- January 15, 2021
- Posted by: Art Berman
- Category: The Petroleum Truth Report
U.S. oil production has fallen more than 2 million barrels per day since March 2020. It will fall much lower.
Output has fallen from almost 13 mmb/d in late 2019 to below 10.5 mmb/d in October 2020 (Figure 1). EIA forecasts an increase in November to 11.0 mmb/d and then an average level of about 11.1 mmb/d for the rest of 2021.
EIA Forecast is Impossible
EIA’s forecast is impossible. It does not account for the low level of drilling and for the high decline rates of U.S. wells. It seems more likely that production will drop by at least another million barrels per day below October’s level later in 2021.
A major problem with EIA’s model is that it assumes a 2 month lag between well start and first production. Data shows that for 2019, the production-weighted average from well start to first oil production was about 4 1/2 months (Table 1). It further shows that the average lag from first oil to offset of legacy production decline is about 7 months. That means that future production for the next six to twelve months is locked into low rig counts. No matter how radical, an increase in drilling will not result in higher production until much later in 2021.
It is reasonable to look back to the last time that oil prices and rig counts collapsed. When that happened in 2014-2015, U.S. oil production fell -1.1 mmb/d (-11.6%) (Figure 2). Production in 2020 has fallen more than twice as much even after output recovered from the shut-ins during April and May.
That is because decline rates have changed. In 2014, U.S. decline rates were about 22% annually. Today, decline rates have increased to 43% (Figure 3). Also, the production base is 45% higher than in 2014 so more wells are required to maintain output.
Some expect production to follow a trajectory similar to EIA’s forecast in Figure 1 but at a level somewhat lower than an average 2021 level of 11 mmb/d. That is improbable because reduced rig count has not yet affected production. Most of the decline so far is because of the large number of abandoned or suspended wells.
What About DUCs?
Many reasonably expect that DUCs (drilled uncompleted wells) provide a solution to the lag between drilling and production. There are, after all, about 5,800 DUCs in the main U.S. tight oil plays. These are already drilled and could be converted into producing wells for the cost of completion which is about half the total well cost.
Most DUCs, however, are uncompleted for a reason namely, that their owners don’t believe that their performance will be as good as wells that they chose to complete instead.
Even assuming similar performance, the larger problem is that large numbers of DUCs are already being completed and official EIA 914 production remains less than 10.5 mmb/d. North Dakota publishes monthly data on DUCs that can be compared with active, producing wells.
DUCs currently account for about 35% of new Bakken producing wells and about 25% of completed wells since March have been DUCs (Figure 4). During the 2015-2016 oil price and production collapse, DUCs in the Bakken reached about 40% of completions. It is, therefore, reasonable to expect that current DUC levels may be close to a maximum. Whether Bakken data applies to other plays is, of course, unknown.
More importantly, there are just too few wells being completed to expect U.S. production to maintain 11 mmb/d in 2021.
Five key regions of the United States—Texas, North Dakota, New Mexico, Oklahoma and the offshore Gulf of Mexico—account for 80% of total output. Figure 5 shows incremental new wells and new production for those regions from 2014 through July 2020 (12-month average). At least 400 new wells must be added each month to offset declining legacy production and maintain 11 mmb/d for the U.S. Instead of adding new wells, fewer wells were drilled in each successive month after March 2020. Not surprisingly, incremental monthly production has been falling and that is completely consistent with the declining overall production levels shown in Figure 1.
It doesn’t matter whether wells are newly drilled and completed or DUCs—there are simply too few wells being added to maintain present levels of production.
How Far Will Production Fall?
The good news is that well completions and rig counts have turned around and are now heading in the right direction. The bad news is that it will take many months before drilling and production equilibrate. How far will production fall?
The truth is that no one knows. Oil production is part of a complex system. Its interdependencies and feedback loops make it dynamic and adaptive. There are unresolvable uncertainties. The best approach is to identify and describe the key patterns that characterize present state: rig count, decline rates, lags and leads, completions and incremental production rates. These offer the most probable but only notional projections of those trends.
In Figure 6, I show three scenarios based on rig count and EIA’s production forecast for 2021. These should be viewed as trend lines rather than forecasts.
In the base case, output begins to decline in April 2021 and decreases to 9.1 mmb/d by September 2021. In the low case the production minimum is estimated at 7.8 mmb/d and in the high case, 9.9 mmb/d.
Most decision analysis is based on choices between favorable prospects. In the present exercise, the assumption is that the future state of oil production depends on new drilling and new completions. An increase in the rate of well abandonments add another element of complexity and uncertainty.
The number of active U.S. wells has fallen by more than 34,000 (-12%) since August 2019 (Figure 7). Most of that is from wells drilled before 2015. Those wells account for about 25% of current production.
Older wells were abandoned during the last price collapse but the scale and degree were smaller. From April 2015 to March 2016, active wells fell by 8,800 (-3%).
Much of what is happening today in oil markets began in mid-2018 when investors began began their flight from oil companies. The price collapse in late 2018 was a strong signal to producers that was largely forgotten when prices quickly recovered.
