Bakken Part 2: Four Takes on Play Analysis

July 14, 2011 by

Editor’s Note: While VISAGE rebranded to VERDAZO in April 2016, we haven’t changed the VISAGE name in our previous blog posts. We’re proud of our decade of work as VISAGE and that lives on within these blogs. Enjoy.

I’d like to start by thanking the gDC (geoLOGIC Data Center) and geoLOGIC Systems for supporting this blog by providing VISAGE with access to their data…. and now for the blog…..

In our previous blog post we looked at the growth of the Bakken and who the main players are.  When looking at type curves and the variability of the associated data, it becomes very clear that there are many, many factors that need to be considered when doing type curves. The most important consideration is the selection of wells that are comparable and representative of a particular location or situation.  Some things to consider in the Bakken are leg count, frac count (if any), horizontal length, location, horizontal azimuth …. and the list goes on. We’ll show you how the horizontal Bakken wells (since 2007) collectively behave and how much they vary. We can hone in on some more specific analyses with your help (please post your comments).

Take 1: Type Curve (Rate vs Time)

This chart shows the following for Bakken Horizontal Wells post Jan 2007:

  • The average rate vs time “type curve”, which is the average of producing wells
  • The P90 and P10 percentile trendlines (these represent the boundaries in which you will find 80% of the wells. Percentile trendlines are discussed in detail in our blog “So what is the problem with production type curves?
  • The P50 is the median (i.e. half the wells are greater than this number and half the wells are less than this number).
  • The type curve of cumulative production (used to help estimate expected payout on a new well). If you are interested in seeing this with P90 and P10 lines please let us know.
  • Producing well count  is important when doing any kind of type curve as it tells you how many wells contributed to the mean value calculation. When this value gets too low, the sample size may not be large enough to be representative.

Take 2: Percentile Distributions

In our last blog we looked at how the production rates have been changing over time and noted that early indications are that 2011 wells will outperform previous years. An increased understanding of the play combined with drilling and frac technology have had a profound impact on what we can expect from the Bakken. Visual tools like percentile plots can help communicate just how variable the data is. In this chart each dot represents a well. The colours represent different years. While we can see the P50’s have evolved with each production year, the values seem to have a similar range of variability each year.

Are Initial Rates a Good Measure of Performance?
Last blog we also looked at the top 10 wells in the last year. It was pointed out to me that there are many criteria for what a “top well” is. Initial rates are one way of measuring wells, but in some circumstances you may choke back a well for sand control … in such a  case that well may look better in the context of cumulative production in the first 6 months. Let us know what your criteria are for measuring “top wells” by posting a comment.

Take 3: Horizontal Direction (Rate vs Time) 

Type curves are more than just a means of characterizing “typical” well behaviour, they are a visual tool for comparing collections of wells. In this case we can see that the direction the horizontal well was drilled (i.e. azimuth) impacts production behaviour. The notable aspect of this chart is that while E-W wells do not appear to have the best initial rates, they tend to sustain a better rate after 9 months of production than other orientations.

Take 4: Horizontal Direction (Rate vs Cumulative Production)

Another visual tool used to compare well collections is a normalized rate vs cumulative production type curve. This allows us to compare how rate changes with cumulative production over time. I chose month 32 as an example point of comparison.

There you have it. Four takes on play analysis and well behavior in the Bakken. If you found this interesting or would like to see other analyses please post a comment.
Stay tuned for more analysis in the next few days!

Disclaimer of Analysis Results
It is important to note when doing any of this kind of analysis that the selection of wells be scrutinized for comparable, meaningful results. Given we are looking at the play in its entirety, the results presented are general and are intended to fuel your understanding of play analysis. We would be happy to run analyses for specific well lists that you may have. Please feel free to contact us.


Special thanks to geoLOGIC systems for providing VISAGE access to data from the gDC (geoLOGIC Data Center)

Data Disclaimer: gDC is the data source for the information presented in this blog post. The well list used was provided by an independent geologist. If you see issues with the data or have concerns about the well list used for this posting please contact us.