North American LNG: Ready for Lift Off
October 2017
The first wave of major US LNG projects are already hitting global markets, and projects in the pipeline could lift the US export capacity to 80 – 110Mtpa by the mid-2020s. Projects already due for completion by 2019 will easily lift the US to the number three place and just behind major LNG exporting nations Australia and Qatar.

The impact of Henry Hub 

US based LNG exporters based in the US’ southern Gulf region will be exporting gas sourced at Henry Hub prices, effectively setting a transparent mechanism for a ceiling on natural gas. The current pricing forecast, over the mid-term for Henry Hub is at US$3.00/Mbtu, and this contrasts with the traditional oil-linked pricing formulas that have guided the global LNG sector since its inception for over four decades. 

But the application of Henry Hub as a basis for feed-gas costs has a greater impact than a mere pricing mechanism. Because this is fundamentally a market based gas sourcing model, Henry Hub implies an effective de-coupling of the LNG trading facility with individual, upstream field operators. Critically, and unlike almost all other global LNG business models, the US model should allow traders to acquire the gas upstream of the LNG facility. The LNG terminal can then perform the liquefaction and loading service in exchange for a tolling fee, while the traders continue to own and commercialise the gas both upstream and downstream of liquefaction. As such, commercialisation of each of the facilities in many cases is a “take or pay” of the use of the liquefaction and ship-loading capacity, not of the gas itself.   

This links not only with the theme of “portfolio marketing” of gas, which is being increasingly taken up by customers of most major oil and gas companies who command extensive LNG project portfolios, but by traders and potentially speculators who can liquefy and sell-off via global arbitrage, their hedged positions over natural gas.  

While the Henry Hub priced gas is mostly low-cost methane from unconventional well-fields in the Eagle Ford shale and the Permian Basin, AME expects that the facilities themselves have by now secured hedged, long term, back-to-back contracts with upstream suppliers to offset any potential supply-side shocks. 

Flow Reversals 

As recently as 2011, proposals on the table included LNG import terminals to supply the US domestic market. Fast forwarding to 2017 and these former terminals are redundant as the US has achieved self-sufficiency and is becoming a net exporter of gas. The first wave of LNG export terminals kicked off in early 2016 with the first 4.5Mtpa train at Sabine Pass. Sabine, run by Cheniere, will soon have four trains in full operation. This will be followed in 2018 by Sempra Energy’s Cameron LNG, Dominion’s Cove Point, the Elba Island terminal and Freeport LNG. Except for Cove Point and Elba Island, all of the terminals are based along the southern Gulf region of the US. Cove Point is in Maryland and will use gas from the Marcellus Shale, while Elba Island, located in Georgia and with direct access to the Atlantic Ocean, is expected to be sourced from Henry Hub gas. 

Without exception, all are constructed on the sites of former LNG import terminals, and take advantage of pre-existing infrastructures to facilitate the regulatory hurdles and reduce start-up costs. This reversal of flows allows the low-cost reuse of existing asset. This also includes Cheniere’s other project, Corpus Christi, which is expected to start operations from its first, 4.5Mtpa train in 2019. Although previously approved as an import terminal, Corpus Christi still represents a grass-roots site and is a departure from the theme of flow reversals. At Corpus Christi, the prior-approval was likely to have eased the way for re-permitting as an export facility. Elsewhere, the use of existing infrastructures is a major underpinning for the other projects as it includes re-usable pipelines, LNG storage and ship-loading assets. 

Sabine Pass currently has four trains constructed for a total capacity of 18Mtpa, with the fourth and most recent set for an imminent start-up and the fifth due to be commissioned by 2019, increasing capacity to a giant 22.5Mtpa. Train 6 is still in the feasibility stage and needs firm contracts before it can proceed as an investment. Cheniere, via Sabine Pass, has already penetrated markets globally, making deliveries in Latin and South America to destinations in Argentina, Mexico, Uruguay, and Brazil, and further into Spain, Portugal, China, India and Japan. But it already faces stiff competition from Russian Gas into Europe, who can deliver natural gas into Spain at around US$2/Mbtu less than its U.S. counterpart. 

The 5.25Mtpa Cove Point, owned by Dominion, was originally built as an import terminal to supply peak winter demand in the New England markets. It is also due for start-up within months and is well located to service markets in Western Europe. Further to the South, the 2.5Mtpa Elba Island LNG, is being built around a modular concept and will take delivery from Shell of pre-fabricated plant modules. Although it is no longer an owner, Shell has underpinned the entire capacity of the plant for its global markets portfolio.  

Switching to the Gulf of Mexico region, the three-train Cameron LNG is set to become a major gas exporter at around 15Mtpa of capacity, delayed to 2019 for Train 1 but with a fourth train expansion already in the advanced planning stage. Freeport LNG’s first 4.64Mtpa train is planned for a 2018 start up and has pre-sold its much of its base-load capacity to Japanese utilities. Meanwhile, Corpus Christi, due for a 2019 start, is expected to follow on the business model already established at Sabine Pass by owner Cheniere. Originally planned for at least five trains, Corpus Christi should have three trains producing around 13.5Mtpa by the early 2020s, while trains 4 and 5 are still at the commercialisation stage and not yet confirmed. Depending on project completions, AME expects that this first wave of US LNG projects will represent more than 64Mtpa of baseload capacity within three years.