In the first and second parts of this series I provided an opposing viewpoint to an article published on theoildrum.com. In my most recent article, I produced rough estimates of Bakken well production. I was able to show the Williston Basin produces significantly different data from one county to the next. In the third part of this series I will use specific well production data to show the Red Queen article provides a skewed representation of the Bakken.
I have compiled data on 40 wells in Mountrail County, and have found very good results. These wells were located in four different fields around the Sanish Field. I did not include any Sanish Field results as this field was one of the focal points in the "Red Queen" article. Since well completions are continually evolving and producing more resource, I only used wells completed in 2011 with a years worth of production. I included specific completion data to show why certain wells performed better than others.
One variable the "Red Queen" article did not account for was lateral length. EOG Resources (EOG) and Enerplus (ERF) have done a large number of short laterals. These companies, and others continue to use this approach. Short laterals are generally around 5000 feet versus long laterals which are approximately 10000 feet. As one could guess, short laterals produce about half the production of a long lateral. Long laterals have contact with a larger area of the formation. The number of stages are also important. The higher the number of stages used the better the production. As a well increases the number of stages, each stage will be in a shorter length. Shorter stages produce better stimulation as the hydraulic horsepower of the pump trucks produces greater pressure. Increased volumes of water and proppant also provide for better production. When the formation is fracced the water pushes the proppant into the fractures. This proppant "props" open the fractures allowing for more oil to flow into the lateral. The better the proppant the longer the fractures are held open. Sand does not hold open these fractures as long and this shuts off oil flow sooner. Keeping this in mind, each operator calculates well cost and offsets this with what it believes production increases will be. We are seeing more capital used to drill and complete wells, as the increased revenue from production more than offsets these costs.
In the table below I have listed 2011 wells in western Mountrail County. This is a very good area, but the results I found were better than I originally expected.
|Well||Operator||Days||120 Days||360 Days||Lateral||Stages||Choke||Water||Proppant||Ceramic|
|19985||Hess (HES)||383||69601||116202||9341|| |
|19403||Continental Resources (CLR)||403||47870||89381||8749||24||20/64||1985482||2309500||649200||TF|
|19339||Exxon Mobil (XOM)||409||24619||47727||10198||18||20/64||2394068||2757849||Sand||MB|
This research covers a large area. The wells used are as far apart as 43 miles. Of these forty wells, only three wells produced less that the 70000 barrels of oil in the first twelve months. Of these three, one well was a short lateral that still produced more than 66000 barrels of oil in that time frame. This is important as the "Red Queen" article used that annual production estimate to model the low end of the Bakken production curve. Eleven wells produced more than 150000 barrels of oil, and that does not include the short laterals that produced more than 75000 barrels. Over one quarter of these wells produced at least 33% more barrels of oil than the highest point of Mr. Likvern's curve when differences for lateral length are accounted for.
The Red Queen article uses Mr. Likvern's annual estimated production to define how it will grow over 29.5 year well life. Here are his estimated ultimate recoveries or EURs based on those annual production numbers.
|70000 Barrels of Oil||240000 Barrels of Oil|
|85000 Barrels of Oil||290000 Barrels of Oil|
|100000 Barrels of Oil||340000 Barrels of Oil|
In summary, those familiar with North Dakota production know these EURs are low. In fact, it is lower than most if not all estimates given by oil producers in North Dakota. The "Red Queen" production model is too general to be useful. It provides little in differences to area or oil producers. Kodiak (KOG) continually has better well results than its competition. This is why Kodiak is comforable with some of the highest well costs in the basin. Western Mountrail is different from western Williams County. The middle Bakken in Mountrail is not only thicker, but it is deeper as well. Well pressures are greater in Mountrail which produces much larger initial production rates. These rates deplete faster, but still produce significantly greater resource. Well costs are on average $2 million higher in this area than other more shallow locations. Western Williams County has lower IP rates, but deplete slower. This is why we see 24 hour IP rates north of 4000 Boe/d in Mountrail County produce EURs of 900 to 1000 MBoe. Western Williams wells produce 24 hour IP rates south of 1000 Boe/d, but still have EURs of 450 MBoe. It is important to break down the Bakken into areas, to properly judge economics.
Disclosure: I am long KOG. This is not a buy recommendation. In the first table I have used stock symbols for public companies, while Slawson's(private)company name is listed in full. Also in the first table, cumulative production numbers are measured in barrels of oil. Lateral length is measured in feet. Water is measured in gallons and proppant in pounds.