Projecting US shale oil production (after December 2016)

In July 2016 and February 2017 I published posts (which you can find here and here) in which I showed how much oil existing horizontal wells were going to produce in the future, if we would assume that their decline behavior would be similar to those of earlier wells, combined with a certain terminal decline rate and economic limit.

Following these simple rules, I will repeat the same exercise in this post; now wells are included that started before 2017, and their future oil production is estimated. There have not been any differences in the approach, or the coverage, and therefore, if you are interested in a more detailed discussion on how to interpret these results, please have a look at the links given above as well.

This interactive presentation only contains the oil production from horizontal wells, from the main oil producing states that I cover (CO, MT, ND, NM, TX, WY). Oil production from vertical wells is excluded, and no estimate is made of how much wells that started since 2017 may contribute in the future. Neither is gas production considered.

Still, with the results posted here, we should be able to get a general sense of how much these existing horizontal wells will produce in the future, what the differences are between the basins/operators, how ultimate recoveries have changed over time, and how much is recovered in each of the years that these wells are online.

Let’s go over these views one by one.

“Projected Production”

Here we can see the historical, and projected oil production from 59,266 horizontal wells that started production before Jan 2017 in the following states: Montana (since 2003), North Dakota (since 2005), Colorado, Wyoming, New Mexico, Texas and Wyoming (all since 2010).

The view is split between Dec 2016 and Jan 2017, which is from when the projections start.

If you compare the results here with my previous post half a year ago, you may notice that I have somewhat under-estimated the actual production of earlier wells. E.g., in the previous post I had total production from wells prior to 2016 at about 1.3 million bo/d in Jan 2018, while that is now a bit over 1.4 million bo/d.

In that post I already gave possible reasons why that may happen (wells may produce under the economic limit I set, e.g. because of additional gas production, or other reasons, and the effect from recompletions is not estimated). I belief those reasons are still valid. Furthermore, the exponential terminal decline rate that I assume for wells late in life, on which little existing data is available, may be too pessimistic. Therefore also the results shown here will likely turn out somewhat too low.

“Production Profiles”

The average production rates, and cumulative production, for all these wells can be seen in the 2 graphs shown here. If you wish to see only historical, or projected production, you can use the “Show production” selection at the top. Historical production is marked with dots on the curves.

If you click on a year in the legend, the related wells will be highlighted. You may then see some overlap of the historical and projected production rate profiles; this is because wells are grouped by the year in which they started. A well that started earlier in a particular year will have more months of actual production data than wells that started production later in the same year, and therefore the projected output of this earlier well will start later.

The odd-looking upticks at the end of each of the historical curves are caused by the fact that I had 2 additional reported months for North Dakota, which I used here. So those last 2 marks consists only of wells located in this region, and as these wells decline at a lower speed on average than elsewhere, they cause these curves to move upwards at the end. For 2010 you can see that this happens for more marks, but the reason is the same; early 2010 more Bakken wells started production, than in other regions.

There are a few more artifacts that you can find here, especially at very low production levels, but I belief that on a high-level, these shapes are reasonable projections.

The other axes settings (linear, and log-log), both provide other very interesting views on this data.

“Estimated Ultimate Recovery”

Here I wanted to share some insights into the total historical & projected oil production from all these wells. Both charts here use the projected production profiles, as seen in the previous overview.

The top chart shows the (gross) total oil production for each of the 10 largest operators, and how much they already produced versus projected production. Just use the “Operator” selection if you wish to see other operators compared here.

The bottom chart shows for all selected wells the estimated ultimate recovery (EUR), for each of the basins. The aim here is to show the differences in well EURs for each basin, but also how these recoveries are distributed over the years that a well is producing. This latter can be seen by the colors used; for example, the blue color at the bottom is the portion that is recovered in the first year on production.

You can use the “First flow” selection to select only wells that started in a particular year. For example, if you only select the years 2015 and 2016 you’ll see that the average recovery from wells in the Permian is higher than before.

On Tuesday I plan to have a post on a new state, followed with updates on the Eagle Ford and the Permian in the days after.

For this presentation, I used data gathered from the following sources:

  • Colorado Oil & Gas Conservation Commission
  • Montana Board of Oil and Gas
  • New Mexico Oil Conservation Commission
  • North Dakota Department of Natural Resources
  • Texas Railroad Commission. I’ve estimated individual well production from well status & lease production data, as these are otherwise not provided. About 7% of the horizontal Permian wells in Texas are excluded, as these were mixed with too many vertical wells on a lease, making reasonable well profile estimations impossible.
  • Wyoming Oil & Gas Conservation Commission


The above presentation has many interactive features:

  • You can click through the blocks on the top to see the slides.
  • Each slide has filters that can be set, e.g. to select individual or groups of operators. You can first click “all” to deselect all items. You have to click the “apply” button at the bottom to enforce the changes. After that, click anywhere on the presentation.
  • Tool tips are shown by just hovering the mouse over parts of the presentation.
  • You can move the map around, and zoom in/out.
  • By clicking on the legend you can highlight selected items.
  • Note that filters have to be set for each tab separately.
  • The operator who currently owns the well is designated by “operator (current)”. The operator who operated a well in a past month is designated by “operator (actual)”. This distinction is useful when the ownership of a well changed over time.
  • If you have any questions on how to use the interactivity, or how to analyze specific questions, please don’t hesitate to ask.

