In any analysis of which oil and gas player was the most technically competent and cutting-edge, Shell would have to rank near the top, if not number one. Indeed, the company's own history boasts that it is "one of the largest and most innovative energy companies in the world" (here). Shell has consistently and successfully delivered complex and high technology projects for over one hundred years: from Russia's first offshore gas field in the icy waters off Sakhalin; to the deepest oil and gas project in the US Gulf of Mexico – Stones – in 2,900 metres of water; to the world's largest gas-to-liquids plant, Pearl GTL in Qatar.
This makes the expensive fiasco of the Prelude Floating Liquefied Natural Gas (FLNG) project in Australia even more embarrassing for the company. Shell and its minority partners have spent an estimated US$17 billion to US$19 billion on the Prelude project, including the seven production wells tied back to the facility in 250 metres of water. Most of this investment has been spent on the gargantuan FLNG vessel itself, which was built at Samsung Heavy Industries in Korea.
Incredible video of the construction and mooring can be found here.
At 488 metres LOA, the Prelude FLNG vessel is longer than the Petronas Towers in Kuala Lumpur are high. It displaces 600,000 tonnes when fully laden, the same volume of water as five of the largest aircraft carriers in the world. Some 260,000 tonnes of steel went into building the hull and topsides. At 74 metres wide, it is wider than the wingspan of a Boeing 747 Jumbo jet (64 metres). The mooring turret, designed by SBM, is nearly 100 metres high.
Shell hoped that the Prelude unit would produce 3.6 million tonnes of LNG a year, as well as 1.3 million tonnes of condensate, and 400,000 tonnes of liquefied petroleum gas. Onboard, the methane is chilled to −162°C (−260°F) and, at maximum capacity, the facility will draw 50 million litres of cold water from the ocean every hour to help cool the natural gas.
Production is stored in six LNG tanks (total capacity of 221,000 m³), four LPG tanks (total capacity of 90,000 m³), and six condensates tanks (total capacity of 126,000 m³), the same volume as 175 Olympic-sized swimming pools, Shell's breathless publicists claimed.
Around 1.6 million hours were spent designing this engineering masterpiece.
Unfortunately, the Prelude FLNG doesn't work. It hasn't worked since February 2 when it was shut down after a run of alarming safety incidents, culminating in a power failure. The technical issues are documented by Peter Milne on the Boiling Cold website here.
The toilets not flushing is actually the least of Shell's concerns. Even in the brief period when it was functional, Prelude was only producing at 50 per cent capacity. It will not produce again until next year. When, exactly, it will produce at full capacity, nobody knows.
Prelude has been an expensive, disastrous white elephant. Inpex, which owns 17.5 per cent of the project, wrote off more than one billion US dollars in an asset impairment on Prelude in August, plunging the company into a loss. Shell owns a 67.5 per cent stake in Prelude and announced at the end of June (here) that it was impairing its Australian gas assets, principally Queensland LNG and Prelude, by up to US$9 billion (with a "b") due to lower gas prices compared to the assumptions on which the projects were based.
There are only three other operational FLNG projects in the world; two in Malaysia operated by Petronas (PFLNG Satu and PFLNG Dua), one of which hasn't yet actually been commissioned but is on location, and one in Cameroon operated by Perenco, but owned by Golar LNG, using the vessel Hilli Episeyo. A fourth FLNG, being built by Italy's ENI in Korea for deployment off Mozambique, is due to be installed and start operation in 2022.
Like Prelude, ENI's Coral Sul monster is also designed by Technip and is of a similar size, but slightly smaller (production capacity of 3.4 million tonnes of LNG annually, and "only" 432 metres LOA and 66m wide, which is, incidentally, still wider than the wingspan of a Boeing 747, in case you were wondering). Operating in more than 2,000 metres of water, Corul Sul will be the world's first ultra deep-water FLNG unit, with an estimated project cost of US$8 billion.
Ironically, the cheaper, simpler projects, run by companies not known for their technical prowess, in countries much poorer than Australia, are in full production. The Perenco project was based on the conversion of a Golar LNG tanker hull built in 1975 on which four LNG trains, each with approximately 500,000 tonnes capacity per year, were fitted.
The Golar hull was forty-three years old when the project produced first its LNG in 2018. In July this year Hilli Episeyo celebrated loading its fortieth cargo for export offshore Kribi. Converted by Keppel Offshore and Marine in Singapore, the designers have claimed that the converted FLNG vessel saves approximately 33 per cent of greenhouse gas emissions compared to an FLNG newbuild.
Keppel is building a track record in these conversions and is working to convert another LNG tanker to an FLNG plant. This will be sheltered behind a breakwater in shallow water for BP's Greater Tortue Ameyin project off the coast of Senegal and Mauritania for delivery in 2022. To Keppel's relief, BP lifted force majeure on the project on October 1 (here).
