced. On routine pipeline patrols they discovered cracked field joints with "rust seeping through the coating indicating disbanding." They are not convinced that all joints coated before the stoppage have been properly surveyed and a proper remedial strategy is in place. They are also concerned about welding, and the effectiveness of cathodic protection on pipeline that has been in the ground for over one year. They are also very concerned for the corrosion damage caused where unburied pipe had been allowed to gather water.

  132.  The documents reveal a major schism within BP, which the WP report has done nothing to reflect. It seems the field joint coating issue has come a full circle since mid-2002. Then, BP Baku commissioned Derek Mortimore to review the field joint specification coming from the BP design/projects team in London. BP managers like Paul Stretch were over-ruled and Mr Mortimore marginalized. Now a new group of concerned BP managers are also trying to review the work of the same group and have also been over-ruled.

THE FAILURE OF QUALITY CONTROL AND ASSURANCE

  133.  The WP report tells us BP put "full time coating inspectors" in Azerbaijan and Georgia from January 2004. Previously they were part time and had other responsibilities. BP also plans to put "senior pipeline inspectors and BP/BTC corrosion specialists" in the field. This is an implicit admission that BP's Quality Assurance Programme failed during the construction phase. The WP report goes on to say this QAP failure made the cracking problem worse, in that the coating process should have been halted earlier.

  134.  ECGD also shares in this failure. In its 17 December 2003 letter of decision to underwrite the BTC project, ECGD stated that its "independent engineers" (WorleyParsons) had scrutinised BP's "quality control system." Clearly this is not true, because if they had then they would have noticed it wasn't functioning.

  135.  BP already has a very bad QAP record in Alaska, where it operates the TAPS pipeline. Currently it is under investigation by the federal investigator, the Environmental Protection Agency. The EPA is investigating claims from over 12 whistleblowers who had worked as inspectors in BP's corrosion division in Alaska. They have variously alleged the QAP system is undermined by "falsification of inspection data and improper procedures" and cost saving initiatives.

  136.  There is also a well-documented history of harassment and intimidation of BP workers who raise concerns internally.[93] It is within this "shoot the messenger" culture that I would suggest the unseemly attack on Mr Mortimore by BP and WorleyParsons should be seen. It is equally disturbing that WorleyParsons didn't seek to speak to Mr Mortimore or interview the concerned BP Baku senior managers who contracted him. This left the wholly misleading impression that he is a bitter, Jurassic consultant detached from innovative thinking and unsupported by anyone inside or outside the oil company.

  137.  There is no effective, independent, well-funded regulator in Azerbaijan or Georgia. Consequently, BP is doing an investigation into itself, with no effective oversight.

  138.  In the UK, if the field joint coating issue and associated problems had been discovered on a pipeline under construction that intended to run near highly populated areas and major mineral water sources like the Malvern Hills would BP be able to get away with this? The HSE would in all probability have stopped the job, issued a 28-day notice and overseen the remedial work it had a hand in drafting.

  139.  The thrust of the WP report is to attribute blame on Mr Mortimore, the befuddled whistleblower, and on Pipeline Induction Heating, the British company BP now publicly blames for misapplying SPC 2888 coating to the field joints.

  140.  WorleyParsons never interviewed PIH. Had it done so, PIH would have told them Mr Osborne's specification was "incomplete" and SPC 2888 was "the wrong product". Furthermore, sources close to the company have told me that PIH has not incurred any financial penalty for its alleged misapplication of SPC 2888. Indeed, if PIH had cocked up as badly as BP and the ECGD have publicly suggested, one would not expect it to be paid to carry out the remedial work. It is understood PIH are also in line for the contract to coat the Southern Caucasus (gas) Pipeline.

  141.  There are complex negotiations going on at the moment over how to quietly sort this problem out without ending in the civil courts, where a lot of dirty washing would be aired. The negotiations are between Mr Osborne and BP's compliance department on one side and the contractors (Spie Capag and CCC) and PIH on the other.

  142.  Who had primary and secondary responsibility to ensure the field joint coating was fit for purpose, is a key questions. The issue of criminal negligence—whether warnings were ignored—is another and also whether procurement fraud played a part in the choice of the failed coating.

  143.  Experts consulted say primary responsibility lies with the client, BP, as it had overall responsibility through its compliance department to ensure and enforce the Quality Assurance Programme. The QAP is a document developed in agreement with BP's contractors. The single sourcing of SPC 2888, also puts BP in a very difficult situation.

  144.  Things get trickier because BP's compliance manager David Winter and in house corrosion engineer, David Fairhurst sided with Mr Osborne and SPC over Mr Mortimore, and others in BP.

  145.  PIH does have a responsibility to point out failings or problems in its instructions from the client before or during the work, in this respect they may have failed.

  146.  Another bizarre anomaly is the situation in Turkey. This is the longest section of the BTC pipeline with 90,000 field joints. Botas, the Turkish state oil and gas company, is to carry out the construction but Bechtel/BTC has an overall design assurance role. You could be forgiven for thinking that SPC 2888 is also being used on the Turkish section. It isn't. They are using Protegal 3210. If the name sounds familiar it's because Protegal was one of the three liquid coatings Advantica tested for BP in July 2002. In that test Protegal came last and SPC 2888 came first. At the time of writing I am still waiting for answers from Botas about whether they were made aware of the Advantica test. And from BP/Bechtel about why they allowed Turkey to use a field joint coating it disqualified for poor performance in Azerbaijan and Georgia. If Botas weren't told this raises major questions about BP's quality assurance. If they were told and dismissed the Advantica test, it raises questions about the reliability of the test that put SPC 2888 first.

