Could US data centers and AI shake up the global LNG market?

Shale gas has revolutionised the US gas industry over the last 15 years. Huge volumes of low-cost upstream gas have helped push coal out of the power generation mix and underpinned the burgeoning global trade in US LNG exports. The sheer scale of the resource had forged our view that Henry Hub prices would stay cheap as chips for the foreseeable future, underpinning US LNG’s competitiveness.

However, changing dynamics in US power and gas markets have led us to take a more bullish view of how gas prices develop. In the past two years, we have raised our forecast for gas demand from the power sector to reflect the difficulty the US faces in achieving its very challenging renewables build-out targets. Additionally, the spree of LNG contracting activity following the Russian invasion of Ukraine has led us to increase our forecasts for LNG exports, easily the biggest growth segment for North American gas demand.

The second big upgrade came in our North America gas strategic planning outlook. It takes into account the explosive growth in data centers and AI that’s unfolding, along with the reshoring of power-intensive industries such as chip manufacturing.

Today’s data centers are power hungry. They already account for around 5% of US electricity demand, needing 2 bcfd of gas, assuming 60% is met by gas generation. AI recent developments provide fresh momentum and have led us to increase electricity requirements by an additional 5% through to 2050.

As a consequence, we now expect total US gas demand to increase by 30 bcfd (300 bcm) by the early 2040s compared with 13 bcfd (130 bcm) previously.

There is, of course, plenty of gas supply to accommodate more demand in North America - the issue will be at what cost. The incremental gas demand will lead to the earlier exhaustion of low-cost resources, both from liquids-associated plays and the dry gas basins of the US Lower 48. Additionally, our view is that associated gas from tight oil will start declining towards the end of the next decade with even the Permian hitting the wall in the mid-2040s.

As a result, Henry Hub prices are likely to experience further upward pressure from the mid-2030s as higher cost plays are tapped, from the Marcellus in the Northeast to supply demand centers along the Atlantic seaboard.

Wood Mackenzie’s ‘old view’ was that Henry Hub would be range-bound between US$3/mcf and US$4/mcf (in real terms) through 2045. We’re now expecting prices to hit US$4/mcf more than a decade earlier. Beyond 2035, our modelling suggests the market gets progressively tighter, lifting Henry Hub closer to US$6/Mcf (in real terms) through the 2040s. That’s up to 45% above our forecasts of two years ago.

This is certainly good news for upstream producers who have experienced years of modest prices in a well-supplied market. But what are the implications for LNG markets? First and foremost, there remains a huge global appetite for LNG. We forecast 230 Mtpa of new supply will be needed to meet demand growth to 2050, adding 30% more volumes to capacity already onstream or under development. The US will have a big part to play in delivering that supply. While price is critically important, US LNG has other advantages including flexibility for buyers on cargo destination, lower cost cargo cancellation options and swift project build times that conventional projects can’t match.

Most buyers, including those in price-sensitive markets newer to LNG, will continue buying LNG on 15-to-20-year contracts and look at the economics over the life of the deal. For most of that contract exposure, Henry Hub-indexed LNG looks a pretty safe bet. The risk, based on WoodMac forecasts, is skewed towards the back end of a long-term contract.

Asian buyers will, as ever, compare Henry Hub-linked contract prices with oil-indexed contracts from conventional LNG projects. But should Henry Hub prices drift upwards in the longer term, oil-indexed contracts from conventional projects will become a more attractive option, assuming the same Brent price or lower, and the same indexation. Over time, though, that arbitrage would close as higher Henry Hub prices would provide headroom for sellers to increase oil-indexation levels.

It’s also worth bearing in mind that buyers lament the lack of new, competitive sources of supply other than the US and Qatar. One potential positive that could emerge is that higher Henry Hub prices spur investment in new conventional sources of LNG outside these two countries.