A recent article from theoildrum.com, titled Is Shale Oil Production from the Bakken Headed for a Run with "The Red Queen"? addressed issues with respect to Williston Basin production. This article has received significant attention, as it not only is a good paper, but also goes over a large amount of information with respect to middle Bakken production. Rune Likvern's research produced results that run contrary to my findings, so I will cover this topic to hopefully shed some light on the subject. I would like to add that Mr. Likvern does have a very good grasp of this subject. Although my findings have been contrary to his, I would like to add that Mr. Likvern did a very good job on his article and deserves recognition for his work here.
Mr. Likvern's findings from his in-depth study of time series for production from some individual wells are:
1. Estimated breakeven price for the average Bakken formation well in North Dakota is $80-$90/barrel of oil.
2. The average well yields around 85000 barrels of oil over the first 12 months of production and experiences a year over year decline (depletion) of 40%.
3. The recent trend for newer "average" wells is one of perceptible decline in well productivity (lower yields).
4. As of 2007 and recent months, the total production of oil from the Bakken has shown exceptional growth and the specific average productivity has been sustained by starting up flow from an accelerating number of new wells.
5. Based upon present and observed trends for principally well productivity and crude oil futures it is challenging to find support for the idea that total production from shale oil from the Bakken formation will move much above present levels of 0.6-0.7 Mb/d on an annual basis.
These findings are worrisome for the Bakken investor as it shows unconventional wells in the Williston Basin are not economic at today's prices. His production numbers include the middle Bakken, Upper Three Forks, the second bench of the Three Forks and Sanish (Pronghorn) formations. So this is not a true indicator of middle Bakken production, but more of an average for unconventional wells in the Williston Basin. I would also like to point out that historically the middle Bakken is not the formation that has produced the most oil in North Dakota. Madison total production is almost three times that of the middle Bakken at over 923 million barrels.
In figure 2 of Mr. Likvern's article, it shows specific areas of development in the Williston Basin. It is thought that these are the best areas in North Dakota. Ross, Stanley, Sanish and Parshall fields have all been prolific with respect to crude oil production. Other important areas are Elm Coulee field and areas to the east or west of the Nesson Anticline. Mountrail County was one of the first areas of development due to a lower depth that creates a higher pressure environment. This helps to create very large initial production rates, when coupled with middle Bakken shale thickness being over 100 feet in this area. Field development in this area has much to do with Statoil (STO) in Ross, Whiting (WLL) in Sanish and EOG Resources (EOG) in Parshall fields. Each had large acreage positions in this area and were able to develop these fields on a larger scale.
In Figure 3, the chart shows how the price of oil has affected the number of wells drilled in middle Bakken/upper Three Forks. There is some truth in this comparison, as a $60 oil price is needed to profit from these wells on average. When the price of oil drops below this, it gets tight even in the better areas. Technology has also been a driver in increasing the number of wells and total production, as it has helped margins through better IP rates and EURs.
Addressing the price of oil and how it equates to the number of wells drilled is not as simple as it looks. Oil price can differ from one producer to another given its hedges. A $60 oil price would hurt a company with a lower percentage of production hedged, than many of the MLPs that hedge virtually all production. Looking at the drop of oil price in 2008, it significantly affected the number of wells drilled. When oil prices declined from $140/barrel to $40/barrel a decline in the number of wells drilled affected all plays in the United States. Anything is possible, but it is highly unlikely we will see WTI sell for $40/barrel anytime in the near future. Another factor is acreage. Companies like Kodiak (KOG), ConocoPhillips (COP), Newfield (NFX) are able to continue development of acreage in western Mountrail and northeastern McKenzie counties given the production garnered from these areas being much better than say western Williams County.
Another issue often misunderstood is well cost and how this relates to the oil price needed for a well to be economic. Well costs continue to head upward, but as this has happened well quality has increased as well. Companies like Whiting have maintained a well cost in the $7 to $8 million range, but have done this sacrificing long term EURs by using less water and proppant. Kodiak is the opposite of Whiting, as it uses more water and proppant. It also uses better proppant, as we are starting to see its wells deplete as a much lower rate. This is why its well costs are in the $10 to $11 million range. In essence, the increase in total well costs are from a better well design, which translates to better overall recovery.
Figure 5 of this article produces a diagram used to show selected well results and how this equates to average production in the Bakken/Three Forks. I did like this diagram, especially with respect to cumulative barrels of oil produced. Many of these types comparisons focus on 24 hour IP rates and not longer term production. Initial production rates can vary significantly with just a slight difference in the size of choke used. On the other hand, 90 IP rates give a much better view of EURs. What is missing in this collection of data is very difficult to qualify without very specific data. We cannot generalize the whole Williston Basin, as some areas are not even economic, while others are very profitable. Another problem stems from the operator. Each oil producer has its own way of drilling and completing which not only affects production, but also well cost.
Figure 5 uses 7 wells to show the difference in an average Bakken well and some of the best. These wells are all in Sanish Field with six operated by Brigham and one Whiting. Two of these wells were drilled in 2011 with the other five in 2010. The 2011 wells were both drilled early in the year.
|Well||IP rate||90 day IP||Total Oil Produced||Lateral||Date||Choke||Proppant||Water||Stages|
|Clifford Bakke 26-35-1H||4438||1301||322011||9453||10/10||158/64||3874140||3516325||38|
|Sorenson 29-32 2H||4661||914||214979||10053||03/11||181/64||4106600||3570032||38|
|Brown 30-19 1H||2789||646||142412||9600||04/11||118/64||3671960||3274966|
|Arvid Anderson 14-11 1H||2834||742||200871||9465||11/10||96/64||3867540||3351162||38|
|Jerome Anderson 15-10 1H||2678||584||201639||9030||03/10||136/64||3673580||3004512|
|Liffrig 29-20 1H||2037||518||166641||9130||02/10||181/64||3642780||2900565|
Although these wells were completed a couple of years ago, it has the same well design as many of the producers are using today. What the Red Queen article misses is the geology is only a small part of what makes a good well. There is no doubt Alger and Sanish fields are some of the best Williston Basin acreage, but fields like Westberg and Poe have produced excellent results as well. The author stated that the Liffrig well is the average producer in North Dakota, but that was also a time when the majority of producers were still using twenty stage fracs and getting comfortable with the geology.
In future articles on this subject I will tie in more recent results and how those compare with the wells completed two and three years ago. What these well results will show is how closely production is tied into well design and has little to do with small sweet spots from one mile to the next. This is important, as much of the information used in this research back in 2010 had very large production differences from one operator to the next. We will also see that very large specific areas are economic. Those companies that fail to spend money to produce oil in the Bakken will find depletion numbers much higher than that of a company that understands the long term benefits of using the best completion methods.
Disclosure: I am long KOG.
Additional disclosure: This is not a buy recommendation. IP rates measured in barrels of oil per day. Water measured in gallons. Proppant is measured in pounds. Laterals measured in feet.