“If the LNG terminal offshore Tuscany explodes, the Leaning Tower of Pisa will fall over”
This was stated by an Italian professor in the local newspaper during the approval phase of the now operational LNG terminal offshore the coast of Tuscany. To me it’s just another example that neither a PhD nor a Professor title is proof of a capability of rational thought. The fact is that an explosion of such magnitude is impossible. But let’s not get ahead of ourselves, let’s start with the basics.
The main safety aspects of LNG can be divided into two main topics:
- Cryogenic effects from LNG
- Flammability and explosion, including BLEVE and RPT
A spill of LNG might occur. A tank may rupture due to external impact. A flange may leak. A pipe may break. In such events, the initial consequence will be cryogenic effects from being exposed to a liquid at -163 degrees Celsius. Humans will freeze and steel will go brittle. Brittle steel may easily break and lead to secondary failures of other equipment.
Once the LNG has leaked, it will form a pool of liquid LNG. This pool will start to evaporate and form a cloud of gas, primarily consisting of methane. This gas will start mixing with air and once it reaches a mixture with between 5 and 15% gas, it is ignitable. The following sequence of events will depend on whether there is an ignition source, and whether the gas cloud is contained within a confined space.
Ignition source? Yes or no?
- No: The gas will continue to evaporate, disperse at ground level while cold, but start to warm and rise to the sky as methane is lighter than air, and thereafter drift away until the entire liquid pool is gone. Nothing will be left.
- Yes: The gas cloud will ignite.
If yes to above – Confined space? Yes or no?
- No: There will be an initial poof, not very violent, as the gas cloud ignites, and it will burn back to the pool as a flash fire. The gas will continue to burn as it evaporates off the LNG as a pool fire until the pool is gone.
- Yes: There will be an explosion causing overpressure and drag loads and potential damage on structures and buildings
Worst case scenarios for LNG accidents, such as the one described by the professor above, typically calculate an explosive release of all the energy in the cargo. This means that all of the 130 000 m3 of LNG stored in the terminal would have to leak, then evaporate to turn into 78 million m3 of gas, then mix with air to turn into 780 million m3 of ignitable gas/air mix (assuming 10% gas). If those 780 million m3 could be contained inside a confined space, then you would have a pretty potent bomb.
A balloon with sufficient resistance and a diameter slightly larger than 900 meter could make such a confined space. And of course you’d have to wait until all volumes are put into the balloon before igniting it. Needless to say, this is not a credible scenario. Another unlikely scenario is that all the gas evaporates and floats at ground level and avoids all ignition sources until a perfect moment when the whole gas cloud is within the 5-15% mix with air, and then ignites.
Most likely, the event causing the leak will also ignite the gas, causing only a low pressure flash fire and then there will be a continuous fire as evaporated gas burns off the liquid pool. This will be a catastrophic fire, similar to fires of any other fuel, but it will not cause explosive overpressures.
BLEVE is a phenomenon that can happen when a pressurised liquid gas tank is subjected to a sustained external heat source such as a neighbouring fire degrading the structural integrity of the tank. The degradation of the integrity can lead to a sudden rupture of the tank, and in the event of such a rupture the boiling liquid simultaneously expands and ignites causing a powerful explosion and thermal dose. BLEVE can only occur with pressurized tanks, it can’t happen to tanks with atmospheric pressure which is what is used for all large scale LNG terminals and ships.
RPT is a phenomenon that may occur when LNG is released onto water. The water will cause quick heat transfer into the LNG making it a superheated liquid. Once evaporations starts the LNG will evaporate instantly and cause a pressure pulse. RPT is a flameless explosion that can be compared with the cracking noises (small explosions) when heating cooking oil with small amounts of water inside. Significant damages caused by the phenomenon are not expected and have not been observed.
In addition to the main safety topics above, there are a few other special phenomenon of more operational interest:
- Roll-over may occur in tanks where the LNG is stratified, leading to a build-up of pressure in the lower layers which can eventually release causing sudden vaporization (boil-off). Normal procedure now ensures regular mixing to prevent stratification.
- Sloshing is the effect of waves inside LNG tanks on ships. The consequence of this may be damage to the tanks and potential leaks of LNG.
Yes, natural gas burns. And this is good news, because this is why it can serve as a fuel. Also, under certain circumstances, natural gas can cause explosive over-pressures This is why any equipment utilized for gas and LNG must be designed in ways where the likelihood of events with high consequences are minimized. As an example, pressurized gas tanks must be located in places where external impacts are unlikely, or they must be shielded from such impacts. Another example is arrangement of equipment so that potential jet fires will not directly hit critical equipment.
In order to ensure safe design, risk management is an integrated part of the design process. To be more concrete, this is a process that follows certain defined steps: hazard identification – risk analysis – risk evaluation – risk treatment. This is a quantified process, so that it can ensure that the safety of the object is within whatever acceptance criteria that apply.
The risk analysis part of the risk management process can take many shapes and forms. It depends on what sort of risk needs to be analysed. There are various computer tools available for this analysis, for example tools for analysing the dispersion of gas clouds and the severity of fires and explosions. For a wide range of LNG leak cases, relatively simple tools can be applied, but for more complex geometries, then CFD (Computational Fluid Dynamics) models are needed.
With a proper risk management approach in both design and operation, gas and LNG are equally safe, or even safer than other fuels.
Should you wish to dive deeper into LNG safety, I can recommend a book by my colleague, Robin Pitblado: LNG Risk Based Safety.
