Final article in a series of five articles by Bill Campbell, retired HSE Group Auditor, Shell International, about risks relating to the Shell Prelude FLNG project.
Prelude FLNG risks are on par with modern offshore oil and gas facilities say Shell – but are they? Let’s discuss
With the implementation of the recommendations post Piper A, turned quickly into legislation, the potential consequences of hydrocarbon releases have been markedly reduced, but Floating FLNG facilities cannot comply, other than that front end gas feed from the reservoir will be shut in and the process gas flared, huge amounts of volatile hydrocarbon liquids remain stored in the hull, which is also the primary structure supporting the process, utilities and the living quarters.
Prelude for example has in its hull, tanks with a capacity to hold 220,000 m3 of LNG, when the cryogenic liquid is returned to gas this equates to 132 million m3 of methane. It also has capacity for 90,000 m3 of LPG and 126,000 m3 of Condensate, with an overall capacity Shell states equivalent to 175 Olympic swimming pools.
If you compare this to a modern North Sea gas platform, they generally have a jacket supporting the Gas Process and utilities with a permanent bridge – protected from blast by the bulk of the utilities modules – across to a second independent jacket containing the Temporary Refuge (TR). It appears extraordinary to suggest that a FLNG installation can have the low risk levels that can be attained by such an arrangement.
Offshore installations have the unique requirement to be self-sufficient should an emergency occur. This legislation discussed above when enacted was credited with significantly reducing risks on North Sea offshore installations by limiting the potential for escalation and allowing through this process Operators (including Shell) to demonstrate in their submitted Safety Cases they could meet the stringent TR survival criteria.
That this legislation was needed is reflected in the continuing failure of North Sea operators to prevent (despite this being publicly stated to be the Industries highest priority) hydrocarbon leaks. These are still regular occurrences, from 86 in 2008 to 43 over the last reporting period. The 43 consisted of 35 significant releases and 9 major releases. According to the Regulator’s own definition a major release for example has the potential to quickly impact out-with the local area to affect the TR, its escape routes and other areas of the installation, causing serious injury or fatalities. This happened on average every 40 days during the last reporting period. For Prelude, or any other FLNG installation for that matter, to come anywhere near the low risks of a modern gas platform it would need to be immune from the chronic disease that affects the North Sea by having during its operating life no significant or major hydrocarbon leaks at all, full stop, period.
For LNG plants, size matters!
The second issue that makes the risks of FLNG so high is the minute footprint of the hull deck. Shell states that Prelude takes up a quarter of the space taken by a land plant of similar capacity but can that statement be supported by fact? Are there any onshore plants in the world producing LNG, LPG, and condensate with similar throughput and including the offloading terminal and storage tanks that occupy an area just 4 times the deck area of Prelude?
The FLNG concept places a hazardous substances plant on a vessel on the high sea, where the confined footprint of the hull of these vessels cannot provide the separation distances afforded to land plants. These distances, considered necessary by law are to protect, not only employees within the plant fence, but communities that may be near the plant. The footprint of the Prelude hull is 488 by 74 m, or 36,112 square metres, some 3.6 Hectares (Ha), or circa 9 acres in old money.
To overcome this lack of space on Prelude the various process modules are stacked on top of each other requiring a large amount of passive fire protection, active systems such as seawater deluge being inadequate to cope with gas jet fires… To reduce risks it is common for hazardous substances plants, including LNG plants on land, to occupy up to hundreds of acres. This is because in simple terms, mitigation of the effects of explosion and fire and loss of containment is reduced by distance separation. The greater the distance the less is the probability of escalation or collateral damage from an initiating event.
Researching through the internet I cannot find land usage for a LNG plant either in existence or proposed that is anywhere near the meagre proportions of the Prelude deck, granted Prelude is a huge vessel because it has to be, and the continual overemphasis by Shell of its size detracts from the fact that as an LNG plant it takes up a very small area.
The total Prelude deck is 3.6 Hectares (Ha) or if you like, this equates to 1 Hectare per million metric tons per annum (mtpa) of LNG produced. This can be compared to Western LNG at Nisiski, Alaska, a 3 mtpa capacity plant which has the smallest land area I can find at 24 Ha or 8 Ha/mtpa. So it is 8 times the footprint of Prelude. To give you an idea of what this means is that this footprint would make the Prelude deck 3.9 kilometres long if the width were to remain as is at 74 metres.
Pluto LNG, Western Australia, is a 4.3 mtpa plant but covers 80 Ha, or 18.6 Ha/mtpa. Woodfibre, Vancouver, is a 2.1 mtpa plant and covers 86 Ha or 41 Ha/mtpa and Gladstone, Queensland, has an initial 4 mtpa capacity but has a land area of 190 Ha or 47.5 Ha/mtpa. If we discount the Gladstone plant which may expand its capacity in later years the average of the first three plants in terms of footprint per million metric tons of LNG produced per annum is a staggering 22.5 Ha/mtpa, so not four times larger than the Prelude footprint, as Shell press releases suggest, but circa 22 times larger taking up an area of not 3.6 hectares but circa 79 hectares. To scale this up and give you an idea of what this means Prelude would have to be over 10 kilometres long to align with this footprint of the average onshore LNG plant using the small sample discussed.
So can anyone, anywhere, come up with a 3 to 4 mtpa LNG plant on land that takes up just 14.4 Ha which is 4 times the actual Prelude footprint of 3.6 Hectares which Shell declares is the area that Prelude would take up if installed on land. If such a plant exists Shell perhaps can tell us where in the World it is.
