Bakken Update, Is Production Slowing Down?

The Bakken production data published by North Dakota is out with October production data. ND Monthly Bakken Oil Production Statistics (Bakken Only) and ND Monthly Oil Production Statistics (All North Dakota).

The below chart is North Dakota Bakken barrels per day and All North Dakota Barrels per day. As you can see it’s mostly Bakken and very little “the rest of North Dakota.

Bakken Barrels Per Day


It looks like there is a pattern developing. I charted the month to month percent change in the chart below with a line indicating a possible trend.

Bakken Percent Change

Basically I think you can ignore everything before July 2011. That is because that was the month the surge started, additional wells almost tripled that month and Bakken production has tripled since that date.

From the Director’s Cut:

The drilling rig count was unchanged from Sep to Oct and the number of well
completions dropped only slightly from 225 to 202, resulting in only a 1% increase in oil
production. All McKenzie County roads were shut down for 3-4 days due to rain.
McKenzie County has 1/3 of the drilling rigs and 29% of state production. This means
production likely would have been 10,000-15,000 barrels per day higher without that
weather event. Days from spud to initial production increased from 100 to 114. The
uncertainty surrounding federal policies on taxation and hydraulic fracturing regulation
continues to make investors nervous.

We estimate that at the end of Oct there were about 460 wells waiting on completion
services, a decrease of 40.

202 wells were completed in October but production was up less than 1%. Production was up 8,495 bp/d in the Bakken and 8,652 bp/d in all North Dakota. I think that my previous estimate of 150 to 155 new well completions in order to keep production flat was a bit low. If we had only 17o new wells completed in October I don’t think there is any doubt that production would have been down. But perhaps in good weather when all roads were open then 155 wells would have been enough.

Actual change (increase) in Barrels Per Day:

Jul-13 51,510
Aug-13 37,482
Sep-13 20,359
Oct-13 8,495

Additional wells were 195 in the Bakken and 211 in all North Dakota. Additional wells are new wells + shut in wells re-opened – wells shut in. So that number could be either more or less than new wells.

Helms gives us the price of oil, presumably the price Bakken producers are getting:

Sep Sweet Crude Price = $92.96/barrel
Oct Sweet Crude Price = $85.16/barrel
Nov Sweet Crude Price = $71.42/barrel
Today Sweet Crude Price = $73.00/barrel

Apparently Bakken crude does not fetch the same price as WTI. That is understandable because Bakken crude is so light it can almost be considered as condensate. The reason for this is this is really “tight oil”, so tight that only the smallest molecules can escape.

Bakken Barrels per day per well is charted below. As you can see bpd per well had started to drop pretty dramatically before the surge. Then almost tripling the number of new wells per month increased the bp/d per well quite fast. Now it appears to be dropping again. Of course it this is expected as all those old wells are declining quite rapidly.

Bakken Barrels Per Day


Bakken barrels per day per well had started to recover somewhat to 135 after dropping to 129 but dropped back down to 132 in October. Bp/d per well reached a post surge high 144 back in May and June f 2012.

EDIT: It made the Bakken Breakout: Oil production increasing at a slower pace

EDIT: And a couple of other really important stories on the web:  Inexpensive oil vanishing at alarming rate And: Some ND drilling ‘becoming uneconomic’

This entry was posted in Uncategorized. Bookmark the permalink.

95 Responses to Bakken Update, Is Production Slowing Down?

  1. dcoyne78 says:

    Hi RalphW,

    We could do two separate models, but one of the inputs to my model is the change in the number of producing wells which can be any number that someone wants to use.

    For historical input we use the data from the NDIC and adjust this slightly to estimate older wells that will get shut in over time, but the adjustment is pretty small because in Dec 2004 there were only 189 wells producing in the Bakken and we assume the well life is 30 years, we use the 1953 to 2004 data to estimate when these 189 wells get shut in and if on average the new wells from 2005 or later get shut in after 30 years that would start in 2035. The effect on the overall model is small.

    Note that the answer is not really “always”. It depends on future real oil prices and how many wells are added in the future. If after the wells stop being added in Jan 2014, the oil price never rises back to current real prices, or for some other reason (a change in fracking regulations for instance) the number of new wells added can only rise to 150 per month, then the conclusions would be different.

