Statewide shelter-in-place orders, worldwide business shutdowns, market meltdowns, medical calamities. Much of what is going on right now is unprecedented in the modern era, and there are no guideposts to help predict what happens next to the world as we knew it. But in the boom-bust energy sector, it is déjà vu all over again. We have seen steep drops in prices, drilling activity and production enough times to have some idea about how this is likely to play out. Granted, this time around it is particularly bad, but that doesn’t change the sequence of events that we are likely to experience over the coming months and years. Today, we’ll look back at what happens to Shale-Era basins after a price collapse, focusing on the inherent lag between a major reduction in activity level and the inevitable production response.
Anybody who thought the $4.85/bbl increase in crude oil prices on Thursday was anything other than a head fake was kidding themselves. The price of CME/NYMEX Cushing WTI crude oil for April delivery (the last day of trading for the contract) fell by the end of the day to $19.46 before the settlement was posted at $22.43/bbl, down $2.79/bbl. The WTI May contract closed at $22.63, down $3.28/bbl. Forget fundamentals. Crude oil prices are now totally dependent on the next news flash, press release, rumor or stock-price shock. About the only thing we can be sure of is that it is likely to be a long while before we see the sunny side of $50/bbl crude oil again.
As we detailed in Paint It Black on Friday, producers are responding to that harsh reality by cutting their capex budgets even deeper than their initial 2020 plans. Even before last week’s price carnage, the 42 E&Ps that we cover in the RBN blogosphere had announced that their total capital spending would be cut by 26% from previous guidance and a reduction of 36% versus 2019. No doubt we’ll see many more cuts in the coming weeks. But there are a couple of important things to note. First, with only a few minor exceptions, E&Ps are still drilling, collectively planning to spend tens of billions of dollars to bring more oil and gas to the market. Second, these companies are in survival mode. Consequently, those billions are being spent drilling the sweetest of the sweet spots using the most productive drilling and completion equipment out there, so volumes are not falling off nearly as fast as are capex dollars. For example, on March 16, EOG Resources announced a 31% capex reduction but said that it expects 2020 production to be flat compared to last year. Granted that’s down from their previous guidance of 10%-14% growth, but it is not a decline. When all the guidance comes in with first-quarter results, we expect that our universe of companies will come in collectively somewhere around a 6-8% decline in production targets. Down, but not out.
To be sure, crude prices are off sharply, and we can expect cuts in active rigs to come very quickly. On Friday, the Baker Hughes U.S. rig count was down by 20 to 722, with 19 of those being crude-focused rigs, 13 of which were out of the Permian. But even though the rig count will be withering at a record pace, that does not mean that we’ll see a steep decline in production. One reason is DUCs — drilled-but-uncompleted wells — that producers can turn on to generate revenue without the need of a drilling rig. Another is that there are still a lot of rigs out there running. And, as we said just above, the rigs and crews that remain in operation are the cream of the crop, and they are drilling in the absolute best locations.
Of course, eventually a reduced level of drilling will slow production growth, then flatten output and likely even bring it down. But even if crude oil prices stay dirt cheap for a long time, the fall in production will stretch out over months. And at some point, the decline will flatten out. That’s because of the innate production curve — or type curve — for shale wells. Steep in the front. Flat in the back. So wells drilled in the recent past will be on the steep part of the decline, and producers won’t be drilling enough new wells to offset the decline of those existing wells. But over time, the preponderance of wells will reach the flat part of the curve, and the pace of the decline will slow. That’s what we saw in the Eagle Ford and the Bakken back in 2015-16. But an even better example is what happened to Haynesville gas in 2011-14. Haynesville, located in Northwest Louisiana and Northeast Texas, is a gas play, but the way it played out is a classic. In fact, the Haynesville experience makes a great case study in which we can segment the rise, decline and recovery of a shale basin into five distinct stages.
