Midscale trains for US exports are gaining ground, writes Ed Reed
The decision to install modestly sized trains at Kinder Morgan’s Elba Island project revolutionised thinking in the US around liquefaction options. With a phased start-up at the project over this year and next, it has already triggered a number of developers to work on similar plans – scaling down trains to get projects to the starting block.
A move to smaller LNG trains seems to fly in the face of conventional liquefaction wisdom. Large trains offer economies of scale, bringing the unit cost of production – essentially the price per tonne or million British thermal units – down. The first LNG plants, such as Algeria’s Arzew, involved trains with 0.5 million tpy of capacity. By the 1980s, trains had scaled up to around 2.5 million tpy, and 3.3 million tpy in the 1990s. Qatar brought trains online in 2010-11 with 7.8 million tpy of capacity, the largest such units in operation in the world.
The US, though, seems to be heading in the opposite direction. While the earlier plans such as Sabine Pass used trains with 4.5 million tpy of capacity, there has been a shift in later plans towards clustering small trains to reach similar quantities.
“It’s a balance between capital expenditure and production capacity. Bigger projects are very sophisticated and highly efficient. Big plants are highly complex to maximise efficiency and LNG production and, therefore, maximise returns,” Gasconsult’s CEO, Bill Howe, told NewsBase Intelligence (NBI) in an interview earlier this year. This focus on a large-scale bespoke plant “may not be sustainable in an era of lower energy prices,” he continued.
Leading this move has been the Elba Island project. This plant is due to bring six trains online this year and another four in 2019, each with capacity of 0.25 million tpy.
While Elba Island is relatively modest, both in train and total size, success at the project has helped drive a shift in construction thinking, paving the way for more companies to consider smaller trains as an easier option. Following Elba Island, a number of other companies are considering their options along similar lines.
The next step
Cheniere Energy, the US’ first LNG mover, has talked of a midscale option for its next step at Corpus Christi. Two trains are under construction and a third is expected to reach final investment decision (FID), perhaps this year. While those three trains follow the Sabine Pass model, at 4.5 million tpy, the company has proposed the next step at the site could involve seven midscale trains.
This would bring total Corpus Christi capacity to 23 million tpy, suggesting the seven trains would provide 9.5 million tpy – or around 1.36 million tpy each.
Tellurian, in which Charif Souki – formerly CEO of Cheniere before being ousted, shortly before first LNG was achieved at Sabine Pass – is involved, is also working on a small train plan. In November 2017, the company said it had signed four fixed price lump sum turnkey (LSTK) contracts with Bechtel, worth US$15.2 billion, for a plant near Lake Charles, Louisiana.
The Driftwood development consists of 20 trains, each producing 1.38 million tpy. The first phase involves eight trains, producing 8 million tpy of LNG, with two storage tanks and a loading berth. The subsequent three phases include four trains each and 5.5 million tpy.
In the November statement, Tellurian’s CEO and president, Meg Gentle, said the engineering, procurement and construction (EPC) cost for the project worked out at US$550 per tonne, describing this as “one of the lowest-cost liquefaction construction projects worldwide”. While the number cited by Gentle is low, this number is likely only to be applicable once all 20 trains have been built – the first phase, given the associated infrastructure needs, will be higher.
A decision to opt for smaller trains reduces the capital intensity, Gasconsult’s Howe said, although it loses some of the efficiency gains. “It’s a case of getting the balance right.”
Securing sufficient offtake contracts for a 20 million tpy plant can be challenging, particularly in the current environment, where pricing is uncertain and buyers are reticent to commit. Typically, operators seek to sell around 70-80% of production before work starts.
“Smaller trains can be matched to offtakers and work can begin as soon as those volumes are contracted. A train can be dedicated to one or two offtakers,” Chart Industries’ director of marketing, Paul Shields, told NBI.
“Based on the economies of scale for baseload facilities, building a one-train 5 million tpy plant is unlikely to be cost-effective, while that’s a clear opportunity to cluster trains,” he continued. The concept of a “sweet spot” for plant size is being redetermined, not least because of Tellurian’s Driftwood plan, which “blew a lot of past ideas out of the water,” Shields said.
