IMO regulations require a minimum safe distance between LNG tanks and the ship’s shell, which further restricts the naval architect’s design space. RINA article describes the idea of LNG tank placement which provides more design freedom without compromising safety, or even improving safety.
Safety considerations
IMO’s draft international code of safety for ships using gases or other low-flashpoint fuels (IGF Code) stipulates a minimum safe distance between a storage fuel tank and the ship’s shell of 1/5th of the ship’s beam, B. At other levels the distance should never be less than 760mm (IGF 2014).
This requirement has been taken from existing IMO regulations for sea going gas tankers (IGC 2014), which has proven satisfactory over the past 50 years. In fact the B/5 requirement has been copied from the early SOLAS regulations on damage stability which is inspired by damage statistics recorded between 1948 and 1966.
It shows that, given a collision, the probability of the damage penetration not exceeding 1/5th of the ship’s beam equals 55%. It also shows that in 45% of the collision cases, this distance is exceeded. Obviously these statistics are outdated; ship structures have changed significantly since the late sixties of the previous century, so have ship sizes.
Double wall structures, bulbous bows, framing direction, improved material, tenfold increased gross tonnage (to name a few) all have influence on the resistance and impact during a collision. Another aspect is crashworthiness, that is the ability of a ship structure to resist a collision, which depends largely on the actual structural design and applied materials. This aspect is now ignored in IMO regulations.
A common way of dealing with safety, or rather risk in general, is through conducting a formal safety assessment. It defines risk R, as a multiplication of the probability of an undesirable event occurring by the effects of such an event. Likely events with severe effects are unacceptable. However, when effects are ‘slight’, a high occurrence probability is acceptable. Events with severe consequences, like an air craft crash, are acceptable only when their probability of occurrence is sufficiently low.
Crashworthiness and the probability of flooding
About 10 years ago, a European project was carried out on the effect of crashworthiness on the probability of flooding (Crashcoaster). The results were reported to IMO in an INF paper submitted by Germany (IMO 2003). It demonstrated that predicting the collision energy absorbing capacity of a ship structure is feasible without excessive efforts.
This approach is in fact adopted by the European inland waterway authorities albeit for tankers carrying hazardous cargo aiming at reducing the probability of loss of containment, rather than damage stability. It is now part of the regulations on the carriage of hazardous cargos on inland waterways (ADN 2014, chapter 9.3.4). As a consequence, many inland waterway tankers are now fitted with a crashworthy side structure
Crashworthiness of LNG tanks
Following the work done in the European project, an initiative was taken to investigate the crashworthiness of LNG storage tanks with respect to ship collisions. The main goal was to establish the actual tank vulnerability with the ultimate aim to calculate the probability of tank rupture and consequential loss of containment.
Since currently no technical evidence exists on the effects of a rupture of a pressurised vacuum tank carrying LNG fuel, one must fear for the worst, that is suffocation, BLEVE , fire and/ or loss of structure due to the brittle fracture of ship steels at cryogenic temperatures. Hence the application of such tanks is only acceptable when the probability of rupture and consequential total loss of containment, is sufficiently low.
There were two technical issues to solve:
- Will the tank rupture when the volume reduction due to the tank deformation during a collision exceeds the volume of the initial vapour space over the liquid (liquid full condition) ?
- Will the structural material of the tank still behave in a ductile manner when subjected to a collision impact in cryogenic conditions?
- The crashworthiness FE simulation software can cope with fluid/ gas structure interaction as well as temperature dependent material behaviour. The software was amended with a routine which calculates the steep internal pressure build-up in the tank in case of a ‘liquid full’ occurrence. Also tank material tests were done with full thickness specimens at cryogenic temperatures and realistic deformation rates.
Towards the tank ends the diameter of the tank increased during the crash, due to the ‘liquid full’ condition. It is remarkable that the tank shell does not rupture, not even at large deformations. Apparently it acts as a fender. Moreover it is very encouraging to see that the simulation results match the test results satisfactorily, both with respect to structural strength and as gas-liquid-structure interaction.
Having established the validity of the simulation method, it was decided to analyse a typical pressurised cryogenic vacuum tank of 35m3, representative for inland waterway shipping. Figure 8 shows the predicted deformation in case of a collision by an inland waterway push barge.
A ‘full’ tank tends to show a larger collision energy absorbing capacity compared to tanks with smaller filling ratios, mainly due to the ‘fender’ effect. It is interesting to note that collision energies available at the European inland waterways are assumed not to exceed 22 MJ (ADN 2014). Hence the absorbing capacity of the tanks is substantial. When scaled to sea going vessels, similar effects are to be expected.
A weak point of fuel tanks, as they are adhering to IGF instead of IGC, is the pipe penetrations. For storage tanks, one dome with various safety measures is required, while for fuel tanks multiple penetrations are allowed. Although the tank may not rupture, penetrations of piping into the tank probably will due to this inherent weakness. The associated outflow areas are expected not to exceed those of the piping. If this is indeed the case, loss of containment does not occur instantaneously and is expected to be manageable. There is no heat source, so evaporation goes slowly. This, however, still needs to be investigated.
Replacing the B/5 requirement
In this paper, it is shown that the crashworthiness of LNG fuel tanks which contain fuel can be calculated. Moreover, the energy absorbing capacity is significant as fuel tanks act as fenders. However, the piping is a weak point and needs greater attention.
With these results available it is proposed to consider the vulnerability of pressurised cryogenic LNG tanks with respect to collisions explicitly. As yet, no probabilities of tank rupture were determined. However, such probabilities can be calculated through correlating existing damage statistics databases, in conjunction with automated identification of ships (AIS) shipping data, with collision impact absorbing capacities (crashworthiness) of cryogenic pressurised vacuum tanks together with the resistance of the ship structure. With such evidence, acceptable safe distances between tanks and the ship’s outer shell can be determined for actual ship designs instead of hiding behind the B/5 rule, especially considering this rule is based on statistical data that is more than 50-years old.
The explicit approach of crashworthiness and effect analysis of ‘slow’ loss of containment is considered to be most beneficial particularly for dredgers, short sea shipping, ferries and cruise ships.
Source: The Royal Institute of Naval Architects
In the outbreak, I was outspoken with you propecia before and after has changed my being. It has become much more fun, and now I have to run. Just as it is incredible to sit.