Before we describe those stages, we should first go through a brief tutorial on type curves, focusing on a typical well in the Haynesville. Figure 1 shows the type curve (dark-blue area), which indicates how the volume of gas the well produces changes dramatically over time. The vertical scale on the left is daily production in MMcf/d while the horizontal scale is months since well completion. In the first two years of the well’s productive life (dashed red box), production is declining rapidly. The initial production rate starts off at a very healthy 14 MMcf/d, but after a year or so, the well will be producing “only” about 6 MMcf/d, and after another year its output will be down to ~3.5 MMcf/d and starting to flatten out. But the “estimated ultimate recovery” or EUR curve (purple line, right axis) makes it clear that one-third of the well’s production is recovered in the first two years of production while the type curve is declining rapidly, but the remaining two-thirds of production (dashed light-blue box) happens over the next couple of decades while the curve has a very shallow decline profile. (For this example, we are going to keep it simple and not go into the prospects of recompletion and enhanced recovery techniques.)
The main takeaway is that production from new wells starts high, declines quickly but then flattens out. So the fewer the number of new wells drilled, the greater the proportion of older wells are in the total well population, and those older wells are already out on the flat part of the curve.
Figure 1. Type and EUR Curves for a Typical Haynesville Gas Well. Source: RBN (Click to Enlarge)
With that type-curve information in mind, you can start to wrap your head around the notion that, in a highly productive shale play, a ramp-up in drilling-and-completion activity can result in a relatively quick rise in production, and a ramp-down in drilling will drive a relatively swift change in the short-term trajectory of production growth, but that it takes time for a true decline in production to set in. The Haynesville’s experience over a roughly 10-year period — from 2007 to 2017 — helps to demonstrate that point. The play went through five stages, which are shown graphically in Figure 2 and outlined here:
1. Good Times – 2007 to early 2011. Spurred by Chesapeake Corp.’s discovery of extensive shale-gas reserves in the Haynesville in 2007-08, the rig count in the play (brown line, left axis) took a choppy rise from 110 to 160. During that time, Haynesville gas production (green line, right axis) was increasing fast, eventually skyrocketing from 5 Bcf/d in 2009 to 10.6 Bcf/d in 2011. Nothing like that had ever been seen before. But by 2011, the flood of incremental gas supply was overwhelming the U.S. market and natural gas prices plummeted from the $4.30/MMBtu level in the first half of 2011, sliding over the next 18 months to average $2.80/MMBtu during 2012. The Haynesville was hit hard; production there was strictly dry gas, with very little NGLs to supplement producers’ bottom lines. So the focus of U.S. drilling activity quickly shifted to shale plays that produced either crude oil (like the Bakken); “wet,” NGL- and condensate-rich streams (like the Eagle Ford); or the emerging, highly prolific Marcellus natural gas play, all of which offered higher returns.
2. Rig-Cutting Time – Mid-2011 to mid-2012. Producers abandoned the Haynesville in droves. The working rig count fell from 160 to 40 by mid-2012. But check out what happened to production. As the rig count fell, production continued to rise. And even after the rig count hit rock bottom, Haynesville production was still relatively flat. As we demonstrated above, shale’s decline curve math has a built-in inertia factor. Fewer wells were being drilled, but the wells that were being completed after the drilling cuts started focused on the best locations using technological upgrades, thus were so productive that they offset the drilling slowdown. But there is a limit to how long that can go on.
Figure 2. The Five Stages of Haynesville Natural Gas. Sources: EIA, Baker Hughes, RBN (Click to Enlarge)
3. Production Cuts – Mid-2012 to early 2014. By mid-2012, that limit had been reached, and Haynesville production declined from 10.6 Bcf/d down to 6.5 Bcf/d in 2014, even with only 40 or so rigs running. But then, all those wells that had been drilled two, three or four years before had reached the flat spot in their type curves.