    I think you are suggesting that if wells were not added for two years that the new peak will be lower than current levels. I will test this hypothesis using the following scenario:

    Lets say no wells are added for 10 years, in this case the eventual decrease in new well EUR is also delayed by 10 years. To simplify (but make the model less realistic) I will have no new wells added starting in Jan 2014 and resume adding new wells in Jan 2023 (at a lower level of 150 wells/month), but note that this will delay the start of the decrease in new well EUR from June 2015 to June 2025 the maximum decrease in new well EUR is also lower to keep the TRR at 8.5 Gb(mean USGS estimate). Also note that Brent oil prices (in May 2013$) would need to fall to $61/barrel by Dec 31, 2013 to make this scenario plausible, this is the current breakeven real oil price for the average Bakken well on a point forward basis.

    Chart below for this scenario (which I consider quite unrealistic, it is a “what if” type of exercise.)

    For comparison I have shown an alternative scenario below that adds 150 wells per month from Sept 2013 until 37,800 wells are reached in Feb 2031


  2. Watcher says:

    So a peak is reached sooner by continuing to drill? Versus suspending drilling and allowing a decline to assert for a while and then restoring drilling — unless you wait 12 years?

    Seems unlikely. The decline rate is still pretty high after just 2 years of no drilling, and you’re saying peak is only a couple of years away anyway with full drilling. So what is changed? Decline rate in 2 years is what, 120K bpd/month? If you stop drilling now for 2 years it would only be down to what, 50K bpd/month?

    Did your graph carry forward the residual decline rate? It looks like only new decline rate is embedded.

    • dcoyne78 says:

      Hi Watcher,

      I think I may have not understood what you were looking for.

      I was attempting to create a scenario where the Dec 2013 level of output (which my model predicts will be 930 kb/d) would not be surpassed by a secondary peak if drilling resumes after being suspended. If drilling was suspended for 12 years and only 145 wells were added at maximum after that delay, the new peak would be below 930 kb/d.

      The comparison you might have been looking for was something less than the current “predicted peak” say 1.2 MMb/d, to take Ron’s estimate. If you stop drilling for a while the new peak will be lower and the longer the delay, the lower the peak, note that my initial scenario at 200 wells added per month (no suspension of drilling) was 1.4 MMb/d, and the scenario with a 1 year delay had a somewhat lower peak, when the delay was 10 years the peak was even lower (but that was in part because I was trying to make it lower to attain what I initially thought you were looking for so the wells added was reduced to 150 wells per month).

      So I will try again with a scenario to match a 1.2 MMb/d peak with 175 wells added per month.


      Then a new scenario which stops all new wells after Dec 31, 2013 until Dec 31, 2015 then they jump back up to 175/month Jan 2016.


      Note that the legacy decline does not increase when no new wells are added both because the output decreases so 6% of 800 would be less than 6 % of 900 and also because the rate of decline decreases as wells get older, see chart below.


      Also if we drill 200 wells per month and reach peak then the legacy decline is less than you would think because new well EUR has started to decrease so the 1st month output of new wells might be 300 b/d times 200 wells so legacy decline would be about 60kb/mo at peak, note that currently legacy decline is about 45 kb/mo, the EIA estimates are off if we use the Hughes estimate of 400 b for first month output and NDIC data for change in wells and output data.


  3. Pingback: Michael Klare: Have the obits for peak oil come too soon?

  4. Pingback: Peak or no Peak, Oil is Unsafe at any Price | Informed Comment

  5. Pingback: Michael Klare: Have the Obits for Peak Oil Come Too Soon? | naked capitalism

  6. Pingback: Arab News Blog » Peak or no Peak, Oil is Unsafe at any Price

  7. Pingback: Michael Klare: Have the Obits for Peak Oil Come Too Soon? | Jo W. Weber

  8. Pingback: Think “peak oil” is a discredited idea? Think again | Grist

  9. Pingback: Peak Oil Is Dead. Long Live Peak Oil! (The Nation) | Uma (in)certa antropologia

  10. Pingback: Have the Obits for Peak Oil Come Too Soon? | VantageWire

  11. Pingback: » Forse il necrologio per il picco del petrolio potrebbe essere arrivato troppo presto – Michael T. Klare

  12. Nony says:

    The Nov2013 increase for Bakken was 20,000 bpd. (Equates to a 240,000bpd annual increase.) So I don’t think looking at the change from Aug, Sep, Oct shows a real slowing, given next data point was healthy.

    FYI: NOV2013-NOV2012= 237kpd increase, while NOV2012-NOV2011=227kpd increase. Seems pretty linear, when looked at for more than small data point. I guess you could plot the whole monthly time series and use trailing 12 months increase (so taking seasonality out).

  13. Pingback: Peak Oil becomes an Issue Again after the IEA Revises its Predictions | Olduvaiblog

Comments are closed.