FLNG facilities are therefore at increased risk because of the confined area of the hull footprint. If Shell does not accept this then the industry certainly does. With reference to a paper by the Technip HSE Design Department quote there are many safety issues associated with the development of this new kind of facility. The potential risks due to the processing of hazardous substances which are magnified due to the close proximity of the liquefaction facility to the living quarters. FLNG facilities present new challenges in safety that result from a combination of complex processes, hazardous process fluids, harsh marine environment and a reduced footprintcompared to equivalent onshore installations unquote.
Total also got into the act quote in this type of floating plant there can be no compromise when it comes to safety, it’s confined spaces must accommodate living quarters as well as operations (liquefaction, storage and offloading) that involves large inventories of flammablegases. The major risk is blast that could set off other explosions in a domino effect unquote.
There are other risks that appear specific to Prelude FLNG which have been managed differently by others. The living quarters on Prelude, and because of the orientation of the vessel into the wind as it rotates around its single point mooring, is almost constantly downwind of the process and some 100 m from it.
The Total risk reduction strategy, and also that of Technip in their prototype deign for an FLNG facility, supports tandem offloading using cryogenic rated flexible pipe to transfer the LNG to the carrier vessel stationed 100 metres or so to the rear of the FLNG installation, rather than side to side transfer, as is the case with Prelude. This limits in their words the risk of collision. They also locate the accommodation at the bow to avoid being downwind of the process and as far as possible from the offloading area where the risk of LNG leakage is the highest in their opinion.
Fixed offshore installations, modern or otherwise, in the Norwegian or UK sectors of the North Sea take into account the prevailing wind direction statistics, and locate living quarters where they are most of the year upwind of the production and drilling processes – as a risk reduction measure. So in summary there are 4 factors that make the risks of Prelude FLNG higher than for conventional offshore installations handling gas and for a conventional LNG plant dispersed over a large area on land.
- Prelude stores within its hull, the primary structure that supports the process, a huge inventory of volatile hydrocarbons. The risks are therefore high because the potential consequences are higher should a major conflagration occur
- Offshore LNG plants such as Prelude cannot comply with the separation distances provided by the land area of onshore plants so the potential for collateral damage and uncontrolled escalation is greater. Although Prelude is a huge vessel it provides a very small footprint as an operating hazardous substances plant
- Prelude working almost continually with side to side offloading onto a carrier vessel has the combined risks of simultaneous operations as explained in previous articles and in particular the risk of collision, fire or explosion on the carrier vessel with subsequent collateral damage to Prelude. Spillage of the cryogenic LNG during offloading could also cause fracturing to the deck plates of one or both vessels moored in close contact
- Side to side loading increases the risks of LNG leakage impinging on the integrity of the living quarters. Other than in becalmed conditions, the living quarters being located downwind of any such release is such that the resulting methane-air vapour cloud that forms could simply circumvent the blast wall. Another risk consideration is that the change of phase from liquid to methane is accelerated when the LNG comes into contact with the sea when it vigorously reacts with seawater
To expand on Paragraph 4, it is the evaporation volumes as the liquid returns to its natural state that causes concern, just 100 m3 of LNG will convert into 60,000 m3 of methane and as the gas disperses it will enter its explosive range at 150,000 ppm or 15% methane in air when it reaches a volume of 400,000 m3 and is still explosive at 50,000 ppm or 5% when theoretically the vapour cloud has a volume of 1.2 million m3. If we consider the dimensions of the vessel at 488 long by 74 wide by 33.3 metres down to the water line this represents a volume of circa 1.2 million cubic metres. So 100 m3 of LNG can in windless conditions create a dome shaped cloud that engulfs the entire vessel and if a wind exists, as is the norm, such a cloud could engulf the TR, its escape routes and the lifeboats in a matter of minutes.
By implication therefore FLNG offshore installations must operate at higher risk than other offshore gas installations. Prelude FLNG specifically is designed with its living quarters downwind of the process and uses side to side loading both factors which others in the Industry accept increases risk.
There needs to be a more honest and open debate about the risks of FLNG. As I read the input of both Technip and Total to the subject their contributions are more sober, more reflective of realty and clearly some considerable assessment of FLNG risks have been undertaken by them. In contradiction to this are the bland, unsupported by any detail, statements by Shell as indicated by the title and introduction.
Again if I can refer to Technip quote FLNG being a new technology is not covered by existing Regulations, codes and standards, and good engineering practices as they do not account for the specific safety concerns of this new concept such as cryogenic risk with the process modules of an offshore facility unquote.
Whether you agree with the statement or not it is evidence of a more cautious and considered approach by the Industry than that adapted, at least publicly, by Shell, to what will be a very risky business. Risk assessment is not, or should not be an emotional or subjective exercise. It is not about pessimism or optimism, it’s about can it happen? And further, does it happen frequently? And can the consequences should it happen be potentially catastrophic? Sadly the answer to all these questions is yes, and any risk matrix worth its salt will recognize this as fact based on historic failure data and not ignore it.
While it is true that LNG leakages at onshore plants are few and far between when they have occurred the results can be catastrophic. There is also more commonly spillages during offloading, this has happened to Shell amongst others causing damage to the vessel deck plates. Also the venting LNG vapours has been ignited on Shell vessels (by lightning) both at the terminal and on the high sea.
In my opinion, and its only an opinion, and I would welcome your opinion, because an open an honest discussion is needed. For Prelude to attain the risk levels of a modern conventional offshore gas installation as described in this article the risk of hydrocarbon release on board that vessel needs to be negligible, but this appears to be an unattainable target, certainly in the highly regulated North Sea, and despite this being that Industry’s top risk reduction target.
Prelude will need to be a very special vessel, with very special people to be leak proof, perhaps it can, but I have my doubts.