UK block 211/19 and 211/25 awarded to Conoco

National Library of Norway

person By Petter Rønningsen, National Library of Norway
The National Library is one of the most important sources of knowledge about Norway, its citizens and Norwegian conditions. It is responsible for collecting, preserving and making available collections of all types of media.

Pursuant to the Legal Deposit Act, all material published in Norway – regardless of media format – must be deposited with the National Library. These testaments to Norwegian culture and social life are thereby preserved for posterity, and represent an important source for documentation and research. Activities are developed in collaboration with various social institutions in the fields of libraries, media, training, research, archives, museums and art. New services are also created for the public.The National Library represents an important resource in a number of areas, such as the infrastructure for Norwegian research – including its role as a research library. It also serves as an instrument of cultural policy, of long-term conservation of Norwegian cultural heritage, and for celebrating important anniversaries of authors. In addition, the library has gained an extended responsibility in linguistic policy through the job of establishing, developing and operating a Norwegian language bank. It has also been made responsible for developing Norway’s library sector.
Digitising, long-term conserving of digital records and developing digital library services represent a key part of the National Library’s activities.
It has begun the extensive job of digitising its complete collection, a project expected to take 20-30 years. Material is made available in accordance with the Copyright Act or agreements with copyright holders.
With almost 450 employees, the National Library is headed by the national librarian. Its organisation is split between Oslo and Mo i Rana in northern Norway.

Role in the Statfjord industrial heritage project

The National Library is equipped to make the part of its collection relevant to Statfjord available through the industrial heritage project. That derives both from its role as the depository for all material published in the public arena, and from its current project to digitise its whole collection. The library has been responsible for digitising the books, journals, radio programmes and pamphlets which can be found on this website.
It has also received digital files and metadata for films from Norgesfilm and for photographs and objects from the Digital Museum , which have been made available through the search page. In addition, the library has implemented and operates the entire web solution for the Statfjord industrial heritage, including the search function.
The latter makes it possible to search both in metadata (name, title, author and so forth) and in the actual content of a book, for example, or other printed works. A relevant digital object can be viewed or listened to.
Other digital material can also be found in the National Library’s digital collection (in Norwegian only) at www.nb.no .

UK block 211/19 and 211/25 awarded to Conoco
Published January 2, 2019   •   Updated January 3, 2020
© Norsk Oljemuseum
close Close

Building the Statfjord A topside

person by Trude Meland, Norwegian Petroleum Museum
The Aker group had been commissioned by Mobil in August 1974 to fabricate and outfit the topside for Statfjord A. It had already undertaken similar contracts – Vindholmen Verft in Arendal had built the topside for Beryl A and Aker Verdal north of Trondheim was building a similar structure for Shell/Esso’s Brent B. Both these platforms were destined for the UK sector of the North Sea.
— The Statfjord A topside under construction in the dry dock of Aker Stord. Two old hulls, named Tom and Tina, were used as "foundations" for this work. Photo: Tor Resser/Norwegian Petroleum Museum
© Norsk Oljemuseum

Under the letter of intent signed between Mobil and Aker, the module support frame (MSF) for Statfjord A would be fabricated at Aker Verdal with outfitting by Aker Offshore Contractors (AOC). The latter also had the contract for mechanical outfitting in the shafts of the concrete gravity base structure (GBS).

When it turned out that the topside would be larger and heavier than first agreed, the Aker Verdal yard ran into capacity problems.[REMOVE]Fotnote: Haga, T. (1993). “Stordabuen går offshore” : Arbeid og faglig politikk ved A/S Stord verft 1970-83 (Vol. 1993-4, AHS (trykt utg.)). Bergen: Gruppe for flerfaglig arbeidslivsforskning, Universitetet i Bergen: 260. Stord Verft south of Bergen, which had only built ships – primarily oil tankers – before, was hit when the tanker market collapsed in the wake of the oil crisis. An order for 11 supertankers was cancelled, leaving the yard short of work.[REMOVE]Fotnote: Grove, K., Heiret, J., & Stord jern- & metallarbeiderforening. (1996). I stål og olje : Historia om jern- og metallarbeidarane på Stord. Stord: Stord metall- og bygningsarbeider[e]s fagforening: 161. og Myklebust, A., & Aker Stord A/S. (1994). 75 år på Kjøtteinen : 1919-1994 : Jubileumsbok for Aker Stord. Stord: [Aker Stord].