  147.  At the very least, the BP Projects team is hopelessly conflicted as the appropriate body to design and oversee the remedial strategy. The BP board in London has allowed this situation to develop and endorsed the aggressive PR strategy of toughing it out in front of the select committee. The government, through its ECGD department, has adopted this line as evidenced by ministerial statements, responses to parliamentary questions, Mr Weiss' appearance in front of the committee on 11 May and more recently the preposterous WP report. Like BP, the ECGD refuses to accept its own mistakes and instead commissions study after study to try and make the original selection of SPC 2888 look responsible. As the field joint scandal has escalated with the announcement of the select committee inquiry BP has indulged in a further cover up. The evidence lies in the extraordinary BP document dated 18 June 2004.

  148.  An independent investigation into the commercial and scientific aspects of the selection of SPC 2888 is, I submit, the only safe way forward for the people and environment of Georgia, Azerbaijan and Turkey, and the British taxpayer.

1 October 2004

---

---

79   I enclose copies of two of these front-page articles. Should the committee want to copy tapes of the documentaries please contact me. Back

80   Not printed. Back

81   The SPC UK sales agent at all material times is a man called Mike Bird. He has been employed by SPC since at least March 2001, although E.Wood recall Bird and SPC's Canadian chairman, Bob Alliston, coming to their offices in Yorkshire in 1999. Back

82   We had also heard claims that a batch of SPC 2888 paint had to be placed in quarantine because it was in effect watered down or had separated. So far these allegations remain unconfirmed but the committee could ask BP to provide details of this. The ECGD, we understand, was not informed. Back

83   At first blush this may seem far-fetched. But anyone familiar with BP's activities in Alaska will know how its hand in a spying operation on employees, external critics and a US congressman was exposed in the eighties and nineties. Back

84   The Turkish section of the pipeline is not using SPC 2888 and is being constructed to a different spec by another contractor. Back

85   ECGD statement to The Sunday Times dated 14 February 2004. Back

86   In August 2004 the Georgian government suspended work in the Borjomi region, where the national mineral water company is based, following reports from independent consultants that the BTC pipeline needed more environmental and security measures. BP had not told the Georgian government about the earlier field joint coating stoppage in November 2003. Back

87   BP also refused to explain why the report was redacted, what role SPC played, the names of its investigation team dealing with the failed field joints and the number of field joints that were buried in Azerbaijan and Georgia by February 2004. Back

88   Mr Powers now claims I misquoted him in my earlier submission. I re-checked those bullet points against the tape-recording of our conversation on 25 February. They are accurate, as are these. I shall release the tapes should the Committee require it. Back

89   In my earlier submission I pointed out the Advantica test on SPC 2888 was carried out using a wash primer. But in the specification Mr Osborne wrote soon after this test, nominating SPC 2888 as the sole source field joint coating material, he excluded a wash primer. This leaves a number of unresolved questions: Was the wash primer one of the reasons why in certain tests SPC 2888 appeared to outperform its two other liquid coating competitors? And does its absence from the field joint coating specification in part explain why SPC failed when applied to the BTC pipeline? The WP report does not address this properly. And neither BP nor Mr Osborne has provided an explanation. They simply ignored the question when the Sunday Times put it to them. Mr Mortimore had drawn the wash primer question to BP's attention in November 2002. In January 2003 BP commissioned Advantica to do a test, which is referred to in the latest WP report. Advantica determined the use of a wash primer was "detrimental" unless used with flame treatment. Mr Osborne is now using this to explain why the wash primer was not included in his field joint coating specification dated October 2002 and therefore not used on the BTC pipeline. But even this has been contradicted by SPC during an industry conference in Germany. Jim Banach, a leading SPC representative, gave a paper in October 2003 that confirmed a wash primer and flame treatment were best practice. What are we to believe? This is further evidence the selection process was at best incompetent. Back

90   Ian Thompson of Advantica refused to return repeated calls to his office. Back

91   After the welded field joint has been coated it's lowered into a ditch and backfilled with small, sharp rocks dropped at a height in bulk on the pipeline. Naturally the plastic and field joint coating must therefore be resistant to penetration, abrasion and impact. The WP report shows BP only did a backfill test on how the SPC coating behaved when rocks were dropped on it after SPC had been selected. Furthermore, BP has taken what experts consider to be an extraordinary decision not to bed and pad the pipeline with a fine material like sand. Is this a cost cutting measure, perhaps in line with its Pipeline Cost Reduction Initiative? Dr John Leeds wonders what prior testing of SPC 2888 was done to see how it behaves on a full and operational buried pipeline that is repeatedly rubbing against large diameter rocks instead of sand. Back

92   BP is the operator and joint main shareholder of the SCP gas pipeline. The WP report confirms SPC 2888 will be used to coat its field joints. Back

93   See Financial Times `Probe into BP Alaska Widens' by Sheila McNulty, 8 August 2004 and The Guardian, 12 July 1999 and New Solutions magazine 2000 by M Gillard, M Jones and A Rowell. Back

---

Prepared 4 May 2005