LONDON, Nov 26 (Reuters) - Plentiful and cheap supplies of natural gas, coupled with near-record prices for diesel and gasoline, provide a seemingly ideal environment for projects that convert natural gas into liquid transportation fuels. Yet most gas producers have hesitated to commit to new projects.
Gas-to-liquids (GTL) beats other options like pipelines, liquefied natural gas (LNG) and compressed natural gas (CNG) for smaller gas deposits stranded thousands of kilometres from consuming markets.
GTL plants are the most attractive way to realise the value in natural gas when oil prices are above $60 per barrel and gas prices are below $8 per million British thermal units, according to the 2012 "Global Energy Assessment", a landmark study compiled by the International Institute for Applied Systems Analysis.
Four commercial GTL plants have become operational over the last 20 years: Mossel Bay in South Africa (1992), Bintulu in Malaysia (1993), Oryx at in Qatar (2007) and Pearl also in Qatar (2011-14).
Combined capacity will be almost 225,000 barrels per day once Shell's Pearl project is fully operational (mostly diesel with smaller quantities of naphtha, lubricants and waxes).
South Africa's Sasol (Mossel Bay and Oryx) and Royal Dutch Shell (Bintulu and Pearl) dominate the market, though other companies including BP, Chevron and Exxon have proposed projects or built demonstration facilities over the last decade.
Shell is currently expanding Pearl and Sasol is working with Chevron and the Nigerian National Petroleum Corporation to finish a GTL plant on the Escravos River in Nigeria.
Escravos, which is a copy of Oryx, will convert about 8.5 million cubic metres per day of gas, currently flared, into 22,000 barrels per day of diesel and 11,000 barrels of naphtha.
Both Shell and Sasol are reportedly considering projects on the coast of Louisiana in the United States. Sasol has plans to expand its production in South Africa and to build new plants in Uzbekistan and Canada. But even the two market leaders have hesitated to make firm commitments to new projects.
Other major oil and gas producers have either avoided GTL altogether, or abandoned projects early. Exxon Mobil, Conoco and Trinidad have all cancelled projects in the last decade. BP produced 300 barrels per day of synthetic crude oil at a test facility between 2002 and 2009 but has not chosen to build a full-scale plant.
The case for a big expansion of GTL production appears compelling. Nearly 40 percent of the world's natural gas reserves are too far from consuming centres to be economically delivered via pipelines: GTL, LNG or CNG may be the only way to exploit them.
GTL plants can unlock value by converting plentiful and cheap gas into scarce and valuable liquids. GTL yields lots of diesel (70 percent) rather than gasoline (25 percent), which could help bridge the worldwide shortage of middle distillates.
GTL burns cleanly (it contains almost zero sulphur) and engine performance is better than for diesel from a conventional refinery (GTL diesel has a cetane number of 70 compared with 40 for refinery output). It also avoids the expense of fitting and operating energy-intensive hydro-desulphurisation units to conventional refineries.
Together with LNG, GTL is the only way to reduce wasteful venting and flaring of stranded natural gas from oil wells in countries like Nigeria, Russia, Iran and Iraq.
GTL plants are currently about 60-65 percent energy efficient. Their carbon efficiency is 75-80 percent. But projected improvements over the next decade could push energy efficiency over 70 percent, and carbon efficiency to 90 percent, comparable to conventional refining.
Escalating costs have done much to undermine GTL. The technology is relatively mature. But many developers have opted for very large-scale plants in a bid to maximise economies of scale.
GTL plants have had to compete for scarce engineering talent and raw materials such as high-strength corrosion resistant steels with the enormous number of LNG projects launched over the last ten years.
The result has been enormous pressure on the relatively scarce engineering companies and suppliers capable of delivering advanced GTL plants on this sort of scale.
In much-cited 2005 report, Professor Michael Economides of the University of Houston, put capital costs for a 65,000 barrel per day GTL plant at about $25,000 per barrel per day ("The Economics of Gas to Liquids Compared to Liquefied Natural Gas").
But following cost overruns, Shell's Pearl project ended up costing $18 billion, or about $110,000 per barrel per day for a 140,000 barrel per day plant. The Sasol-Chevron Escravos project has seen even worse cost inflation, and will likely end up costing at least $180,000 per barrel per day for 33,000 barrel per day of output.
"Construction delays are chronic. Costs escalate as the giant projects create their own economic weather for engineering, labour, steel, shipping and other services," explains Arctic Gas, which coordinates gas transport projects in Alaska for the U.S. federal government ("Can gas to liquids technology get traction?" June 2012).
Escalating costs are only one aspect of a wider issue. "The challenge for gas to liquid technology is not high cost, it is high risk," according to the U.S. Department of Energy's Advanced Research Projects Agency (ARPA-E).
GTL projects have to cope with multiple risks simultaneously: price changes in both the feedstock and product markets, construction costs, cost of catalyst, and how big a plant to build.
ARPA-E found the biggest risk was not the cost of gas, or even construction, but the realised price for the products when the project comes onstream (which explains why Pearl has been successful despite being wildly over-budget).
It takes so long to develop a large-scale GTL plant there is no guarantee the configuration of gas and diesel prices which underpinned the original investment decision will still prevail when it eventually enters into service.
"Progress has been stymied by the astronomical cost of building GTL plants, and the career risk facing chief executives who undertake such investments, as well as by oil prices that have swung between rock bottom and sky high," Arctic Gas observed.
ARPA-E highlights the tremendous swing in oil prices from under $40 per barrel to more than $110 between 2000 and 2011, the typical timeframe for developing and building a GTL project. Long delays and high costs give rise to enormous market-timing risks ("Gas to liquid technology" February 2012).
The solution may be to think smaller. Process engineers generally prefer bigger plants because costs go up in line with the surface area while revenues go up faster in line with volume. The Sasol-Chevron GTL reactor vessel is enormous (see picture here:).
In Shell's two 1200-tonne reactor vessels at Pearl "the surface of the cobalt catalyst is so vast that if it were spread out horizontally it would encompass an area almost 18 times greater than Qatar itself," Arctic Gas says.
Smaller scale plants would be quicker to build, more responsive to changing market conditions, and a lot less risky. Building lots of small plants rather than a few big ones would also maximise the opportunity to reduce costs through learning effects.
WorldGTL and CompactGTL are already developing small-scale modular systems for offshore platforms and associated gas production. Petrobras is piloting a micro-reactor based GTL system developed by CompactGTL for some of its offshore oil installations.
If GTL is ever to contribute a significant fraction of diesel demand, developers will have to be convinced the divergence of gas and oil prices will be sustained in the long-term, and find ways to manage project risk more effectively.
In the meantime, the brightest outlook may be for smaller-scale plants that can be developed more quickly with less risk.