Aside from the capital question, small and midscale technology providers must also convince potential buyers about the technological validity of their offerings, in particular over thermal efficiency.
Chart is working on a number of projects, including Tellurian’s Driftwood. In comments to NBI, the service provider noted its brazed aluminium heat exchangers, which it said delivered “improved efficiency and lower installed cost versus other heat exchangers”.
There are concerns around the “scaling” problem, Gasconsult’s Howe noted, but this can be tackled through simplifying the system. “In traditional LNG plants there’s a lot of work that goes into pre-treatment of the gas; CO2 [carbon dioxide] and heavy hydrocarbons, for instance have to be reduced to minimal levels. At Gasconsult, a lot of work has gone into optimising those processes.”
Howe went on to say one advantage behind Gasconsult’s liquefaction process was that it did not use an external refrigerant to chill the gas. “We use the natural gas itself. Compared to mixed refrigerant processes this allows us to reduce the equipment count by a minimum of 16 items, which saves space, weight and cost.”
Chart’s Shields noted that a modular approach cuts the expense of on-site installation, with as much as possible built in a central facility and then shipped to the location.
There are a number of players in the smaller-scale liquefaction business, while the conventional baseload projects are dominated by a small number of technology providers. Operators, therefore, have more scope for getting a better deal.
Operators aiming to cluster trains to achieve export volumes focus on midscale trains, which can be easily repeated. Smaller trains can operate independently and be used to access smaller markets, linking small deposits to regional markets, often to displace diesel use.
Stabilis Energy began producing from its 100,000 gallon per day (60,000 tpy) plant in George West, Texas using a standardised train from Chart, the C100N. This plant can fill two trucks simultaneously, with LNG going to local demand, including oilfield services in the Eagle Ford shale.
Commenting on the George West project, Chart’s Shields described it as “a local merchant supply arrangement. There are a couple more [projects] involving LNG supplies to the Caribbean, one intended to fuel a power station. There’s a trucking model, where LNG can be moved to inland regas facilities, and it also works on a peak shaving model.”
Chart, for instance, has three standardised small-scale train options, up to about 255,000 tpy. For larger projects, midscale trains – up to about 1.5 million tpy – can be built on a modular basis.
Siemens, in a statement in mid-March, noted the commissioning of a micro-scale high-pressure LNG system in Dawson Creek, British Columbia. This facility, owned by Altagas, can produce around 30,000 gallons per day (18,000 tpy), using Siemens’ LNGo-HP system.
The company noted the ability such a plant to produce stranded deposits, in rough terrain or remote regions. Dresser Rand commissioned a small-scale low-pressure liquefaction plant at the Ten Man facility, in Pennsylvania, in 2016, for Frontier Natural Resources.
Small plants with marine access can also be used to supply shipping, with a number of companies – such as Carnival the cruise operator – moving into this. Impetus is likely to increase for LNG as a marine fuel as 2020 nears and ship owners can no longer run high-sulphur fuel oil (HSFO) straight.
Additionally, “milk run” tankers are growing in acceptance, potentially opening the way for short-distance small-volume LNG distribution.
The rise of midscale trains has been driven by a particular set of circumstances, the location, in the US, where substantial engineering ability is based, and fuelled by the problem of the shale gas glut. Furthermore, the changing nature of the LNG market – away from large volume, long-term contracts – has forced developers to take radical approaches.
While large-scale trains have caused some problems in parts of the world with cost overruns, such as in Australia, this has not been the case in the US. Furthermore, related non-liquefaction costs remain constant, such as storage tanks, pipelines and quays.
Midscale trains can come to market quicker and for less money, but the operating margins will be slimmer than on larger units. Tellurian’s Driftwood plan will be a revealing test of this midscale technology drive.
The oil and gas industry is slow to adopt new technologies but it seems that a drive for smaller trains is on the rise. While not every US plan will come to fruition, such an adoption will help drive innovation and bring costs down, making subsequent plans that much easier.