The oil-price collapse of 2018 should have sent a clear message to producers to change their behavior or risk further crushing price reactions going forward.
—2018 Oil Price Collapse: More Than a Correction (March 2019)
The oil-directed rig count fell 15% from 1,006 in February 2019 to 846 in February 2020 but that was not enough to regain trust with investors. Covid forced the oil-directed rig count to 178 in August and nothing can be done from that level to change falling production. Completing a drilled, uncompleted well is a considerable capital expense and producers simply don’t have the money to complete enough of them to make a difference.
Whatever the magnitude of production decline or its precise timing, it is important to recognize what is coming. The lower-for-longer ruling paradigm has been accurate and useful since the oil-price collapse in 2014. What is happening now is different.
It is unlikely that the tight oil business will recover from the effect of Covid-19 and lower oil prices. Markets will continue to send higher price signals until rig counts recover to the 800 or so rigs needed to support EIA’s 11 mmb/d forecast.
The public and many investors have the peculiar belief that the world will be just fine without oil. The world will be fine. It has survived meteor impacts and mass extinctions but humans are more fragile. Higher oil prices are the last thing the global economy needs right now.
Art, I couldn’t agree more. Commodities are rising and oil price is set to rise, in the midst of a global economic crisis. A perfect storm is brewing and no amount of money printing can fix that. If things take a turn for the worst the economic crisis could be followed by a monetary crisis. Energy per capita and standard of living are going down for the majority no matter what. That could easily add a social crisis whose first signs we are all seeing. Peter Turchin predicted the increase in social instability 10 years ago in Nature Vol 46, 4 February 2010, pg 608. The pandemic was just a catalyst for what was already brewing. We are living in interesting times.
Art, many thanks for your evidence-based analysis. Indeed, humans (and the food and clean water supplies we depend upon to sustain us) are more fragile than planet Earth.
The Laws of Physics and the limitations of chemistry and biology are non-negotiable.
Climate scientist Zeke Hausfather tweeted earlier this month (that also included a temperature graph from 1850 to 2020):
“Based on the latest ERA5 data, 2020 was the warmest on record over land by a fair margin, tied for the warmest for the global as a whole, but only the 3rd warmest over the oceans.
The world’s land (where we all live) was 1.94C (3.5F) warmer than the preindustrial period in 2020.”
The latest climate science indicates humanity will inevitably see an increasingly hotter and more hostile planet Earth for humans (and the increasing endangerment of our food and clean water supplies) in the coming few decades. A +1.5 °C global mean temperature rise (above Holocene Epoch pre-industrial age) is already ‘locked-in’ and likely reached before 2030. Many climate scientists think global mean warming will likely surpass +2 °C well before the end of this century, due to greenhouse gases (GHGs) already present in the atmosphere.
See report “Climate Reality Check 2020”: https://www.climaterealitycheck.net/
Either humanity finds and deploys effective and timely ways to rapidly reduce human-induced GHG emissions from now on, including drastically reducing fossil fuel (coal, gas and oil) dependency, or we risk +3 °C or higher global mean warming later this century, forcing billions of people to migrate away from progressively more uninhabitable equatorial and mid-latitude regions, drastically reducing arable land, fisheries and clean water supplies, and likely inducing civilisation collapse.
Accelerating global sea level rise will also progressively inundate many highly populated coastal regions in the coming decades. A recent scientific paper suggests:
“The lessons from the paleo record inform us that it is possible, when pushed by greenhouse gases, for the climate to change rapidly and for ice sheets to drive several meters of global sea-level rise over a century timescale.”
Professor Hans Joachim Schellnhuber, founder, Potsdam Institute for Climate Impact Research, Germany, has said:
“If we don’t solve the climate crisis, we can forget about the rest.”
Indeed, we are living in interesting times!
[…] costs, but also because oil shortages figure large in our near future. As oil geologist Art Berman […]
Smart analysis. The recovery will be very onerous because the reversal of a total shutdown presents a series of more demanding activities and costs, which large producers will also be able to face but the small ones do not think they can do it.
Thank you for the content. Your graphs are excellent and I appreciate the historical data. Predicting the future is always tricky. I read a lot of oil and gas articles and the ones that I find most enjoyable are the ones that say the EIA forecast is wildly off, especially when the EIA has overestimated by 1 million barrels per day of production or more. Sometimes EIA’s forecasts are wildly off, like the start of 2020 when they predicted 13.3 million b/d in their STEO forecast for 2020 but who could have predicted COVID? In 2018, their Jan STEO forecasted 9.9 million b/d and shale blew the number out to 11 million b/d, another 1.1 million b/d miss. If your thesis is right though shouldn’t the EIA’s predictions be wrong in the past, especially in the declining rig period from 2014 to 2016? Their records shows that their Jan STEO predictions were right more often than they were wrong and if anything, they underestimated production. Here’s the data.
Jan STEO latest
2014 8.5 8.8
2015 9.3 9.4
2016 8.7 8.9
2017 9 9.4
2018 9.9 11
2019 12.1 12.2
2020 13.3 11.3