Bygging av A-dekket,
Statfjord A topside under construction at Stord Verft. Photo: Aker Mekaniske Verksted/Norwegian Petroleum Museum

Aker accordingly submitted a proposal to Mobil in February 1975 to transfer the job of building the Statfjord A topside from Verdal to Stord. Mobil accepted this, and Stord Verft thereby acquired its first offshore assignment. The contract was signed on 5 April 1975.[REMOVE]Fotnote: Moe, J. (1980). Kostnadsanalysen norsk kontinentalsokkel : Rapport fra styringsgruppen oppnevnt ved kongelig resolusjon av 16. mars 1979 : Rapporten avgitt til Olje- og energidepartementet 29. april 1980 : 2 : Utbyggingsprosjektene på norsk sokkel (Vol. 2). Oslo: [Olje- og energidepartementet]: 178.

But Stord Verft was soon to discover that fabricating offshore structures differed significantly from shipbuilding. While the latter was a form of mass production, a platform had to be custom-built.

Bigger than expected

The original contract allowed Stord Verft to take on a wider scope of work. As initially conceived, the topside was to comprise a single MSF where the equipment was installed in modular form – almost like a large Lego construction.

Bygging av A-dekket,
The Statfjord A topside is starting to take shape. Foto: Aker Mekaniske Verksted/Norwegian Petroleum Museum

Since the yard had spare capacity, however, Aker initiated negotiations with Mobil on a new integrated structure. Rather than all the modules forming a conventional modular system, the support structures for a number of these would be incorporated in the MSF, and their equipment then installed in them.

The contract with Aker’s Stord Verft embraced the following elements.

Design and fabrication of the steel topside, with an estimated weight of 6 800 tonnes. This included steel procurement, towing the topside to the mating site and mating with the Condeep GBS.

Design and fabrication of the MSF for the integrated M6, M8, M11, M12, M14, M17, M18 and UM7 modules and installation of the deck, with an estimated weight of 1 828 tonnes. This also included steel procurement.

Outfitting the cellar deck and the above-mentioned modules. This would be done to the extent that approved drawings, specifications, equipment and materials were available at Stord Verft.

While contract negotiations were under way between Aker and Mobil, appraisal wells being drilled on Statfjord revealed that the field was larger than originally thought. On the basis of that knowledge, it was resolved to double production capacity on the platform from 150 000 barrels per day to 300 000. In order to be able to handle such large volumes of oil, the topside facilities had to be expanded from one to two process trains. That naturally doubled the size and weight of the equipment. This in turn meant that the MSF had to be strengthened to handle the weight.[REMOVE]Fotnote: Moe, J. (1980). Kostnadsanalysen norsk kontinentalsokkel : Rapport fra styringsgruppen oppnevnt ved kongelig resolusjon av 16. mars 1979 : Rapporten avgitt til Olje- og energidepartementet 29. april 1980 : 2 : Utbyggingsprosjektene på norsk sokkel (Vol. 2). Oslo: [Olje- og energidepartementet]: 141.

A partially integrated topside would be significantly more complicated, but Aker justified the design changes submitted to Mobil on 6 February 1975 by noting that steel weight would be reduced. Mobil accepted the revised design and a new contract was signed on 5 April 1975.[REMOVE]Fotnote: Moe, J. (1980). Kostnadsanalysen norsk kontinentalsokkel : Rapport fra styringsgruppen oppnevnt ved kongelig resolusjon av 16. mars 1979 : Rapporten avgitt til Olje- og energidepartementet 29. april 1980 : 2 : Utbyggingsprosjektene på norsk sokkel (Vol. 2). Oslo: [Olje- og energidepartementet]: 178.

