Beating the cracking problem of FLNG! Assessment of brittle fracture risk
For FLNG adopting the mixed refrigerant process where liquefied high pressure C1 and C2 are used, even a small amount of leakage may cause brittle fracture in FLNG made of steel. Although its frequency is uncertain, the potential damage could become unexpectedly devastating; the FLNG may be forced to stop production and leave the offshore field for permanent repair in a shipyard. Obviously, the operator’s loss of profit would be significant.
Protective measures therefore must be appropriately designed in terms of technical effectiveness taking account of associated risk such as fire & explosion, and its likelihood too. Therefore, prior in-depth risk analysis is essential.
Cryogenic service in gas liquefaction plants is markedly different from common offshore crude oil production units. This means straight application of the conventional methods and parameters of process leak analysis may not be accurate in forecasting what may occur at FLNG.
For this reason, leak scenarios for risk analysis are carefully devised so as to reflect past LNG leak incidents, operational practices and plant design practice.
A leak of high pressure cryogenic fluid is also called droplet-laden-flashing-jet (Bosh). Unless the vaporized portion of two-phase jet is reasonably assessed, pool sizes which are key to quantify brittle fracture cannot be properly estimated. Therefore, a special CFD solver is used which incorporates algorithm of HEM (Homogeneous Equilibrium Model) and surface pool model which allow calculation of 3 dimensional dispersion of liquid & vapor as well as the resultant liquid pool formed, with temperature and depth.
More than 100 HEM simulations have been conducted. They showed that the tendency of pool formation is rather limited for cases with larger hole sizes and lower leak points, and pools reach equilibrium within a relatively short period (3- 5 minutes).
Leak frequency for different, representative hole sizes on each type of equipment are calculated incorporating characteristics of cryogenic service, and also introducing a newly developed CCDF: Complementary Cumulative Distribution Function. Distribution of leak hole size follows in principle the power-law while there is an obvious physical upper limit due to equipment sizing. The CCDF developed well accounts for the phenomena and observations.
Finally, risk results are presented in the form of cumulative occurrence frequency of wet pool size by relevant areas. It is observed that the pool formation over process decks tends to be larger and more frequent. The typical indication would be in the region of 10meter dia. at 10-3 occurrences per year per train, whereas it is 5meter dia. at 10-4 for the hull upper deck.
The derived cryogenic pool sizes are further converted to facility damage to estimate the repair cost and downtime. This process requires clarification of two aspects: the mechanism of brittle fracture and the conditions of each area exposed to the leak.
The onset of a fracture theoretically involves certain combinations of three parameters: microscopic flaw size in the steel product, (tensile) stress working, and material fracture toughness. Microscopic flaws reside in welded parts with higher probability than others. Also, welded parts contain residual stress due to high heat input. The Fracture toughness of steel is weakened when the material comes into contact with cryogenic fluid.
There are many welding lines on both the process deck and hull upper deck. In conclusion a simplified rule -when a cryogenic pool meets welding lines, it causes brittle cracks – is introduced.
The risk of cryogenic pool formation is higher at the process module deck than at the hull deck. However, this needs to comprehended the other way round due to susceptibility of damages to structural members.
The Hull upper deck is one of the primary members of the hull structure, and high repair quality is demanded when damaged, which leads to the conclusion that in case of meter-long damage to the hull upper deck, the work must be done in a well controlled environment like a dry dock. On the other hand, the process module deck is as such that can be replaced with gratings and does not carry importance as essential load bearing members since heavy equipment installed on the process deck is practically supported by primary and secondary beams beneath them.
Based on the above deliberation, it was concluded that due to the enormous amount of loss of profit for hull repair off field, the hull upper deck should be sufficiently protected even if the likelihood is remote, whereas the level of protection of the process module deck can be left to the risk allowance of unit owners.
Special carbon steel has been developed having standard chemical composition like hull structural steel. It is produced through TMCP (Thermo-Mechanically Controlled Process) to refine crystal grains and obtain a higher level of fracture toughness.
The developed steel went through a large scale structural experiment and proved it is able to stop the propagation of long major cracks even at -45 deg. C, contrary to D grade steel which fails at only +10deg.C. However, it is said that the characteristics could not be seen in the HAZ (heat affected zones).
In conclusion, the hull upper deck must be sufficiently protected from brittle fracture due to its susceptibility. By having the process deck level completely plated, the chance of cryogenic splash reaching the hull upper deck level can be made fairly remote. Special TMCP steel with higher crack arrestability can provide better relief than ordinary steel, but still needs to be protected, especially at HAZ. The optimum combination of employing special steel and insulation material such as syntactic epoxy foam material would be one reliable solution.
Kensaku Maeda will be presenting FLNG Cryogenic Leak Risk Analysis and Study on Employing TMCP Steel with Higher Crack Arrestability at the Cryogenics stream at Gastech 2012 Exhibition’s Centres of Technical Excellence (CoTEs). To learn more, register here for this free seminar on 10 October 2012 at the ExCeL centre in London.
Sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.