The Malaysian units have boiler plate capacity of just one third the capacity of Shell's unit, at just 1.2 million tonnes of LNG annually for PFLNG Satu, and 1.5 million tonnes for PFLNG Dua. The 365-metre LOA PFLNG Satu operated without much ado on the Kanowit Field in 80 metres of water off the Malaysian state of Sarawak from 2017 to 2019, when it relocated to the Kebabangan gas field, exactly as the proponents of LNG had believed was possible. The ability to move from location to location to produce from smaller fields is a major advantage of FLNG over fixed structures and pipelines to shore.
The full brochure for PFLNG Satu is available here. The vessel's tanks can store up to 177,000 m³ of LNG and 20,000 m³ of condensate, which Technip says is capacity equivalent to about 79 Olympic-sized swimming pools. Clearly, Olympic pool volume is becoming a standard unit of measurement for FLNG vessels. Who knew?
The slightly larger PFLNG Dua sailed from Korea in February to Rotan field off Borneo to be moored in 1,500 metres of water. No reports of first export loading have yet been received, and we understand that commissioning has been delayed by Covid-19 restrictions.
This graphic from Westwood and the SPE shows the problem: Prelude is both significantly larger and significantly more expensive than the other three completed FLNG vessels, and far more pricier than ENI's unit. In Prelude's case, bigger is definitely not better, especially when it doesn't work.
Shell's woes have a major impact on Australia. When Prelude loaded its first LNG shipment for export last year, the Australian Petroleum Production and Exploration Association (APPEA) said that Australia was "continuing to reap the rewards of more than AU$350 billion (US$250 million) worth of investments that will see LNG exports surge to more than 85 million tonnes by 2020" (here)
That hasn't happened. And Prelude isn't the only Australian LNG plant facing operational and reliability issues. Earlier this year, Chevron found cracks on propane heat exchangers at the onshore Gorgon LNG plant, which led to ongoing production shut-ins for repairs on one train and for inspections on the other two trains. Like Prelude, Gorgon was plagued with delays and cost overruns and finally came in with a total project cost of US$54 billion.
Woodside's North West Shelf LNG plant also shut down a train for maintenance. As a result, EnergyQuest, a consultancy, estimated that Australia shipped just 6.2 million tonnes of LNG, in 91 separate loads in September.
APPEA's prediction of 85 million tonnes of exports this year will be disappointed by a significant margin. Australian LNG plant capacity utilisation fell from 88.8 per cent in August to 86 per cent in September, EnergyQuest reported. Prelude's remained zero.
I am not going to cast aspersions on Australian engineering prowess with LNG plants, given that Equinor's Hammerfest LNG plant in northern Norway recently suffered a catastrophic fire on September 28, and will be closed for repair for most of 2021 as well.
Prelude's problems are unfortunate, because the world needs more LNG as a transitional fuel to replace coal, which is significantly more polluting than gas (here). And the world needs FLNG to bring stranded gas reservoirs online in medium to deep waters and in remote locations, which cannot be economically viable when tied back to shore-based facilities.
We have seen in both onshore hydrocarbon fracking in North America and in offshore wind in Northern Europe how widespread implementation has increased efficiencies and delivered economies of scale, so that the costs of both wind farms and fracking have dropped. For renewables, this has created a virtuous circle, so that as more turbines are installed, incremental gains in technology and operational efficiency are won, and costs drop further, making the economics even better over time. If more FLNG schemes were to go ahead we would see major cost and operational improvements to the concept.
Petronas's decision to move into FLNG was designed to address the country's falling LNG exports by exploiting gas fields too far from shore and/or too small to justify a subsea pipeline and fixed infrastructure to connect them to Petronas' existing onshore LNG plants in Borneo.
Meanwhile, ENI chose FLNG for offshore Mozambique because of geological issues between the shore and the deep-water gas field:
"FLNG is the best solution possible, both from a technical and production point of view, as well as in terms of the environmental impact," ENI's Coral FLNG project development director Stefano Rovelli has said. "The field is quite far from the coast, and the seabed has a canyon that is up to 13,000 metres wide and up to 300 metres deep, which makes it difficult to lay pipes to transport the gas to the shore. With an FLNG, the project's footprint is minimal, limited to the underwater area where we will install three manifolds, the moorings for the vessel and the flexible lines for connection to the FLNG."
FLNG made sense in Cameroon because of the remote location, the lack of skilled labour in country to build a land-based FLNG plant, and the lack of engineering facilities. It made more sense to convert an LNG carrier in Keppel shipyard in Singapore and then sail it to Cameroon to plug into the gas well and produce than to try to build a facility on land in a country facing an insurgency, corruption, and political instability (here for the woes of President Paul Biya).
The same arguments would apply in Tanzania and Equatorial Guinea. When it bought BG in 2015, Shell acquired large undeveloped offshore Tanzanian gas fields, just north of Mozambique's. Off Malabo, the Fortuna project in Equatorial Guinea has long been considered an FLNG opportunity, if the right cost model was in place. FLNG is a solution which requires implementation.
