Peak Oil is You

Donate Bitcoins ;-) or Paypal :-)

Page added on August 5, 2017

Bookmark and Share

Is The EIA Exaggerating U.S. Oil Production?


Oil is back at $50 per barrel, restoring some semblance of confidence in the market. But that is just about as much as we can expect in terms of a rally, according to most analysts, with momentum likely to dissipate from here.

But not everyone agrees. Many are worried that oil prices will crash again next year as OPEC scrambles for an exit strategy, but there is actually a bullish case for oil that is not outlandish.

First, crude oil inventories continue to fall. The EIA just released another week’s worth of data, showing another drawdown in inventories. It was a bit more modest last week – 1.5 million barrels – compared to previous four weeks, but the drawdowns continue. U.S. crude oil inventories are now down more than 50 million barrels from the peak hit in March, with stocks back within the five-year range.

But a larger reason why oil prices could deviate from expectations and actually rise quite a bit is because the market is poetically assuming a lot more oil is set to come online than might actually be the case. As Andy Lipow, president of Lipow Oil Associates, notes in a CNBC column, the market has already factored in further production gains from Libya and Nigeria, a major reason why market pessimism really spread in the month of June. But because that assumption is baked into today’s price, if those two countries – beset with violence and instability – fail to come through, then a lot less oil will reach the market than is generally assumed.

Moreover, the U.S. could also disappoint. As discussed in previous articles, the shale drilling boom is starting to slow. A bottleneck of services is held back the completion of some wells, while many shale companies have actually started to throttle back on their spending and drilling activity. In the past few weeks, a list of companies have announced cuts to spending plans for this year, including Anadarko Petroleum, ConocoPhillips, Whiting Petroleum, Hess Corp., and Pioneer Natural Resources. Those five companies alone have slashed a combined $850 million in spending.

The cuts, coming in response to the recent dip of oil prices into the mid-$40s, will likely translate into much lower production next year. As a result, the projection from the EIA that shale will hit 10 million barrels per day (mb/d) in 2018, or even the more recent downward revision to just 9.9 mb/d, could be hard to reach.

However, there is one other important uncertainty that could point to higher oil prices in the near-term. Not only will future production come in lower than expected, but what if current production is actually lower than we think? What if we are overestimating the shale boom?

The market bases its expectations for U.S. oil on the EIA weekly production surveys, which have shown a steady climb in output right up through July. The most recent data for the week ending on July 28 pegs current U.S. oil production at 9.43 mb/d.

But the thing is, these weekly estimates are not always accurate, and are often subject to revisions – sometimes large revisions – later on. On the other hand, the EIA publishes more accurate monthly figures, but only on a several month lag. The most recent data is for May, which has U.S. oil production at 9.169 mb/d.

The problem is that there is quite a gap between these two sources, with the more recent but less accurate weekly figures reflecting much higher oil production.

For example, while the EIA is now pretty certain the U.S. averaged 9.169 mb/d in May, that wasn’t what they thought at the time. Back in May, the EIA published these weekly estimates:

• May 5: 9.314 mb/d

• May 12: 9.305 mb/d

• May 19: 9.320 mb/d

• May 26: 9.342 mb/d

To take a rough average, that would put May’s monthly production at 9.32 mb/d. In other words, back in May, the EIA was telling everyone that the U.S. was producing 151,000 bpd more than it actually was. This is not to put the blame on the EIA – estimating production in essentially real-time is difficult – but only to say that we only get an accurate picture of what is going on by looking in the rear-view mirror. 

Which brings us back to the present. The EIA is now estimating that the U.S. is producing 9.43 mb/d, a figure that has steadily climbed in the past few months. But that would mean that production accelerated very rapidly since May. In the chart below, notice the jump that would need to take place between the monthly figures through May, and the weekly figures up until now.

This is a long-winded way of saying that U.S. oil production could quite conceivably be much lower than 9.43 mb/d right now. In that context, the slowdown in the rig count and the cuts to spending look even more important. Not only will the U.S. fall short of the projections for the rest of this year and next, but it could miss by quite a lot. If that is the case, then oil prices suddenly do not look like they are necessarily trapped at $50 per barrel.

By Nick Cunningham of

20 Comments on "Is The EIA Exaggerating U.S. Oil Production?"

  1. Makati1 on Sat, 5th Aug 2017 5:59 am 

    ANYONE who believes ANY government numbers deserves the consequences. Nuff said.

  2. twocats on Sat, 5th Aug 2017 7:21 am 

    This is all well and good. But in the latest run towards $50/barrel based on whatever the latest bullshit was (Venezuela sanctions?), tons of companies re-upped their hedges in the $50 range. I bet many already had 1/2 years worth of hedges and have locked in an additional 6 mos. So with probably an average of 8 to 12 months of $50 hedge out there, it’s going to be a drag on prices downward if they ever go above $50 until approximately March 2018. Until then, have fun losing money!

  3. MASTERMIND on Sat, 5th Aug 2017 7:34 am 

    Just add it to the list of other fake data. The soviet union did the same thing before they collapsed.

