Finding the balance
“Choose your problem carefully,” says Francesco Dioguardi of DH Industries. “When it comes to the development of LNG bunkering vessels each pathway you take has different consequences.”
His company is responsible for the reliquefaction plants onboard the first US bunkering barge: although it’s been developed at the same time as its European counterparts – one of which is soon to be deployed in and around Zeebrugge – it is a very different animal.
To start with the 107.6m Zeebrugge solution (developed by ENGIE, NYK and Mitsubishi Corporation), presently under construction in HHI’s Korean yard is self-propelled with a service speed of 13 kn.
Most importantly its 5,100 m³ LNG tanks are of a C-type design that, as Martial Claudpierre of Bureau Veritals explains, relies on good insulation to keep the liquid within three or so degrees of its -162C sweet spot.
Further, while typically expected to run at 4bar, these two robust tanks can take over 10bar pressures: a margin that will easily cope with any boil-off gas created by any slight warming of the LNG during transit. However its sphere of operations will probably remain within a few hours sailing of Zeebrugge’s Fluxys terminal and what little boil-off there is will be fed through to the Wartsila dual-fuel propulsion engines.
It has to be said it’s not a space saving configuration: these C-type tanks are of the ‘prismatic’ cylindrical design but as Mr Claudpierre explains, overall the choice of tanks for the LNG bunkering vessel has been driven by the business case “so if there’s bunkering demand is low this leaves the vessel able to work as a shortsea LNG carrier in North Sea area”. In other words, it’s not a regular string – yet – despite a few clients including a car carrier. So the tanks’ ability to keep the LNG in good condition for over a week is a necessary part of the deal.
By contrast, Mr Dioguardi explains the US push barge has been developed to handle a very long route, loading up in Vancouver and bringing the LNG product back down to feed Tote’s Orca class ships in Tacoma as well as handling the much shorter Jacksonville bunkering route. Aziz Bamik of GTT explains the bunkering barge has been designed to be flexible, able to load from trucks, feeder vessels and dockside and “to go wherever Tote most needs to deploy it”. So, unlike Zeebrugge, the US barge is being driven by a very secure business case: the customer is behind the barge in more ways than one.
But the US barge doesn’t have any propulsion system of its own, (due to crewing regulation quirks) so there’s no chance of using boil off gas itself. Given all this the Clean Marine Energy/Wespac venture has forgone the high pressure solution and, admitting the need for onboard chillers, the barge has been designed around a 2,200m³ GTT membrane tank which operates only a little over atmospheric pressure – typically 0.4 to 0.7 bar above normal.
Unlike the C-type solution, the membrane inside the Mark III Flex tank “almost breathes” says Mr Bamik, expanding and contracting to absorb the changing physical and thermal load and helping lower the boil-off rate. And, he adds, the ship’s hull lends the membrane some of its rigidity, making the most of the available hold space and minimising the barge’s footprint.
The Stirling cycle reliquifiers are obviously a necessary part of the solution, preserving the LNG at its peak calorific content: Mr Dioguardi explains that six of them have been set up onboard and this modularity lends a certain flexibility: “The different plants can be turned on or off depending on what’s needed, and there’s always one spare to deal with the returned vapour.”
But what about the potential differences in pressure, especially during transfers between C-type tanks and membrane tanks? Mr Claudpierre points out that although some have worried that transfers from a warmer, higher pressure installation “could blow up a membrane tank like a balloon” there are checks and balances in place along with controls to adjust the pressure, both before and during filling, although he admits that it all hinges on “good training… onboard as well as on the bunkering vessel”.
It’s difficult to say how the costs compare especially as prices vary from one continent to another. Martial Claudpierre points out that if you can afford the space, C-types can be less expensive than membrane varieties (which on a barge also need reliquifaction plants) although according to Mr Aziz the two are “more competitive” in the US. Despite this, Mr Claudpierre adds a scaling up of LNG bunkering demand will generally add to the cost-effectiveness of bigger membrane tanks.
However, it’s not just ‘either-or’: other combinations are coming to fruition. For example, Shell’s big 6,500m3 bunker vessel under construction at STX Korea has both type-C tanks and onboard reliquefaction to assist with LNG transfer as well as keeping it in peak calorific condition. Only time will tell if there’s a clear ‘winner’ and it’s just as likely the business case will continue to be the deciding factor.
Fuel tank challenges
Despite the number of gas carriers plying the seaways, translating appropriately sized C-type tanks into marine fuel systems has not been straightforward.