4. The Zombie Phase – Mid-2014 to late 2016. And then production flattened out. The production decline was minimal. A low price hardly makes any difference in this phase. Output from the wells in place is no longer declining rapidly. The wells are just rocking along, generating cash, and incurring only modest operating costs. Most of the capital has been invested years before. No production gets shut in. Producers need the cash to stay alive.
5. The Recovery Phase – Early 2017 on. A whole new slate of players entered the Haynesville fray, primarily private equity-backed management firms (as opposed to the publicly held producers of the early days). When they came in, they did so armed with new methods and technologies that could achieve much higher output per well (see Don’t Call It a Comeback). The rig count rebounded and so did production.
In fact, after 2017, the Haynesville went on to surpass its old 10.6-Bcf/d peak production in early 2019, and the play’s output now tops 12.4 Bcf/d. While natural gas prices remain sub-$2/MMBtu and the LNG export market may face COVID-19-related challenges in the months ahead, the Haynesville ironically stands to benefit from the collapse in crude oil prices. Super-cheap crude is sure to drag down the rig count — and, eventually production — in the Permian and other crude-focused shale plays that produce a lot of associated gas with their oil. When those associated gas volumes come down, natural gas prices could rise, potentially spurring a new round of drilling in East Texas and Northwest Louisiana.
The most relevant lesson from the Haynesville’s roller coaster, though, may be that while the oil price crash will result in a big decline in the rig count in the near term, it will take many months for U.S. crude oil production to drop into a significant decline phase. In fact, it is likely that Permian crude production turns out to have more inertia than did the Haynesville, due to the significant number of legacy producing wells in the basin that have been on the flat part of their type curve for decades. That, combined with Saudi Arabia and Russia’s plans to flood the global market with their crude, suggests that a lingering worldwide glut of oil is on the horizon.
Thanks for posting, Gary. RBN publishes some really good, in-depth analysis. I get their daily emails.
I would take slight exception to the inclusion of 2007 Haynesville data as there is basically none. A few vertical exploratory wells and then a couple of horizontal completions entering 2008. Even 2008 drilling and production data is sparse. The announcement of the play was in late March 2008 and the land rush commenced but few wells were drilled as companies prioritized leasing, forming drilling and production units through state hearings. getting a number of well permits and bringing in the rigs to drill them. 2009 is the first year when there was significant drilling January through December.
5. The Recovery Phase is missing a key element. One that I don't think I've ever seen highlighted in an industry article. Strange. Although the state approved Cross Unit Laterals (CUL) a few years prior, few were drilled in the Zombie Phase, Mid-2015 - late 2016. When drilling increased all Haynesville operators were in CUL drilling mode or in the process of transitioning to it. The increasing length of laterals got the most attention and coverage but it was the least important for increased production per 1000' of lateral. The advent of drilling through unit boundaries did away with the 330' set back requirement of previous "unit" wells. That was another ~80 acres per unit (~160 acres across two) that had not been previously produced because the gas can not be extracted without a frack. Then came the intensive frack designs. Chesapeake was the leader with their Propageddeon well (S/N 249722). Chesapeake pumped 5000# of sand per linear foot of perforated lateral in that well. The combination of vastly increased proppant loading and that extra 160 acres of stimulated rock made the wells in The Recovery Phase a vast improvement in production volume. The long laterals don't necessarily benefit the majority of royalty interests. Along with pad drilling and multi-well completion operations, they did lower the operator's cost to produce an mcf.
That development pattern is now fading and a new phase designed to hold leases and meet take away commitments with a reduced number of new wells is beginning.
Interesting reply Skip, I never knew that part of it. And yes I really like the way RBN addresses oil and gas issues but I have done a lot of right of way pipeline work in addition to Landman work so I am kind of prejudiced.
Gary, I would refer you, and any interested members, to one of my blogs covering a chronology of the early Haynesville Shale play. And I agree that RBN does an especially good job with pipeline related articles.
A good read and a good trip down memory lane! Thanks Skip, 2008 seems like a lifetime ago
(yeah, it's getting harder to be Hopeful About Natural Gas!)