It is hard these days to attempt to temper the enthusiasm ignited by the spectacular rise in shale oil production. Any attempt to point out flaws in the rosy picture presented by fracking enthusiasts is met with a barrage of counterpoints, which are sometimes hard to argue against. In the past few years, we have been in the quick production ramp-up phase of a brand new source of oil, so it has been hard to say anything even slightly pessimistic about production profiles that look like this:
According to North Dakota's official government data in December 2011, per well production in the North Dakota Bakken was 144 b/d. There were two other months previous to that when average production per well surpassed this number, but I should point out that previous to this, those spikes were caused by a larger than average increase in wells commissioned, which affected the monthly average production because of the relatively few wells in production at the time. Since December 2011, the 144 b/d average was matched in the immediate few months in early to middle 2012, but since then, there has been a steady decline. It is still early to project a definite trending rate, but in the past 18 months, we had a decline of 12%. Per well production in June 2013 was 129 b/d. Year-on-year, production per well is down an average of 8% so far for the first half of 2013, compared with the 2012 average.
Can this trend be reversed? Perhaps, but I doubt it would last longer than perhaps a year in duration, because there are now too many wells past their first phase of production, which means that they are past their maximum rate of production and declining rapidly. In the absence of an inexplicable, temporary ramp up in drilling, this is the new established trend. Up to two hundred wells coming online every month cannot affect the average rate of per well production as much as they used to.
This was expected to happen of course, given what we know about the fast decline rate of shale oil wells. There is a finite number of wells that can be drilled in the formation, estimated at over 40,000 by the EIA and about 6,000 wells were drilled in North Dakota already. The rate of increase in the total number of wells has fallen off sharply in 2013 compared with previous years, when the rate of increase in total wells was between 50-60% per year. If the rate of drilling continues at the same pace as in the first half of 2013, we will have a year on year increase of 33%.
Note: The pink line represents the relatively steady percent increase of total wells in production year on year between 2009-2012 and the sharp drop in that rate of increase we have seen so far in 2013. The blue line, which represents average production per well has been somewhat volatile, but the correlation between the drop in the rate of increase in total wells in production from year to year and the relatively sharp drop in the rate of production per well are definitely correlated.
Assuming the obvious -- that the number of wells drilled every year will stop growing exponentially in order to maintain the 50-60% increase we had between 2009-2012 -- we will have a year-on-year increase in wells of only 10% by 2020, assuming that from now on we will have a 2,000 well increase per year.
Note: For 2013 I assumed the current rate of well additions to continue for the rest of the year. From 2014 to 2020, I assumed that there will be an average of 2,000 wells drilled per year, which is more than any previous year so far.
Incidentally, by 2020 there will be over 20,000 wells drilled in the North Dakota Bakken formation if my assumed pace of drilling will be maintained. Some time around 2020 there will be a gradual decline in wells drilled every year, assuming that the EIA estimate of wells that will ultimately be drilled is approximately accurate. Furthermore, the quality of the deposits being drilled is likely to decline, further deteriorating the per-well production rate.
The fact that per well production started its irreversible decline does not mean that there is an imminent peak in oil production in the Bakken. There will be at least a few more years where drilling about 2,000 wells per year will more than offset the per well decline rate. As the ratio of declining wells to new wells drilled every year gets worse, most likely in less than a decade, total production will start declining. We have no way of knowing for sure how fast this per-well decline rate will be. Assuming that ultimately the total number of wells drilled in the North Dakota portion of the Bakken will surpass the current EIA estimate and there will be 45,000 wells that will ultimately be drilled (Click here for an opposing view to EIA's estimate) and an assumed steady decline rate of 3% per year of per well production, here is what it will look like:
While this may be only a hypothetical scenario, it is probably close to what will actually happen in the event that the EIA estimate of wells that will ultimately be drilled is not too optimistic or pessimistic and assuming that a 3% per year decline rate in per well production is close to achievable reality. For 2013 per well decline was 8% thus far, compared to average per well production in 2012, but it comes due to a sharp decline in the percentage gain in total wells. Assuming that we will see a relatively stable well addition rate of about 2,000 per year, the decline in the percentage gain in total wells will moderate significantly for as long as we will see this assumed rate of well addition in the field.
Recognizing this event is important, because it signals the turning of the page towards a new chapter in the Bakken shale oil evolution. It is a chapter that will settle many arguments between the optimists and pessimists, centering around the question of what volumes of production we should expect going forward. For instance, we will see whether the rate of per well production decline will accelerate or follow a steady linear evolution, depending on the number and quality of new wells coming on line, as well as the decline rate of each well currently in production. We will also get a better idea in regards to the overall profitability of producing this field. We will see companies venture outside the currently established sweet spots and we will get a better idea of what is out there, beyond what we already know. Bottom line is that from now on, those who follow the Bakken story -- out of pure curiosity or perhaps as potential investors -- now have another tool available to double check whether current claims made by industry as well as impartial entities such as the EIA measure up to reality. With every data point we will get in regards to average per well production rates, we will have a clearer picture. It is the beginning of the resolution to a debate that has been raging for half a decade now.