10 thoughts on “Projecting US shale oil production (after December 2016)”

  1. Great posting. I would suggest including a log-oil rate vs. cumulative oil plot (w/ forecast). These are very useful for forecasting final recovery and diagnosing flow regimes (especially in shale with initial b-factors around 1)… Not that I don’t trust ya, but it is difficult to gauge the fidelity of your forecasts without this data (I’d say the same to our reservoir engineers).

  2. Enno
    Thank you, once again, for the excellent research and presentation.

    I have mentioned before that it would be good to compare apples-to-apples, by normalizing the individual production rates of the wells by the number of completion stages. It’s not rocket science. In a low perm rock, if you increase the contact surface area (approximated by the number of stages) you increase the instantaneous production… but you also accelerate the decline.

    It is great to see that rate acceleration phenomenon in your plots. The more recent wells (which can have up to 50-60 stages as compared to 20 or so in the past) have higher initial rates, but they decline faster. Nothing comes for free… It does appear that the average EUR (without any additional EOR) is around 300k bo (it does vary by basin, as one would expect).

    Do you have access to the completion information or is that data too difficult to get to ?

    1. JP,

      Thanks for sharing your insights, and suggestion of another important factor to consider; the # of stages.

      I can now process all completion data in FracFocus. Although this contains very interesting and useful data, which I aim to reveal in the coming period in my posts, e.g. proppant mass, fluid volume, proppant type & mesh, it typically does not contain data on the number of frac stages. Some states do offer some data on this, but it’s far from straightforward to gather it on a large scale unfortunately. Still, you’ve given it a high place on my list of data items to look out for!

      1. I have a question. I clicked on 2016 EUR for 2016. It looks to me that Permian recovers roughly 50% of the oil in the first year. Am I interpreting this correctly?

        1. Krisvis,

          Thanks for the observation.

          > It looks to me that Permian recovers roughly 50% of the oil in the first year. Am I interpreting this correctly?

          That is indeed what this method comes up with. For a detailed discussion on how the method was designed, the assumptions used, please see the earlier posts.

          I don’t claim that the method applied is perfect, or that it has a very high accuracy. Indeed, I typically refrain from discussions about the future. However, I wanted to use an empirical way to figure out very roughly how much we can expect from these wells in the future, just by assuming that well behavior hasn’t changed much. Furthermore, I wanted to do so in a transparent manner, and state all assumptions (which readers of course can disagree with). I offer many tools here so that anybody can make up their own mind about this.

          As written in the post, there are a few reasons why these projections will likely be too low. Also, if you visually inspect the projections in the first 2 overviews, and compare them with the historical data, my impression is also that they appear a little too pessimistic.

          This is confirmed by my quick comparison of the results half a year ago, and the results now. If the level of under-prediction stays constant, maybe it is reasonable to expect that Jan 2018 production from wells prior to 2016 will actually come in at about 1.6 million bo/d?

          Based on this, I think it is reasonable to expect that some of the new EURs projected here (2015/2016 vintages) may turn out 15-30% higher. That is the order of magnitude in projection error that I would still be happy with. However, I think it is unlikely that the projections shown here are more than 50% off, if we ignore the effect from extra capital expenditures such as re-completions and EOR. Otherwise, I’m willing to consider a bet!:-) (especially if a cut-off date is used, not too far in the future)

  3. The montage below is from a 2014 Apache Wolfcamp Simulation using Low, Medium and High GOR cases. It is a 7,000′ horizontal well in a 70′ thick one microdarcy rock. It has about 100 acres of drainage area.

    The well is completed with 23 stages spaced at 300 feet. Each stage has 4 perf clusters and induced fracs. So there are 92 total 350′ long fracs along this 7,000′ idealized wellbore, each with 20 md-ft of frac conductivity.

    The Apache engineer doing this is really smart- the only thing he varies is the PVT properties of the oil to show you the impact of that (something you have absolutely no control over) on recovery behavior. The PVT properties are based on historical measured Wolfcamp data. The initial solution GOR varies from 667 to 1230 scf/STB.
    In the left figure, you see long term oil rate and cumulative oil. In the low GOR case, about 40% of the oil is recovered in year 1 but in the other cases over 50% of the EUR is recovered in year 1.

    The right figure is the early time (500 day) behavior. The low GOR case which will actually end up with a higher EUR is clearly the worse performing in the first 500 days (see Enno’s Bakken early wells outperform the later wells with time).

    I show these theoretical graphs to illustrate that something as simple and uncontrollable as a doubling of initial solution GOR changes predicted behavior dramatically.

    My opinion is that any engineer who thinks he can predict what these wells will ultimately produce on the basis of a year of rate data is BS-ing you, and Enno’s comprehensive statistical method is as good as any at this point. The weakness of his method in Texas is the multi-well lease problem which does not exist in the Bakken or New Mexico.

    It is like a seismic shoot and Enno’s is 10,000 fold.


    1. Jim,

      Great comment as usual; thanks for sharing this analysis here, and for providing further support of these projections.

Leave a Reply

Your email address will not be published. Required fields are marked *