DBO - EIA Short-Term Energy Outlook And Tight Oil Update November 2023
2023-11-14 03:01:00 ET
Summary
- The EIA’s Short-Term Energy Outlook was revised lower in November compared to September.
- A shortage of oil, high oil prices, and the transition to electric transport are expected to lead to a fall in oil demand by 2029.
- About 70 thousand horizontal oil wells are completed by the end of 2029 in the Permian Basin and then the completion rate rapidly decreases as tier 1 and 2 drilling locations become scarce.
A guest post by D Coyne
The EIA’s Short-Term Energy Outlook ((STEO)) was published in early November. The chart below estimates World C+C by using the STEO forecast combined with past data from the EIA on World Output.
The EIA’s Short-Term Energy Outlook ((STEO)) was revised lower in November compared to September (we skipped the October STEO). World C+C output is expected to decrease in the 2nd and 3rd quarters of 2023 and then increase over the next 5 quarters. Annual average World C+C output increases by about 1028 kb/d Mb/d in 2023 to 81804 kb/d and then to 82555 kb/d in 2024, about 445 kb/d below the centered 12-month average peak in 2018. This month’s World C+C estimates are about 400 kb/d lower than September’s estimate for 2023 and 1150 kb/d lower for 2024 due to the revisions in the STEO forecast since Sept 2023.
The chart above assumes World Petroleum Stocks at the end of 2014 were high enough to supply 90 days of the World's average petroleum consumption in 2015. Total petroleum stocks tend to rise over the long term because average petroleum consumption increases and in order to maintain a 90-day stock the absolute level of petroleum stocks must increase. The days of forward supply ((DOS)) on the right vertical axis is the better measure of petroleum stocks because it is very easy to see if the World stock level is above or below the target level of 90 DOS. Based on the EIA’s supply and demand estimates, at the beginning of 2018, 2019, and 2020 petroleum stocks were relatively low (less than 86 days of forward consumption) and at the beginning of 2022, 2023, and 2024 the stocks were at a balanced level (close to 90 days of supply).
OPEC has very different supply and demand estimates for the World than the EIA. This results in the stock levels looking very different, with the OPEC estimates suggesting a balanced oil market at the beginning of 2018, 2019, and 2020 with close to 90 days of supply each year. At the start of 2022 and 2023, the OPEC estimates suggest a significant oversupply with stocks at 97 and 96 days of forward supply in each of those years. High World demand assumptions by OPEC (about 2 Mb/d higher than the EIA estimate) result in a significant shortage by the beginning of 2025 at only 86 days of supply.
The chart above takes the average of the STEO and MOMR supply and demand estimates and uses them to evaluate stock levels based on these averages. By this estimate, the World stock levels look a bit high at the start of 2016 and 2017 and flip to a slight undersupply at the start of 2018, 2019, and 2020. By the start of 2021, there was a significant oversupply (14 days extra), and there was a slight oversupply at the beginning of 2022 and 2023 with DOS at about 92 days. In 2024, the year starts at a balanced 90 days of supply and by the start of 2025, the market is a bit undersupplied at about 87 DOS. Unfortunately, the World does not know if the EIA, OPEC, or IEA estimates for supply and demand are correct, this uncertainty may be one underlying cause of oil market volatility along with war, politics, weather, natural disasters, and unknowns.
The chart above came from an earlier post on the EIA’s IEO 2023, the reasons why I think the DC scenario is a better estimate for World C+C can be found there. This chart is the motivation for a revised shock model with a plateau of around 83 Mb/d, though my expectation is that the plateau will be shorter than the DC Scenario above which assumes the EIA IEO reference scenario for transportation demand for petroleum is correct through 2032. A shortage of oil, high oil prices, and the transition to electric transport are expected to lead to a fall in oil demand by 2029. See the chart near the end of this post.