  4. rockman on Sat, 5th Aug 2017 8:25 am 

    If the EIA is compiling non-adjusted production as reported to the various state regulators then that number is lower then the actual volume. Take Texas for instance. Our producers report exactly what the recover every month. When they file their reports: some companies are late and thus the TRRC number is low. But the TRRC has known this so applies a small adjustment factor. But there is a several month delay in the TRRC publishing its numbers.

    Here is what the EIA says about Texas:

    “The Lone Star State produces the most oil in the U.S. The EIA’s most recent April figures put its total bounty at about 3.71 million barrels of crude per day. But that figure uses the methodology that the EIA is moving away from, and it isn’t perfect.

    The projection starts with data from the Texas Railroad Commission, a state agency which receives reports from producers about how much oil each lease they own has produced. The Texas Railroad Commission figures are messy. Often, the forms aren’t complete or contain errors, and the Railroad Commission will file the problem reports in a pending file that doesn’t get counted.

    That can leave a lot barrels out of the figure. For example, the Texas Railroad Commission reported a preliminary March production of 2.30 million barrels per day, while the EIA gives 3.78 million barrels per day in the same period.

    Over about six months or so, Railroad Commissions figures are adjusted upwards as the paperwork is gradually moved out of the pending file. Eventually, the number ends up within a few percent of the EIA estimate, Long said.

    The Texas Railroad Commission declined to be interviewed on the subject, but a spokeswoman for the group noted it updates its figures as late or corrected reports come in.

    To compensate for the Railroad Commission’s data lag, the EIA looks at how months-old Railroad Commission data has been revised and applies that historic adjustment to the best available data it can get for the present month. The process works well when the past data is similar to present data, but can miss the mark when things move fast enough to create a disconnect between the past and present.

    More at:

  5. shortonoil on Sat, 5th Aug 2017 8:44 am 

    Until one these analysts can state what the economy can afford to pay for a barrel of oil, they might as well be throwing darts! If the price gets too high the economy slows, and demand falls. They want to assume that the economy is not going to react to the price. By assuming that pigs can fly they can put a price on air borne bacon. The economy will react to the price and there is a maximum price that it can afford to pay for petroleum. Anything above that, and the economy slows enough to drive the price back down. We won’t mention the energy considerations involved; they haven’t figured out the supply and demand side of it yet.

  6. boat on Sat, 5th Aug 2017 9:18 am 


    If the world is going to collapse shouldn’t demand be collapsing? Can the hills group explain this.

  7. MASTERMIND on Sat, 5th Aug 2017 9:19 am 


    The economy really can’t function properly on anything more than 25$ a barrel oil.

  8. RD on Sat, 5th Aug 2017 9:58 am 


    April statewide Texas produced 90,251,384 bbls oil and condensate..that is a real number..May est on prelim reports is * 84,591,354 (prelim is obviously low). So in April the number was 3,008,379 bo&bc/day lower than EIA report of 3,700,000..find it hard to imagine that amount will be showing up in august or september..usually the late filings are small producers just getting behind in paperwork..having worked for multiple majors and numerous independents I can assure you it does not take four months to file production..especially now that is done online… it could be off but not 20% after 4 months..ICBW

  9. boat on Sat, 5th Aug 2017 10:29 am 

    Master delusional,

    The world is chugging along fine at $45-50 oil. Hell at that price the number of drilled but not facked wells is over 6000. You remind me of of the world collaspe chicken little group that turned into the boy who cried wolf group. Shouldn’t that money you invested in the market make you happy? Or maybe a 16 year low in unemployment be a sign?
    As trump has made illegals uncomfortable, Texas needs constriction workers, welders etc. Anybody needing an economic jump, come on down.

  10. MASTERMIND on Sat, 5th Aug 2017 10:52 am 

    New Oil discoveries by scientists have been declining since 1965 and last year was the lowest in history -IEA

    International Energy Agency Chief warns of world oil shortages by 2020 as discoveries fall to record lows

    Saudi Aramco CEO believes world oil shortage coming despite U.S. shale boom

    UAE warns of world oil shortages ahead by 2020 due to spending cuts

    HSBC Global Bank warns 80% of the worlds conventional fields are declining and world oil shortages by 2020

    UBS Global Bank warns of Oil Shortage ahead

    The Oil Age may come to an end for a shortage of oil. -Saudi Oil Minister Sheikh Yamani

  11. Outcast_Searcher on Sat, 5th Aug 2017 10:53 am 

    The world was chugging along just dandy, with steady but slow GDP growth with nearly $100 oil from mid 2010 – 2014. And globally, energy efficiency gains continue over time, rather steadily.

    So the idea that the world can’t “afford” oil products based on oil at $50 is absurd, to anyone not biased toward financial catastrophe fantasies.

    People pay what the market demands for oil products, up to the amount they want. They don’t change this behavior until they’re under a lot of perceived financial stress. This behavior was just beginning in mid 2008 when oil prices got above $130 and looked to stay that way for awhile at the time.

    The fact that the first world is generally continuing to see modest economic growth with little sign of inflation acceleration is GREAT news for anyone who is trying to look at the data instead of grind an axe.

