Does EIA Drilling Productivity Report Reflect Real World?

Last October, the Energy Information Administration (EIA) announced it would introduce a new publication that focused on the impact of drilling rig productivity on oil and gas output. This report was developed to assist the EIA in its forecasting of future oil production given that the American shale revolution has altered the historic relationship between active drilling rigs and oil and gas production. Drilling horizontal and hydraulically fracturing shale wells has changed the forecasting landscape. What the EIA concluded was that it could identify the additional barrels of crude oil and thousand cubic feet of natural gas that came from the addition of one average working rig in a basin. It also concluded that measuring this addition would enable the agency to forecast output two months into the future.

The logic of what the EIA is attempting to do makes sense, and reflects the need to capture technology innovations drillers are introducing. These innovations were allegedly responsible for the dramatic growth in shale gas output a few years ago while the rig count remained flat or declined. If more gas was produced while fewer rigs worked, there has to be an explanation tied to improved drilling performance. Capturing this dynamic has bedeviled the land drilling segment, which has struggled to understand exactly what types of new rigs to build, but more importantly, how many of them will be needed to produce the projected oil and gas output suggested by producers and industry forecasters.

We have dealt with this topic before and will continue to explore the nuances of technological change in the oilfield service industry and its impact on oil and gas output, not only because it is an interesting topic, but primarily because understanding the dynamic may enable us to anticipate changes in the underlying oil and gas industry relationship that could create serious economic and industry
upheavals.

We recently authored an article in the Musings discussing the impact of the extremely cold winter on oil output in the Bakken basin of North Dakota. When we examined the EIA’s March Drilling Productivity Report, we were startled by the chart showing a steady increase in the basin’s oil output and the accompanying forecast for future monthly increases. That shock sent us back to look at all the monthly charts for the Bakken’s output projected by the EIA. Every month showed a projected output increase. Given what we knew about the production output during the winter months obtained directly from the North Dakota Department of Mineral Resources (DMR) that showed a significant drop in production, and especially Bakken output, which is what dominates the state’s oil production, we wondered about the EIA’s forecasts.

To explore the subject, we plotted the monthly output estimated by the EIA along with its next month forecast. We then looked at what the starting point for production was the next month along with its projected output one month forward. We did that for each month from October 2013 through March 2014, which gave us an estimate of Bakken production for April 2014. Next, we went to the North Dakota DMR web site and got each of the month’s preliminary and revised monthly production statistics for the entire state and for the Bakken formation. All of monthly data is presented in Exhibit 2.

As the DMR data is reported with a lag of two months, the last preliminary monthly output figure for the Bakken formation is for January 2014 of 871,672 barrels per day (b/d). That estimate contrasts with the EIA’s January 2014 projected output of 1,025,000 b/d made in December 2013. That is a difference of 153,328 b/d, or 17.6% of the DMR estimate and 15% of the higher EIA estimate. If we measure the DMR Bakken production estimate against the EIA’s starting point in January, the difference is 15.2%.

What is a possible explanation for this wide a difference? One suggestion is that the Bakken formation extends into Montana, so by only looking at North Dakota’s output we are understating the basin’s total production. We obtained the monthly oil production figures for Montana for March 2012 through January 2014, as compiled by the Montana Board of Oil and Gas. For the month of January, Montana produced 70,030 b/d, but that figure is for the entire state’s oil output, not just for the Bakken. According to the state, the Bakken formation only impacts three of the 100 counties of the state. Moreover, as of March 21, 2014, there were only seven drilling rigs working in Montana versus 179 in North Dakota. Compared to a year earlier, there were 11 rigs in Montana and 174 in North Dakota. The significant point about Montana oil production is that it appears to have peaked last summer, although we don’t know how much of the production drop-off in December and January is weather-related. It is also interesting that the active Montana rig count also peaked last summer at 14 rigs. The chart in Exhibit 3 shows the trend in Montana’s monthly oil output.

Even if we assume that all of the Montana oil production is related to the Bakken, that still leaves the EIA forecast nearly 83,300 b/d higher than the data from the two state agencies responsible for monitoring the petroleum industry in their respective states. Despite the old joke that the EIA’s estimate miss is “good enough for government work,” we think that large a miss raises questions about the forecasting model. If we question the forecast for one shale basin in the EIA’s report, shouldn’t we question all the basin forecasts? The Bakken exercise was the easiest to undertake ‒- as it is not easy to secure the state production data for the other basins.

We plan to further investigate the relationship between active drilling rigs and oil and gas production in future Musings articles. The industry must understand the dynamics shaping the future land rig market, as it will impact onshore drilling contractors. It is also an important factor in understanding the business trends impacting the onshore oilfield service sector. In the end, understanding these drilling dynamics will help to define the possible future for the oil and gas producing companies.