The first issue has been motion. Ann Rigmor Nerheim of Rolls-Royce (which has pursued gas-only rather than dual-fuel as a more environmentally-friendly option) explains that in a C-tank system the pressure is maintained by a pressure build up unit (PBU). The PBU simply evaporates some of the LNG, feeding it through to the top of the tank. However sloshing very cold liquid LNG around the tank affects this process by condensing the gas vapour, resulting in a reduction of the internal pressure. If this drops hard, the effect can be felt all the way to the engines – especially if a ship bunkers very cold LNG up to its maximum (95%) volume and immediately encounters very rough seas.
So, one part of the answer was awareness of the LNG bunkering temperature. However, investigation threw up a surprise: the severity of the pressure drop can be related to the tanks internal bulkheads: get it right and these dampen the sloshing action, get it wrong “and the sloshing action will be intensified” she explains. Rolls-Royce’s answer combines a better internal bulkhead pattern with a higher capacity PPU and possibly the addition of an LNG pump.
Footprint has presented designers with another issue. Trying to update old templates to accommodate LNG propulsion is challenging says Nina Savijoki of Deltamarin and likewise, retrofits don’t lend themselves to one-size-fits-all answers.
The Finnish company examined a number of options (including bilobe, spherical and cylindrical C-type tanks) on existing designs with range and autonomy kept in mind. Many showed intrinsic challenges: for example the best solution for a 120m, high speed ro-pax retrofit were three cylindrical tanks placed low in the vessel, two fore and one aft which maximised tank volume without impacting vehicle pay load – but as Ms Savijoki admitted “aft tanks are probably not feasible for diesel-electric azimuth arrangements… and three tanks makes it expensive”.
Retrofits have to be taken case-by-case: for example vertically arranged tanks were not much use for the kind of ro-pax previously examined especially as the necessarily narrower tanks would lose volume and further, the height could impact both car decks and stability – but potentially of interest slotted into the internal ‘service’ spaces on cruise ships.
Container vessels proved the most tricky of the set because they already pack in as many boxes as possible. Given class disapproval over placing the LNG against the sides and below of the accommodation block, this left the area aft of the superstructure – but at this point investigations showed up a need to reinforce the scantlings. “So even if we succeeded in keeping the container layout, it still meant reducing the overall container capacity because of the inherent additional weight in form of tanks, LNG and reinforced structures,” says Ms Savijoki.
Despite this, “it all changes completely when you design the vessel with LNG tanks from the start,” says Ms Savijoki. In fact a Deltamarin ropax concept with a membrane tank – the DeltaChallenger – shows “a better car and passenger capacity in relation to size than the reference vessels”.
Further, very recent lattice-type tanks provide new opportunities she says. Developed by KAIST and POSCO the rectangular design is based on a prismatic configuration capable of holding up against high pressures (like C-type tanks) by virtue of its lattice-like inner reinforcing structure.
Ms Savijoki says it’s a pragmatic development as “cylindrical tanks often need to fit it into a box-like area which makes space utilisation inefficient and means fitting two tanks to get to the right LNG volume”. She adds that since tanks are expensive, if these lattice designs could reducing the number installed then it will both drop costs and result in better use of the available room.
New approach
Despite all this, there is actually more freedom to determine where you can put LNG tanks than maybe initially assumed says Alex Vredeveldt of TNO.
He explains the old B/5 rule – which says that ships using gases or other low flashpoint fuels need a minimum distance between the fuel tanks and the ship’s shell of 1/5 of the ships beam – can be relaxed if you can demonstrate ‘equivalent safety’.
Therefore TNO put itself to the task of finding out what actually happens to C-type LNG tanks in a collision in order to base its calculations on a full risk analysis.
However when simulations suggested these tanks, even under cryogenic conditions, wouldn’t crack it invited some scepticism: therefore when TNO took 20L tanks into live tests “bets were placed” said Mr Vredeveldt. Despite initial disbelief the high grade 304 or 316 stainless steel tanks didn’t break even when full of liquid – in fact this even seemed to help absorb the shock. Having verified the simulation, it was used to examine impacts to bigger, 35m³ tanks – and there were some interesting results: it showed that rather than rupture, tanks tended to deform even under the kind of force that could come from collision with an inland waterway barge.
According to Mr Vredeveldt this means “we can demonstrate explicit safety rather than hiding behind the prescriptive regulations which are still often used in ship design”.