This meant a welcome increase in the scope of work for Stord Verft. A number of the modules were originally due to be fabricated at other yards and engineering works in Norway and abroad which were also struggling with the after-effects of the oil crisis. Because Stord Verft was now going to integrate these modules in the MSF, the other fabricators lost sorely needed work.[REMOVE]Fotnote: Moe, J. (1980). Kostnadsanalysen norsk kontinentalsokkel : Rapport fra styringsgruppen oppnevnt ved kongelig resolusjon av 16. mars 1979 : Rapporten avgitt til Olje- og energidepartementet 29. april 1980 : 2 : Utbyggingsprosjektene på norsk sokkel (Vol. 2). Oslo: [Olje- og energidepartementet]: 183. Modules were built by Kværner/Sterkoder in Egersund and Kristiansund (M1A, M1B, M2, M10, M13, M16, UM8, UM9 and the toolroom), Nordland Offshore A/S in Sandnessjøen (UM9B and the telecommunications module), Bodø Mekaniske Verksted (UM2 and M10 top), Wilson Walton, Middlesbrough (M3, M5 and M9), and Curtis Wright/RSV in Rotterdam (M4A, M4B, UM5 and UM6). In addition, Norway’s Leirvik Sveis completed M9. 

In retrospect, it is clear that the local management at Stord Verft took an unrealistic view of the transition from shipbuilding to platform construction. A topside for what then ranked as the world’s largest offshore installation was far more complicated, and new standards had to be met – particularly for welding.

The amount of engineering design work was greatly underestimated, and detailed design proved the first casualty. While the yard had been accustomed to using its own drawing office, it now received detailed drawings from Matthew Hall Engineering (MHE) in London. The latter had the contract for the technical design of outfitting and process equipment.

Design job too extensive

MHE had won an engineering management contract (EMC) for the Statfjord A topside (see the section on the construction contracts for more details). This job embraced conceptual and technical design, management of process equipment and modules – including procurement – award of fabrication contracts and construction site supervision.

Although the EMC had been put out to competitive tender, a tight market meant that Mobil received only two bids. The operator had been concerned from the start about MHE’s lack of experience with similar projects, but had no choice but to award it the contract if the schedule was to be maintained.[REMOVE]Fotnote: Moe, J. (1980). Kostnadsanalysen norsk kontinentalsokkel : Rapport fra styringsgruppen oppnevnt ved kongelig resolusjon av 16. mars 1979 : Rapporten avgitt til Olje- og energidepartementet 29. april 1980 : 2 : Utbyggingsprosjektene på norsk sokkel (Vol. 2). Oslo: [Olje- og energidepartementet]: 168.

It quickly transpired that the design work failed to meet the required standard and would be delayed. MHE had estimated 400 000 hours for the work, with 90 per cent completed by the end of 1975. But the company ran into problems.

In addition to insufficient experience with such projects, it lacked enough personnel and had up to 80 per cent contract workers at one point. When the decision was taken to change the concept to a partly integrated topside, MHE had already been working for nine months on planning the process equipment. A completely new concept meant that the company had to start all over again, and many working hours were wasted.[REMOVE]Fotnote: Moe, J. (1980). Kostnadsanalysen norsk kontinentalsokkel : Rapport fra styringsgruppen oppnevnt ved kongelig resolusjon av 16. mars 1979 : Rapporten avgitt til Olje- og energidepartementet 29. april 1980 : 2 : Utbyggingsprosjektene på norsk sokkel (Vol. 2). Oslo: [Olje- og energidepartementet]: 169.

Only 50-60 per cent of the design work was completed in January 1976, and the number of hours was four times higher than the estimate. Mobil resolved to cancel the contract with MHE and brought in a US engineering contractor – Brown & Root – to replace the London-based company on a gradual basis. Mobil regarded Brown & Root as the best qualified candidate for the job, and awarded the contract without competitive tendering.

In October 1975, Brown & Root also took over the contract for planning, management and execution of offshore hook-up work from MHE. It already had the job of technical design and management for work on the loading buoy, flare boom and flowlines.[REMOVE]Fotnote: Moe, J. (1980). Kostnadsanalysen norsk kontinentalsokkel : Rapport fra styringsgruppen oppnevnt ved kongelig resolusjon av 16. mars 1979 : Rapporten avgitt til Olje- og energidepartementet 29. april 1980 : 2 : Utbyggingsprosjektene på norsk sokkel (Vol. 2). Oslo: [Olje- og energidepartementet]: 156. This meant that the company now held all three of the major assignments for completing Statfjord A.

Topside weight increases

The factor which caused the biggest problems in planning and building Statfjord A was a massive underestimating of the weights involved.