Similar challenges of problems exist elsewhere with other undeveloped gas discoveries. Shell's failure to make Prelude an operational success means that these other fields could remain stranded. For Shell itself, Prelude's problems mean that its Crux natural gas and condensate reservoir is in limbo, despite regulatory approval for the project in August.
Crux's field development plan is based around a massive normally unmanned offshore platform controlled from Prelude, with a 10,000-tonne topside on a jacket 193 metres high receiving production from five production wells. Shell will then lay a 165-kilometre subsea pipeline to take the gas to the Prelude FLNG plant.
Shell owns 82 per cent of the project and its partners have claimed first gas will be due in 2026, by which time one might hope that the Prelude facility would be back in service. Honestly, who knows?
Woodside had originally selected FLNG as the concept for its Browse field complex off Western Australia in 2013 but was deterred by the cost. Now the Browse concept is based on two identical FPSOs which would process and export dry gas to the existing infrastructure at the Karratha Gas Plant via the Browse Trunkline and a second trunkline. But the project has still not received investment approval.
There are numerous other Australian gas fields which could be developed using FLNG, if a cheap and reliable solution was available, including the Equus, Clio-Acme, Cash-Maple and Bonaparte fields. Instead, the owners of these reservoirs have struggled with their negotiations with onshore LNG plants to sell their production to them, and with their financiers over concerns about reservoir quality and durability of production if they install fixed facilities.
Further afield, Shell is desperately seeking a solution to its 35 per cent stake in the Abadi gas field offshore Indonesia. The gas field was first discovered in 2000, but remains completely undeveloped after two decades, as the Indonesian government and Shell and 65 per cent project operator Inpex argue over the development plan.
The Abadi scheme, in the remotest corner of Indonesia in the Arafura Sea right up against the Australian maritime border, would be a perfect candidate for an FLNG facility, if a proven system was available quickly and easily. It isn't, so Inpex has been compelled to agree to a shore-based LNG facility, which will be much more expensive on an environmentally sensitive island (shades of Gorgon, above). As a result, in August this year, Shell announced it was planning to sell its stake in the scheme, further delaying development plans.
Read the Jakarta Post's coverage of the ill-fated Abadi project here, and about Indonesia's largest ever corruption and embezzlement trial here.
The biggest loser of all from the failure of Prelude is Timor-Leste and its long-stalled Sunrise LNG project. Here again, development has been postponed by a tug-of-war between the government, which wants production to be carried out in-country, and the energy companies, which argue that a shore-based LNG plant in an impoverished, remote location is not feasible, especially since the 286-kilometre pipeline across the Timor Sea from the Sunrise field to the proposed Tasi Mane LNG plant in Timor-Leste has to cross a 3,000-metre-deep trench.
You will recall above that ENI in Mozambique opted for FLNG for the Coral Sul field precisely to avoid such a geological feature, and that Perenco in Cameroon opted for FLNG to avoid the cost of a greenfield newbuild in the middle of nowhere. FLNG was long mooted as the solution, but was not cost-effective and Prelude's problems have made it harder.
In 2019 Shell and ConocoPhillips sold their stakes in the field to the Timor-Leste government for US$650 million, giving it a majority 56.56 per cent interest in the upstream fields. The field operator Woodside holds a 33.44 per cent interest and wrote down the value of the asset to zero in August.
This is literally a stranded, unloved asset which would be viable commercially and operationally if a rented FLNG unit was available. Perhaps Dili should talk to Kuala Lumpur and get Petronas in with one of the PFLNG units…
Note that in a country with a GDP of just US$1.6 billion financing of any shore-based facility by the government isn't viable. The Diplomat reported (here) that the government's proposed land-based LNG facilities in-country would cost US$18 billion, more than ten times the country's GDP. East Timor even tried to access a US$16 billion Chinese government loan to proceed with the shore plant. Why they wouldn't copy ENI in Mozambique and spend US$8 billion on a Coral Sul-style FLNG is not clear.
Prelude is a disaster for Shell, but a disaster which is bad for Australian offshore gas prospects, and worse for the neighbours. I never thought I would write that Perenco and Petronas seem to have got it right, but they operate smaller scale, cheaper LNG projects which seem to, er, work and actually produce LNG offshore – which is the purpose of FLNG, whether the vessel is longer or shorter than the Petronas Towers.
For our coverage on the travails of Mozambique and its LNG ambitions, see here.
For more information on ENI's FLNG Coral Sul design and construction, see here.
A summary of Perenco's Cameroon FLNG operations is here.
Our splendid cost graphic comes from the fascinating paper "Financing FLNG Facilities—What Lies Ahead?" by Boris Adokou and Arindam Das of Westwood Energy which was published on May 7, 2018 and which can be viewed here: https://pubs.spe.org/en/ogf/ogf-article-detail/?art=4104
Nikkei reported Chinese efforts to ensnare East Timor in a US$16 billion LNG debt trap here.