  12. peaktard on Sat, 5th Aug 2017 11:00 am 

    @outcast yes i don’t like doom because it’s food for alt-rightard dung beetles.

    even the venerable jay hanson tarded. he predicted war socialism. that’s tupid.

    the war is against extremist tard preachers and dictators like putin. we need siberia because there’s tons of oil there.

    note i don’t want the us to fight alone. we play a supportive role and come in at some point to clean up the mess.

  13. Plantagenet on Sat, 5th Aug 2017 11:00 am 

    The government oil data is at least as reliable as the government economic data.


  14. Anonymous on Sat, 5th Aug 2017 11:26 am 

    The weekly numbers are mostly a model. You have to look at the monthly numbers for real data based on a survey. Revisions to the monthlies are tiny. This has been well written about. It always surprises me how people like you that have been amateur oil production watchers for YEARS still haven’t learned the basics.

    Rockman, the RRC lags are real (glad you finally learned that, you messed it up couple years ago). But they don’t affect EIA estimates which are based on direct producer surveys an IHS data (not RRC if less than 2 years old).

  15. Anonymous on Sat, 5th Aug 2017 11:28 am 

    FWIW, the weeklies are probably running high right now. They ran low a year ago. They just have a hard time dealing with changes in production, especially if DUC inventory becomes a real variable.

    There is no conspiracy. And EIA has estimated/projected low much more than high over the years than high. Peakers just can’t stand what shale has done. Ruined their little retard doom party. Killed The Oil Drum. Etc.

  16. rockman on Sat, 5th Aug 2017 2:12 pm 

    RD – I’ll get back to adjustments in a bit. But looking for that support ran these amazing stats. The link is below but folks need to check out this quick look:

    Texas oil production 2009: 345 million bbls.
    And that gave an avg. of 6 bopd/well from 158k wells.

    Texas oil production 2015: 1 billion bbls.
    And that gave an avg. of 14.2 bopd from 195k wells.

    Those stats tend to provide a bit of balance to all the shale play hype in Texas.

    IOW all those initially high producing shale wells raised the avg. production PER WELL from 6 bopd to just about 14 bopd. So while the shale boom added a increased total production from 345 million bopd to 1 BILLION bopd the average well in Texas remains at a stripper level rate.

    Thee Texas shale play isn’t that much of a correction when look at from that perspective.

  17. rockman on Sat, 5th Aug 2017 2:40 pm 

    From the TRRC: “Preliminary Crude Oil & Gas Well Gas Monthly Production Total – Significant changes to the preliminary production figures will occur in decreasing amounts for approximately six to eight months due to the filing of corrected and late reports by industry. Commission staff anticipates that the production totals following that period are substantially complete, although continued minor changes occur thereafter. There is no point beyond which an operator may not file corrected production reports.”

    From our cohort Ron Patterson: “It takes about 18 months for the RRC reported data to be within 1 percent of the final value (when all the data is in and finalized). It is not clear why the NDIC (North Dakota) is able to compile their data so much more quickly, it may be the fact that they only produce one third or less of the oil and natural gas produced in Texas. In any case, due to Dean’s fantastic work, we have pretty good estimates through at least April 2016.”

    Lots of charts here:

  18. Anonymous on Sat, 5th Aug 2017 5:57 pm 

    You learned! You corrected your old misconceptions. Good job.

    Of course, it is irrelevant to EIA estimates since they are not based on RRC data. They do a direct producer survey (the “914 report”).

  19. boat on Sat, 5th Aug 2017 6:04 pm 


    Go read the methodology on energy reporting at the eia site.

  20. Anonymous on Sun, 6th Aug 2017 9:04 am 

    I have already read it and it backs me up.

    “Previously, EIA used state-reported data to estimate monthly crude oil production. However, some states took several months to as much as two years to report complete (final) data, rendering these data of limited value. Only two or three states reported final data in a timely manner.

    With the expanded Form EIA-914 survey, EIA collects data from a sample of less than 500 out of 13,000 currently active operators of oil and natural gas wells. These operators account for about 90% of crude oil, lease condensate, and natural gas production in the Lower 48 states and a significant share of the total oil and natural gas production in each of the 15 states and the federal Gulf of Mexico for which data are individually collected. Data from this relatively small number of operators make it possible to generate statistically representative estimates of production within two months of the production month.”

    The bottom line is the RRC has huge lags before you get the complete picture. In fact, I am glad to see Rockman note this as in the past he has been guilty of not understanding it. Many other people on peak oil websites (Roger Blanchard, Ron Patterson, and countless commenters) have made the same mistake.

    But in any case, EIA does not rely on the RRC for the monthly time series for any times within last two years (where the lag problem occurs.) Actually even before the 914 report, EIA recognized the problem with RRC data and did their own estimate that was more of an extrapolation.

    But this is Groundhog Day so inevitably we will have future comments saying there was a big peak (from the updating lag) or Texas oilman who will say they have the best regulator so of course the data must be up to date.

Leave a Reply

Your email address will not be published. Required fields are marked *