Bygging av A-dekket,
Stord Shipyard estimated the hours of extra work as result of the weight problems to 150,000 hours, which corresponded to about NOK 75 million including materials. Another consequence of the reinforcement work was failure in the supply of materials. The amount of steel ordered was based on the initial estimates and was therefore too small. With a tight market, delivery time was long. Photo: Aker/Norwegian Petroleum Museum

A platform’s weight can be measured in two ways – wet and dry. Wet weight is measured in operating condition with the various process components are filled with liquids. The GBS and topside are designed to cope with a given weight, and the wet weight must not exceed that level. The dry weight is the sum of MSF and equipment alone. A platform is towed out fully or partly outfitted, and the dry weight of the topside must not exceed the tow-out value.[REMOVE]Fotnote: Rolstadås, A., & Norges tekniske høgskole Institutt for verkstedteknikk. (1981). Prosjektstyring. Trondheim: Tapir: 137.

MHE lacked a satisfactory system for either wet or dry weight control. The most accurate possible weight estimates for both MSF and production equipment were already required in early 1975 in order to design the topside. But the wet weight calculations produced by MHE were too low.

Disputes arose between the design company on the one hand and Stord Verft/Aker on the other about the loadbearing capacity of the topside. MHE’s estimate was 34 500 tonnes, while the contract with Stord Verft stipulated 41 500 tonnes. The overall weight ultimately came to 49 500 tonnes. Extensive design work was required to ensure that the MSF could bear this weight. That in turn led to big delays and significantly more expensive fabrication work.[REMOVE]Fotnote: Moe, J. (1980). Kostnadsanalysen norsk kontinentalsokkel : Rapport fra styringsgruppen oppnevnt ved kongelig resolusjon av 16. mars 1979 : Rapporten avgitt til Olje- og energidepartementet 29. april 1980 : 2 : Utbyggingsprosjektene på norsk sokkel (Vol. 2). Oslo: [Olje- og energidepartementet]: 164.

During the fabrication period, MHE’s engineers fell behind with the design work. In an attempt to speed up progress with the technical design, MHE issued incomplete drawings. This meant that the number of drawing revisions and fabrication changes increased. When drawings were amended, structures already installed had to be ripped out and the work done again.

Tow-out weight restriction

The tow-out or dry weight also caused problems. The loadbearing capacity of the Condeep imposed restrictions on the amount of equipment which could be positioned topside before departure for the field.

It turned out that the maximum tow-out weight was significantly lower than the hardware to be installed. Modules accordingly had to be removed for installation once the platform had been positioned on Statfjord. Such assembly and disassembly, both for this reason and because of revised drawings, had a negative effect on the motivation of the workforce at the Stord yard.

Welding is not just making welds

Welders were the largest group of workers involved in fabricating the Statfjord topside. They quickly discovered that welding tankers and oil installations were two different exercises. The former demanded other material types and thicknesses than were used on ships, and the new kind of steel demanded a different welding technique. In addition, the quality of the work had to be improved and its accuracy improved.

Mobil required that all the welders were certified to specified standards. Instead of the earlier approval by Norway’s Det Norske Veritas (DNV), the American ASME norm was introduced. This was divided into six levels, from 1G as the lowest up to 6G. To secure a 6G certificate, a welder had to be able to weld around a pipe positioned at an angle of 45 degrees while welding “through” – in other words, so that the joint is filled.

Many former ship welders failed to pass the tests for the new certificates, and the yard eventually had problems securing sufficient skilled personnel. Since the work was already behind schedule, the management wanted to bring in foreign welders. But this had first to be cleared with the unions. Welders were brought in from other Aker companies to help out. The shortage of welders meant that prefabrication lagged behind, and delays there had a knock-on effect on assembling the topside.

Welds were carefully X-rayed and checked. The yard was told that each weld had to be marked with a number, so that possible faults could be traced directly back to the welder who had done the job. As the work progressed, it became clear that a number of welds had to be redone. If the checks showed more than five per cent error in a weld, it was rejected. A welder who had too many rejects had to be recertified.[REMOVE]Fotnote: Grove, K., Heiret, J., & Stord jern- & metallarbeiderforening. (1996). I stål og olje : Historia om jern- og metallarbeidarane på Stord. Stord: Stord metall- og bygningsarbeider[e]s fagforening: 170. By March 1976, fabrication was far behind schedule.

Pipework

The pipefitters also found that their work changed with the transition from ships to platforms. There was considerably more pipework and it was more monotonous. The piping was prefabricated and much of the work previously done by the fitters arrived ready-made. Piping was lifted on board from the warehouse by crane.