RIGZONE

10 Comments on "Does EIA Drilling Productivity Report Reflect Real World?"

Davy, Hermann, MO on Wed, 2nd Apr 2014 11:14 am 



How about that big bodacious map of the Bakken to give readers a “feel good” about the extent of the oil deposits. What that maps does not say is the sweet spots are like in two counties in ND. I will leave this article to Rock and Short who are our resident petroleum experts. I just find the EIA overly optimistic with projections. We have seen the classic “excel goal seek” with the EIA in the past. Personally I think they are a sounding board for a government that is run by lobbyists and in this case by oil industry lobbyists.

How that sound DeeC. BTW DeeC here in Missouri when you get a nickname it is cause we like you. I am trying to look the other way with the US bashing and look towards your excellent environmental comments. Besides around ¼ what you say about the US and leadership is spot on.

rockman on Wed, 2nd Apr 2014 12:07 pm 



“Drilling horizontal and hydraulically fracturing shale wells has changed the forecasting landscape.” No it hasn’t. The boom in oil prices changed the forecasting landscape. Which is why no one was predicting a rise in Bakken production back in 2002 when oil was selling for less than $30/bbl even though we had the technology to hoz drill and frac wells just like we’re doing today.

And as Davy points out not only does that “feel good” not highlight the sweet and sour spots it also doesn’t show how heavily and irregularly the Bakken has been drilled already. Here are some details on that subject:

http://switchboard.nrdc.org/blogs/fbeinecke/interior_secretary_visits_frac.html

http://grandemotte.files.wordpress.com/2009/11/bakken-well-locations.jpg

Makati1 on Wed, 2nd Apr 2014 12:59 pm 



Does EIA Drilling Productivity Report Reflect Real World?

Does Santa Claus live at the North Pole?

Does the Easter Bunny exist?

How about any other fiction we tell our kids to warp their minds? (Nope! I refuse to mention religion … ooops! ^_^)

Nony on Wed, 2nd Apr 2014 2:19 pm 



The difference probably has to do with estimation versus waiting for data. For Texas RRC, you end up having to wait 2 years for solid data. ND, you wait for 1.5 months (some small adjustments after that). If their data is based on company interviews or some other leading indicator, then there’s value to it (after all…you are getting an early look). If it’s just the linear exterpolation (TX recently), then not so much. Although the rig counts and well counts are probably up to date.

That said, despite the sniping, I’m quite taken by what EIA is trying to do and appreciate analysis, not just compilation. And the presentation is awesome and helps us think about the dynamics. OK…back to the conspiracy theories about sheeple and TOD-killers and the like.

Davy, Hermann, MO on Wed, 2nd Apr 2014 2:34 pm 



Noo, that sounds warm and fuzzy. I love it…kisses and hugs… kumbaya

rockman on Wed, 2nd Apr 2014 3:11 pm 



Though I occasionally piss on some of the EIA reports they do provide some useful data. I just wish they would emphasize just how dependent their production projections are on prices. They assume a certain high price (typically the current) and assume the price will stay there in the future and thus the drilling/production increases will hold. That alone is not a given. I would be thrilled to see them make such forecasts using multiple price projections such as $50/bbl, $60/bbl, $70/bbl, etc. That should yield some very different production projections. But that’s going to take a lot more analysis then just projecting current trend lines.

Nony on Wed, 2nd Apr 2014 5:54 pm 



1. The drilling productivity stuff (subject of this article) is not at all predictive. It’s just analysis. Year over year differences. Productivity/rig. Breakout by major US play. Separation of new oil production from field ongoing production/decline.

2. Their predictive stuff (outlooks, not what this article is about) did include price assumptions and even different cases (how much price drives how much volume). But I agree with you, there should be even more and a more robust discussion of why they think what they think (in terms of the interaction of economics and geology and even political and foreign policy outlook, and cartels).

3. One analysis of theirs that has price in it, that I quite like:

http://www.eia.gov/forecasts/steo/images/Fig1.png

Look at the 95% confidence intervals (green). That’s NOT their prediction of price. It’s the market prediction (from options and such). There’s a really cool insight in there: a year from now, we might be anywhere from 60-130!

rockman on Wed, 2nd Apr 2014 8:02 pm 



It always amazes me that some folks think the futures market is actually a predictor of future oil prices. For every $ bet that oil will sell for $X/bbl there’s another $ betting that it won’t sell for that price. Obvious one of them will be correct. LOL.

For fun it might interesting to see what futures were 10…7…5…2 years ago and what they “forecast” for the price of oil today and compare it to what oil is actually selling for today.

Nony on Wed, 2nd Apr 2014 9:09 pm 



Hence the interesting feature of the 95% interval band. Not only can you see what the market’s guess is, but the level of uncertainty. Click on the link and consider the graph and what insights are extractable.

Henry Larry on Tue, 30th Apr 2024 12:40 am 



It is crucial to scrutinize forecasting models like EIA to ensure accuracy in understanding drilling dynamics and their impact on oil and gas production trends.
Home Cleaning in Newcastle Washington