Midway through 2013, the natural gas market seems at a tipping point as it balances the vagaries of seasonal weather with the delicate interplay of supply and demand.

Domestic natural gas production was poised at a potential inflection point as 2013 closed in on the halfway mark.

At the end of May, and for the third time in four months, the widely watched U.S. EIA 914 monthly report on natural gas showed a decline in domestic production. It now appears gas production in the Lower 48 onshore market peaked at 73.78 billion cubic feet (Bcf) per day in November 2012 and was down 1.07 Bcf per day in March 2013, the last month for data.

Although it appears that the long-predicted rollover in U.S. gas production is at hand, such signs may yet prove to be a false positive, traceable to arbitrary shut-ins and pipeline maintenance during the first quarter of 2013.

Consequently, it is still premature to anticipate a revitalization in gas-directed drilling in a market that remains oversupplied by between 1- and 3 Bcf per day, with summer weather off to a mild start.

The domestic dry-gas market demonstrates that there can be too much of a good thing. In this case, production gains associated with technological advances in resource recovery, coupled with rising rig efficiency, suggest reserves remain plentiful and easily accessible. The economics of the gas market reflect that dilemma.

Of 32 tight-formation plays in the U.S., drygas shales occupied eight of the bottom 10 in terms of internal rate of return (IRR), according to a December 2012 Credit Suisse shale-gas research report.

That report, which calculated internal rate of returns using strip prices, found the Haynesville/Bossier shale in northeast Texas and Utica dry-gas window generating IRRs of less than 10%, while the dry-gas Eagle Ford shale, Huron shale, Arkoma Woodford shale and Piceance Basin delivered IRRs of 15% or below. Even the best dry-gas shale play failed to produce an IRR above 24%.

The flip side of that exercise compared abysmal IRRs in dry gas to IRRs in oil and liquids- rich plays, such as the super-rich Marcellus shale, the Eagle Ford condensate, the Wattenberg Niobrara and the Mississippi Lime, all of which generated IRRs of 50% or more. In some ways, operators working those liquidsrich plays get free gas as a byproduct of oil production, further dampening possibilities for a revitalization of dry-gas drilling.

Those metrics also explain why the gas-directed rig count has fallen lower than civil discourse in the U.S. political arena. Since January 2012, the horizontal gas-directed rig count, a proxy for shale-gas drilling, has dropped 54%, to 250 units at the end of May 2013, according to Baker Hughes International, while the vertical gas-directed rig count, a proxy for conventional gas drilling, fell 68%, to just 46 units nationwide. Both numbers illustrate the greater macroeconomic issue in natural gas: too much easily accessible supply and modest demand.

Revenue perspective

A review of revenue potential by commodity class provides further perspective. Hart Energy employs an Industry Revenue Model for the domestic oil and gas industry combining domestic production and wellhead pricing reported by the U.S. Energy Information Administration.

At first glance, the Hart model depicts a robust industry that generated $32 billion in revenue during March 2013—the highest level since August 2008. The theory behind the model is that the volume of revenue flowing into operator bank accounts will produce a directionally similar response in reinvestment, with rig count as a proxy. As a rule of thumb, direction in rig count follows direction in industry revenue flow with a lag of some 90 days.

To standardize, it helps to look at revenue per day. March 2013 revenue per day reached $1.035 billion, according to the Hart Energy model, the second consecutive month of billion- plus revenues from oil and gas production. For comparison, the billion-dollar-plus run in 2013 is the first time the industry has topped $1 billion in consecutive months since the March-August 2008 era, when oil and gas prices peaked at $128 per barrel and $10.82 per thousand cubic feet (Mcf), respectively. Daily revenue peaked at $1.34 billion in July 2008.

But the model also indicates that this is a Charles Dickens market, with oil undergoing the best of times while natural gas recovers from the worst of times. The revenue model shows the natural gas market improving dramatically over the past year on a revenue-per-day basis. In fact, daily revenues for dry-gas production edged above $250 million in March, to the highest level since August 2011, largely on the basis of a recovery in commodity prices, coupled with production growth.