The chart above from the EIA’s AEO 2023 reference scenario for US tight Oil output suggests a plateau in tight oil output of about 9.1 Mb/d, this is part of the motivation for a revised tight oil scenario, though I expect the plateau will be shorter than the EIA scenario above. I also found an interesting paper on the Novi Labs website at this link (the full paper can be downloaded at the linked page) which suggests about 24000 tight oil well locations left in the Midland Basin, if we assume there are at least this many locations left in the Delaware Basin this suggests around 90 thousand total wells for the Permian Basin (there are about 45 thousand wells currently), with only 65 thousand of these locations if limited to 20k future tier one and tier two locations.
The scenario above for the Permian Basin is my main revision of my tight oil scenario with other basin scenarios unchanged from my previous tight oil scenario ( link here to the previous post ). About 70 thousand horizontal oil wells are completed by the end of 2029 (from Jan 2010 to Dec 2029) in the Permian Basin and then the completion rate rapidly decreases as tier 1 and 2 drilling locations become scarce. The scenario URR is about 33 Gb which is much less than the USGS F95 TRR estimate of 45 Gb (the mean estimate is about 75 Gb), note that for the North Dakota Bakken/Three Forks the eventual URR is likely to be quite close to the USGS F95 TRR estimate, so this scenario may be quite conservative. A link to the spreadsheet with the above Permian scenario is here , the monthly completion rate in row 4 (delta wells) of the output sheet can be changed to create different scenarios as desired.
The new tight oil scenario is above with a peak of about 9060 kb/d, roughly 500 kb/d less than my previous scenario, URR is 63 Gb, about 9 Gb less than previously.
The chart above is in Mb/d (vertical axis) the low scenario has a 63 Gb URR, the high scenario a URR of 72 Gb and the medium scenario is the average of the high and low scenarios.
The chart above is from the EIA’s official tight oil estimate (spreadsheet at this link ) and shows that Permian output has been flat since March 2023 at about 5000 kb/d. Most of the Increase in US tight oil for the past 2.5 years has come from the Permian basin so we tend to focus here when considering US tight oil output.
The PSM Permian estimate in the chart above looks at C+C output in Texas and New Mexico from the EIA’s Petroleum Supply Monthly ((PSM)) and subtracts Permian and Eagle Ford output from the total C+C in those states to estimate conventional output. For the most recent 12 months, the Permian estimate may be low due to incomplete data, so we look at the conventional output over the months from 13 to 24 months before the most recent data point (September 2023) and find that conventional output was relatively flat during that period. We take the average over these months (Aug 2021 to Sept 2022) and assume conventional output remains at that average level over the most recent 12 months. Permian output is then found by deducting conventional and Eagle Ford output from the TX and NM C+C output total to find the PSM Permian estimate.
I also look at Novi Labs Permian output data from the most recent Permian update (post at this link ), this data is also compared with PSM data for TX and NM in the way that the EIA tight oil data was. In this case, the conventional output was relatively flat for the 12 months from March 2022 to Feb 2023. I assume conventional output continues at the Feb 2023 level for March 2023 to Aug 2023 to find the Novi/PSM estimate. The trend of the two estimates is similar at between 612 and 631 kb/d per year and the difference in the estimates is explained by Novi Labs including all Permian formations where the EIA includes only the Wolfcamp, Spraberry, and Bone Spring Formations. Note that my scenario for the Permian is more in line with the PSM Permian estimate, actual Permian output (when including all formations) may be about 200 kb/d higher than my scenario as of August 2023. The bottom line is that Permian output is likely 300 to 500 kb/d higher in August 2023 than the EIA's official tight oil estimate.
Above we have my revised World C+C Shock Model with the recent estimates from the STEO for 2023 and 2024 shown on the chart, the new scenario has a steep decline starting in 2029, which is earlier than my previous scenario (where it was 2033). The peak is 83 Mb/d in 2026 and 2027 and a 4-year plateau at close to this level from 2025 to 2028, URR is 2660 Gb.
The chart above compares the previous Oil Shock Model (the “high” scenario above) with the most recent scenario (“low”), the medium scenario is simply the average of the high and low scenarios. My best guess is that World output will fall between the low and high scenarios, but it is highly unlikely any of these scenarios will be correct.
Editor's Note: The summary bullets for this article were chosen by Seeking Alpha editors.
For further details see:
EIA Short-Term Energy Outlook And Tight Oil Update, November 2023