The recent fall of natural gas prices to extraordinary lows attests to a misunderstanding of how precarious supply remains, and it has poised the market for potentially explosive price increases in the near term.
Years of waning interest in natural gas simplified the public discussion of gas prices to a single factor: gas in storage compared to five-year average. This spring and summer as storage trended up toward, but still below the five-year average, prices swooned to levels only touched briefly since the early 2000s. The idea of gas storage as the keystone indicator worked in a previous incarnation of the gas market, but it fails to understand the market’s ongoing metamorphosis.
When natural gas was last fashionable, production hovered just over 60 Bcf/d for years on end. Today the supply is growing past 108 Bcf/d. Remarkably, prices trended downward during the transition, even as demand grew as ravenously as 8% per year. Supply stepped up to match demand in an elegant choreography of production, pipelines, and plants with only minor missteps along the way, sparing the market of shortages for many years.
If demand were as stable as in those times past, then the current change in storage trajectory could portend a glut. However, neither storage nor demand is what it used to be. While the seasonal swings in gas demand for heating have not grown, the overall demand has. Storage crucially if less visibly also buffers the multi-year choreography of growth. Storage volumes, though, have not grown with production volumes.
The standard frame of reference —five year range of storage—has been deformed by the extreme demands of 2014’s “polar vortex” and the chronic gas shortage, obscuring reality. For a better frame of reference, the chart below shows nearly 20 years of volumes in storage normalized to dry gas production. It shows how for the better part of the last three years storage trended down as demand chronically exceeded supply. On a longer time frame, normalized storage volumes have trended downward for nearly 10 years as storage did not expand with production. These overlaying trends caused last heating season to end with only 12 days of supply in storage.
During the few times that normalized storage previously reached the nadir seen this spring, the drop caused headline-grabbing leaps in price. This time prices hardly moved. When the depletion trend finally reversed, the market did not see the change as partial abatement of a critical shortage but instead as evidence of a glut. Even with months of higher-than-average injections, last week’s storage ranks as the second lowest in two decades: 31 days of storage compared to a normal average of 47 days. Hardly a bearish signal.
What is more, the market seems not to understand that gas demand will continue to grow aggressively this year and into next. Seven large trains and two clusters of smaller trains of LNG export totaling 5.3 Bcf/d of new demand were scheduled to come online between the first of 2019 and mid-2020. In addition EIA projects generally long-lead industrial demand to grow by 2 Bcf/d in 2019 and into 2020 and forecasts additional growth of NGL fractionation and export.
Last spring the final constraints on Appalachian supply dissipated just in time and for a planned push in demand growth from LNG, but nearly 2 Bcf/d of export capacity planned for commissioning early in the year was delayed. The delay sent the available gas into storage, but that excess supply was always only temporary. Indeed most of the delayed projects are now ramping up. Demand from all of various new industrial and export projects should be inelastic. Even if completed late, the projects will demand gas even if spot prices rise.
To supply needed gas in the near term some modest volume will be recaptured in the Permian as flaring turns to gathering, but mostly the new lines will redirect and not replenish supply. Gas DUCs, especially in Appalachia, have now fallen to the lowest number since the EIA began counting, and new well starts, which exhibit a four- to six-month delay to first production, have been falling since April. Even an eventual increased rig count may provide limited relief since Haynesville pipelines are already full and since unused capacity from Appalachia is limited.
Gas prices reacted irrationally this spring and summer. Prices dipped to some of the lowest figures in twenty years at the same time that storage tightened to near its leanest in the same twenty years even while facing stiff near-term demand growth. Instead of a price increase due to a chronic shortage during breathless demand growth, prices fell on a temporary relief from its paper-thin margins.
The market’s misunderstanding of the signals have now set up the potential for an explosive price increase. Demand growth will arrive about the time supply contracts, and the briefly-increased gas storage may be insufficient to negotiate the swing, especially if winter also arrives with its own aggressive demands.