After August 2011, the gas market slid over a cliff, with revenue following gas prices down, although gas production itself continued to rise. Eventually, daily gas revenues bottomed at $123 million in April 2012. Thus, the March 2013 figures represent a doubling from the market bottom in less than 12 months. While that circumstance would normally constitute a positive sign, the industry is not yet out of the woods when it comes to dry-gas investment.

For one, the improvement in dry-gas revenue overshadows a weak natural gas liquids (NGLs) market that continues to fall below $100 million in daily revenue. While NGL revenues are up versus the $74-million trough in July 2012, those revenues remain below the most recent peak of $122 million in January 2012, thanks to a combination of too much supply and too little infrastructure to process and transport NGLs. As if on cue, rig counts in the gassier liquids-rich gas plays, such as the Granite Wash and Cana Woodford, remain stagnant.

The consensus among analysts is that NGL bottlenecks will continue through 2015; hence there is little stimulus for increasing the rig count in predominantly gas-prone plays.

Meanwhile, natural gas in all its forms has been overshadowed by the incredible bull run for oil. Domestically, the industry generated $686 million per day from oil production in March, the highest level to date for the Hart revenue model. In fact, March 2013 marked the third straight month that daily oil revenue flow exceeded the previous all-time high of $659 million set in July 2008, when monthly average crude oil prices peaked at $128.

Oil currently generates two-thirds of industry revenue flow, though natural gas has worked its way back to a 24% share, up from 14% in April 2012 (though well below the 60% share natural gas commanded in December 2008). While the Hart Energy model paints a rosy picture for the domestic oil and gas industry, the benefits accrue mainly to oil-directed investment.

Furthermore, a recovery in the gas-directed rig count will have outsized impacts on domestic production, thanks to improvements in drilling efficiency and rising recovery factors. The clearest illustration can be found in domestic shale gas. It took 18 years for shale-gas production to reach 52 billion cubic meters per year. That milestone occurred in January 2007, according to the Credit Suisse study. It took only two years for shale-gas production to add the next 52 billion cubic meters per year. The industry doubled production once again—adding another 104 billion annually—within two additional years.

Some argue that high decline rates in unconventional gas will soften the oversupply issue. The Haynesville/Bossier shales serve as the poster child for this argument. The rig count in the two shales dropped from 150 in third-quarter 2010 to 25 in first-quarter 2013. Gas production followed suit, according to the U.S. EIA, and was down 2.17 Bcf per day from a peak of 9.1 Bcf per day between November 2011 and March 2013.

However, the Haynesville only tells part of the story. U.S. gas production is undergoing regional transformation. As gas production fell in the Haynesville, it simultaneously rose 4 Bcf per day in Appalachia on a rig count that declined modestly from 115 units in first-quarter 2011 to 74 two years later. The Appalachian experience illustrates how efficiency improvements from pad drilling and better completion technology require fewer rigs to generate lowcost gas production volume growth.

The demand debate

So what gets the industry out of the natural gas dilemma? Since supply is not an issue, salvation lies in greater demand, generally identified as growing use of natural gas for electric power generation, vehicle transportation or for export as liquefied natural gas (LNG).

The timing of when those demand components will create a material impact on the natural gas market remains the subject of much debate. In general, the transportation option will take a decade to manifest, while the LNG export option is still a half-decade out before volume reaches critical mass. That leaves power generation as a short-term solution.

Thanks to the 2012 collapse in natural gas prices, coal rapidly lost share as utilities switched to natural gas. Coal fell from roughly 45% of market share in power generation in 2008 to 35% in 2012, while natural gas grew from a 20% share to 35% during the same time frame. In 2012, fuel switching added 5 Bcf per day in incremental demand.

Unfortunately, the 2013 recovery in natural gas prices has seen power generators switch back to coal with demand during the first quarter of 2013 off 2 Bcf per day versus 2012. And that leaves the natural gas market delicately balanced at a tipping point, subject to the vagaries of seasonal weather, and torn between an increase in demand via fuel substitution as gas prices fall below $3.50 on the one hand, and the specter of too much production, too fast, when prices rise above $4.50, the threshold operators say they need to increase natural gas drilling.