When ships were still being built, detailed drawings were supplied by the yard’s own engineers, but the individual foreman or fitter could change them if he found it appropriate. That was no longer the case.[REMOVE]Fotnote: Grove, K., Heiret, J., & Stord jern- & metallarbeiderforening. (1996). I stål og olje : Historia om jern- og metallarbeidarane på Stord. Stord: Stord metall- og bygningsarbeider[e]s fagforening: 169. Drawings supplied from London had to be followed to the smallest detail, and all changes had to be reported. The independence and scope of the work was reduced.[REMOVE]Fotnote: Myklebust, A., & Aker Stord A/S. (1994). 75 år på Kjøtteinen : 1919-1994 : Jubileumsbok for Aker Stord. Stord: [Aker Stord].

Working conditions in the workshops were also poor. The worst problem was the heat. The platform steel had to be preheated to 150°C before it could be welded. Although fresh air was blowing into the welding areas to thin out the gases, no effective ventilation was in place. There were smoke and fumes everywhere, while greater use of angle grinders and carbon rod welding increased the noise level.

The Stord Verft workforce noted the transition to the offshore industry directly. Their work became less independent, delivery times were shorter and physical working conditions deteriorated. The many change jobs had an unfavourable impact of their morale, and productivity declined.[REMOVE]Fotnote: Grove, K., Heiret, J., & Stord jern- & metallarbeiderforening. (1996). I stål og olje : Historia om jern- og metallarbeidarane på Stord. Stord: Stord metall- og bygningsarbeider[e]s fagforening:172.

To offset this decline in efficiency, the workforce was expanded to 950 people between June and December 1976. But this did not increase the amount of work actually done.

Designing equipment which was to last “for ever”, and which would never again be seen or accessed, was unusual and challenging. No design codes existed for offshore petroleum projects in 1975. When mechanical engineer Einar Jensen at AOC worked on the rotating machinery, he and his team used design codes developed by Mobil for use on land and DNV codes for shipboard use.

Nobody had ever designed such a large and complex structure as Statfjord A, with all its criss-crossing pipes.[REMOVE]Fotnote: Interview with Eilef Lynghaug, Einar Jensen and Jan-Henry Larsen, former platform managers on Statfjord, by Trude Meland, Norwegian Petroleum Museum, 18 September 2008.  To begin with, all the piping was marked. But these marks were subsequently covered with insulation, so the work was wasted.

Postponement

Bygging av A-dekket,
The Statfjord A topside at Stord Verft. Photo: Aker Mekaniske Verksted/Norwegian Petroleum Museum

Mobil eventually grasped the seriousness of the position, and realised that tow-out of the platform would have to be postponed for a year. It could not go out to the field before the summer of 1977.

In statements to the media, the company emphasised that more of the work on the installation could thereby be done inshore rather than more expensively and lengthily out on the field:

“The more equipment which can be installed on the platform while it is inshore ¬– according to plans, at Stord, the better. In this way, the time at land can be exploited and difficult and expensive work avoided on the field in the North Sea.”[REMOVE]Fotnote: Stavanger Aftenblad, 13 May 1976.  “Størst mulig last en fordel”. 

Because of the maximum tow-out weight, however, this plan could not be fully implemented.

Despite the problems, the MSF was eventually completed. This work had taken place in the yard’s dry dock, but the gates were opened on 30 July 1976 and the structure was towed out into the fjord. At the same time, the GBS was on its way from Stavanger to Stord.

The MSF had been built on two redundant tankers – known as Tom and Tina. When the topside was to be mated with the GBS, the latter was ballasted down until it was almost entirely submerged and the two tankers were manoeuvred between the concrete shafts.

Deballasting the storage cells then raised the GBS, so that the topside was carefully positioned on and then lifted by the shafts.[REMOVE]Fotnote: Godø, H. (1980). Plattformutbygging til havs (Vol. Nr 11-1980, Sosialdepartementets sammendragsserie (rapportsammendrag : trykt utg.)). Oslo: Sosialdepartementet: 45. This mating operation took place on 8 August 1976, and work then continued out in Digernessundet.

