Simulation’s key role in towing of Shell’s FLNG

The Prelude's journey from Korea to Australia spanned 5,900km. Credit: Shell

 

Before it was constructed, Shell knew the towing of its Prelude floating liquefied natural gas (FLNG) facility from Samsung’s Geoje Shipyard in South Korea to Australian waters would be tricky. In times past, shareholders might have written off construction of the 488 m leviathan as too risky before it even began; an operation like this had never been done before.

 

Except, it had. Using modern technology, towmasters could mess-up again and again, often deliberately. Tow lines snapped, tugs broke free, lost power, or collided with the facility causing millions of dollars’ worth of virtual damage.

 

High winds and waves smashed into Prelude’s hull and the channel was blocked making further progress impossible. “We wanted to see what would happen if things went wrong, so we simulated some emergency scenarios,” explained Mark McBride, ships group manager at engineering and ship simulation specialist HR Wallingford. “Simulators allow experienced mariners to try things out that they wouldn’t want to try for the first time in real life and that would include bringing the Prelude FLNG facility out of the shipyard, because you’ve only got one chance to do it right,” he said.

After a huge amount of preparation from everyone involved, over several days, crews ran through various drills in simulators, rehearsing for when they and their vessels would have an asset worth more than USD10 billion in tow. “The idea was to practise the manoeuvres, to simulate the unberthing and then the hooking up to the tugs, and then to carry out a full transit of the channel, and to do that in various tidal and wind conditions so we could identify the limits for safe operation.”

To handle the immense detail needed, each of the Australia Ship Simulation Centre’s six integrated simulators ran across 20 top-end desktop computers, with banks of networked servers for storing the 3D models. Using high definition projectors and screens, the system displayed a total of 350 million pixels of information.

“One of the key issues we had with Prelude was that with such a large structure there was a large sheltering effect on anything that sits in its lee – both in terms of wind sheltering and current sheltering,” McBride explained. “That has quite a big impact and we did a lot of work using CFD [computational fluid dynamics] to allow us to model a sheltering pattern as accurately as we could.”

In the end though, the 27-day tow went off without a hitch. “Shell looked ahead at the forecast for several days so they could pick a day when the wind was going to be at its lowest,” said McBride. “It was all hugely successful. Conditions were perfect on the day and the wind was lower than all the simulations had planned for.”

Simulations are now routine for many companies. Svitzer Marine manager Carsten Nygaard explained how his company had been using them for nearly 20 years.

“Simulations are essential for our company. We use them for training and vetting new tug masters. In other projects we join up with the owners of terminals to assess navigational channels into new terminals and to decide the size and physical dimensions of the tugs that are needed,” said Nygaard. “We can get an idea of the location and decide on the size of tug we need for the specific operation. We’ve actually worked with HR Wallingford quite a lot.”

 

Nygaard discussed Svitzer’s involvement with the new Moín terminal in Limón province, Costa Rica, set to open next year. “At the design phase we were approached to participate in the first simulations, in 2014, before they’d even started building it. We made changes to the design of the channel and turning basin. Everyone could see it wouldn’t work. That’s why it’s good for us to get in at the design stage, because then we can optimise it to our needs and convince the client it’s safer if we do it this way. Then in 2015 we verified the results with engineers, pilots, and client reps.”

 

But even today’s most advanced technology can’t simulate everything. McBride is straightforward about the limits of HR Wallingford’s work. “One of the difficulties is with visual perception. Some mariners – tug masters, pilots, ship masters – have difficulty trying to rationalise looking at a two-dimensional image, trying to see the 3D image in that. Everything seems a little bit closer than in real life. We have ways of helping with this and usually they quickly get used to it.”

 

As the simulations get better over time, McBride explained, HR Wallingford has identified other ways to overcome these little inaccuracies. “I think it will be to do with including more elements in the simulation – small things that a mariner might pick up subconsciously. So, for example, where there’s a strong current flowing there might be a wake showing around a buoy. Of course, nobody might notice if it isn’t there, but the mariner doesn’t subconsciously pick up the fact that there’s a strong current, which could be very important”.

 

It is also difficult to give seafarers the impression they are in a strong, windy environment, said McBride. To do so HR Wallingford generates smoke from the simulated tugs and these plumes react to the wind. The pilot will be used to this kind of visual cue and subconsciously react when the smoke is billowing away from the tug. Similarly, McBride explained that when the pilot increases power to the tug a puff of smoke will be visible, which provides reassurance that the tug master has responded to the command.

“Clearly the simulation is not reality and we understand that,” he added. “We can’t take every single parameter into account. But we model everything on a slightly conservative basis, so the manoeuvring simulation is harder than the real thing. That way, even if there is anything slightly different in the real operation, it will be easier and safer in real life.”

 

Contact Charlie Bartlett at charleswbartlett@gmail.com and follower on twitter @C_W_Bartlett

Share on Facebook
Share on Twitter
Please reload

Recent Posts
Please reload