Productivity Per Rig -- Revisited -- April 3, 2018

Updates

April 5, 2018: In addition to all that is written below, this is another interesting item. Think about this when comparing "rig productivity" in the Bakken with "rig productivity" in the Permian. QEP, in its most recent corporate presentation, said they have found a way to leave neighboring wells on line while new wells are being drilled AND then completed. Meanwhile, in the Bakken, generally speaking, operators take wells off-line (sometimes for months) when a new well is being completed. This fact makes the delta between "rig effectiveness" in the Bakken compared to that in the Permian even more substantial, using BTU Analytics/EIA methodology. Imagine the increase in crude oil productivity if a well did not have to come off-line when a neighboring well was being fracked or re-fracked.

Having thought about all this again, there are a lot of unknowns regarding "rig productivity" comparisons.

Original Post 

In the Bakken, the two most interesting bits of "news" in the past ten days has been the news that QEP will exit the Bakken to focus on the Permian and the BTUAnalytics article regarding EIA's data on rig productivity.

At risk of being way out front of my headlights, some thoughts regarding rig productivity.

1. As I noted earlier, if the analysis is correct, and I bet it is, it is very, very troubling. If that's the methodology that's being used it suggests that like so many others, the EIA has not yet established new ways of thinking about tight / unconventional oil. If that's there methodology, they are using the same methodology for unconventional / tight / shale plays that has been used for decades to track conventional plays.

2. The methodology apparently being used does not take into account DUCs. It seems hard to believe that they would do this (without at least a footnote) but my hunch is that it is impossible to get accurate data unless analysts literally went through well by well. And even then, it probably would not be accurate. As it is, when the NDIC reports data every month, the number of completed wells is a preliminary figure that will be "finalized" the next month. The preliminary figure for completed wells is significantly less than the finalized figure (I just went through a year's worth of that data yesterday).

3. What does this mean? Using their methodology, initial production per active rig is far lower than what active rigs are really producing, once you remove DUCs from the equation. That's what BTUAnalytics says, and I agree.

4. As I wrote that last note, one could argue that the "whole issue" would be more accurate if one simply changed the data heading from "rig productivity" to field productivity or operator productivity, if that makes sense. It's sort of like Tesla's "ramp up." They may have 10,000 cars on the assembly line, but one only counts the cars that actually come off the assembly line completed (and delivered).

5. This is the bottom line for BTUAnalystics: operators are producing (or, better said, will be producing) much more oil than anyone is publicly forecasting. Not good news for Saudi Arabia.

6. That's all the report said. The report did not compare the Permian with the Bakken with the Eagle Ford as I was doing in my post the other day. My hunch is that my conclusions comparing the three fields was not wrong (and if it was, so what? I'm just trying to figure out the Bakken and probably only know 1% of all that is going on -- and BTUAnalystics was a huge piece of the puzzle).

7. But I digress. As I was saying, my hunch is that my conclusions comparing the three fields was not wrong. Although the actual numbers might be different, the story line is probably still accurate. The story line remains accurate if the percentage of DUCs across all fields is approximately the same. That's a huge assumption. The BTUAnalytics was released last July (2017) and the report was probably put together using data for several months prior to publication. In other words, the data is somewhat old (not terribly old) but things change quickly in the oil patch.

8. Back in mid-2017, anecdotally, it seemed that on any given day, 30 - 50% (sometimes more) of the wells coming off the confidential list in the Bakken were being reported as DUCs. It's hard to believe that the Permian operators were reporting, as a percentage, even more DUCs than 50%.

9. So, I guess it comes down to this:

• the methodology to determine rig productivity is wrong because the methodology does not include DUCs
• due to this methodology error, the potential for overall basin production is grossly underestimated (BTUAnalytics conclusion); very, very bad news for OPEC-Russia
• the disparity between the Permian and the Bakken is so wide, it's hard to believe that this is only due to DUCs -- but we will know more two years from now -- maybe

10. Having said all that, there is an even bigger problem. It's generally thought that once a well is drilled, operators maximize production. Nothing could be farther from the truth. Operators manage their overall assets on a monthly basis and they manage each well just as closely -- for any number of reasons, operators will increase/decrease individual well production remotely.