Leaks

But the problems were not over. Cracks in the GBS were discovered in 1976 during maximum submersion in connection with the mating operation. They had arisen as a result of excessive temperature fluctuations and a weak section in the concrete. It took two months to repair the damage.[REMOVE]Fotnote: Hanisch, T., Nerheim, G., & Norsk petroleumsforening. (1992). Fra vantro til overmot? (Vol. 1). Oslo: Leseselskapet: 392. Einar Jensen experienced this episode:

“It was a Friday evening, I think, and we were on our way home to Stavanger. When the high-speed ferry berthed, the project manager was standing there and waving us to one side. We went directly to Sola [airport] and into a specially booked plane. Statfjord A was listing. The platform was shipping water in its cells. Four-five of us were involved. There was little drama, but we took in a good deal of water and didn’t know where it was coming from.”[REMOVE]Fotnote: Interview med Einar Jensen by Trude Meland, Norwegian Petroleum Museum, 2 October 2008.

It transpired that there were two cracks, each 20 centimetres long, in the concrete between one “star” cell and a storage cell. The GBS was taking in 15 000 litres of water per hour or 20 buckets a minute.[REMOVE]Fotnote: Stavanger Aftenblad, 14 September 1976. “20 bøtter i minuttet”. But the structure was never in any danger of tipping over or sinking. The pumps installed in the GBS were capable of discharging a million litres per hour, and had no problem controlling the water intrusion. Without the pumps, the GBS would have sunk 60-70 centimetres per day because of the extra displacement caused by the water.

People from AOC and Norwegian Contractors (NC) devoted the weekend to identifying what had happened. When the cracks were located, NC – the company responsible for the concrete structure – injected them with epoxy, a special and elastic form of concrete. Sealing the cracks was a big job, and the incident led to delays. Installing such modules as the living quarters and the helideck had to be postponed until the damage was repaired.[REMOVE]Fotnote: Hanisch, T., Nerheim, G., & Norsk petroleumsforening. (1992). Fra vantro til overmot? (Vol. 1). Oslo: Leseselskapet: 392.

Two months later, Statfjord A listed again while it was lying in Digernessundet. A test of the emergency shutdown procedure for the ballast system went wrong. The living quarters had been installed at one end of the topside, and differing levels of ballast water in the various cells was used to compensate for the added weight. A service operative from the Swedish factory which had delivered the ballast system was going through the procedure for testing the hydraulics. Einar Jensen reports what happened:

“The procedure involved placing each valve in a semi-open position, and then pressing a red button and timing how long it took to shut down. There were 16 cells with ballast water. He tested first one valve and then another. Everything seemed to be working fine. The final stage in the procedure was to place all the valves in a semi-open position and then press the big button. Since we had varying water levels, things began to happen and the platform started to list. The alarms sounded, and the big button fortunately worked. We had a well-qualified man in the control room who pressed the button as soon as he saw the platform begin to list.”[REMOVE]Fotnote: Interview med Einar Jensen by Trude Meland, Norwegian Petroleum Museum, 2 October 2008.

The procedure followed would have been correct for a system which was not in operation. Since the system on Statfjord A was in fact operational, the final stage should have been left out.

This was a serious incident. The platform listed by three degrees, which means that the topside sank by eight-nine metres along one side.[REMOVE]Fotnote: Gjerde, K., Ryggvik, H., & Gooderham, R. (2014). On the edge, under water : Offshore diving in Norway. Stavanger: Wigestrand: 158. It was evacuated, and nobody was hurt. Subsequent reports have alleged that people jumped from the topside, but Jensen denies this:

“People didn’t jump from the topside, but a scaffolding ladder was installed on each side down to a barge on the water. When the platform listed, the stair on one side became far too short, so that when you reached the bottom step the surface was still some way down. It’s possible that some people who went down those stairs jumped, but I can’t confirm it.”[REMOVE]Fotnote: Interview med Einar Jensen by Trude Meland, Norwegian Petroleum Museum, 2 October 2008.

When Mobil decided that the tow-out of Statfjord A should be postponed by a year, a new date was also set – 3 May 1977. That deadline was met, and the platform was in position on the field by 8 May.

Grouting – filling the space between the GBS base and the seabed with gravel – started two days later. The remaining work of completing Statfjord A and readying it for production could then begin. Much work and many challenges were still to come.

Published April 5, 2018   •   Updated May 13, 2020
© Norsk Oljemuseum
close Close