Example. I quit posting the data below because it did not appear anyone really cared, and it took a bit of time. But look at this, this is January, 2018, data, in the spreadsheet below.

The spreadsheet has 20+ oil fields listed in the far column to the left. The spreadsheet is "ranked" based on the percentage change in production month-over-month (column 10). The first two fields had a significant change in the number of producing wells, so ignore those two fields. Go to the third field, Clear Creek, not a particularly good field. Note that month-over-month the number of producing wells remained the same (76). I would bet that these were the very same wells from month-to-month (in other words, the operator in Clear Creek did not shut in five old wells and bring on five new welsl -- although that's possible). If these are the very same wells, month-to-month, look at the change in production.

• in December, 2017, production for this field: 109,385 bbls
• one month later, January, 2018, production in this field: 156,278 bbls, a 43% increase in production with the same producing wells

Although the increase was not as much, one sees the very same thing in the next four out of five fields.

The Bailey field, an incredibly good field, added six producing wells and yet the increase in production month-over-month was not that great. Why? For one reason, when they bring on a new well, the production is generally atrocious -- simply because some new wells have only been producing one full day when they are included in the monthly numbers.

Field	Dec 17 Prod	Dec Wells	Dec Oil / Well / Month	Percent Change Dec-over-Nov	Change in # of wells	Jan 2018 Prod	Jan Wells	Jan Oil / Well / Month	Percent Change Jan-over-Dec	Change in # of wells
Truax	415,877	190	2,189	16.80%	9	710,783	198	3,590	70.91%	8
East Fork	242,137	123	1,969	-16.75%	1	366,905	128	2,866	51.53%	5
Clear Creek	109,385	76	1,439	-6.01%	1	156,278	76	2,056	42.87%	0
Westberg	165,756	102	1,625	-4.31%	2	201,800	102	1,978	21.75%	0
Lost Bridge	292,199	73	4,003	-18.70%	0	352,464	73	4,828	20.62%	0
Beaver Lodge	131,958	79	1,670	3.74%	0	154,971	79	1,962	17.44%	0
Bailey	640,292	157	4,078	-2.80%	0	742,278	163	4,554	15.93%	6
West Capa	41,150	38	1,083	-3.31%	0	47,214	38	1,242	14.74%	0
Painted Woods	126,353	59	2,142	41.99%	0	142,986	60	2,383	13.16%	1
Little Knife	345,021	161	2,143	6.93%	7	390,388	162	2,410	13.15%	1
Antelope-Sanish	1,503,780	215	6,994	28.63%	12	1,693,577	218	7,769	12.62%	3
North Fork	458,644	73	6,283	16.51%	3	513,087	74	6,934	11.87%	1
Cow Creek	117,841	45	2,619	0.45%	0	130,760	45	2,906	10.96%	0
Sanish	1,300,074	624	2,083	10.94%	6	1,430,162	629	2,274	10.01%	5
Corral Creek	488,558	167	2,925	-9.59%	2	533,834	167	3,197	9.27%	0four o
Stockyard Creek	184,181	89	2,069	-24.87%	1	200,522	90	2,228	8.87%	1
Alkali Creek	621,626	166	3,745	20.37%	3	675,693	166	4,070	8.70%	0
Ross	67,793	74	916	-5.16%	0	72,288	74	977	6.63%	0
Banks	912,593	213	4,284	-9.93%	1	971,819	222	4,378	6.49%	9
Squaw Creek	170,189	47	3,621	-2.23%	0	178,847	47	3,805	5.09%	0

Bottom line: I have to sign off. My battery power is almost gone. I have to find a Tesla charging station to re-charge my MacBook Air.