Saturday, June 15, 2019

Beating A Dead Horse -- June 15, 2019

Updates

Later, 9:09 p.m. Central Time: a reader provides an excellent reply to the original post -- I'll bring it up here for each browsing/search capability:
Rigs matter eventually. Even if they have gotten better. Even if there is some DUC to work off.

I calculate we need about 40 rigs to maintain production.

DUCs are fine, but they are inventory. If you use them up, it's like draining your savings account. If can float you for a while. But eventually, you need to be bringing in as much income as you spend. So, when you look at long term sustainability, I would just ignore DUCs.

Looking at EIA Drilling Productivity Report for Bakken, I calculate about 40 rigs needed to maintain production near current (1.4 MM bopd) level.

The way to do the calculation is explained by EIA and is pretty simple:

Rigs*(new oil/rig) + base decline = 0

Looking at the EIA charts:
https://www.eia.gov/petroleum/drilling/pdf/bakken.pdf

First panel (upper right) shows the new oil/rig is ~1400 bopd/month/rig. Next two panels directly below show the base decline is running about 67,000 bopd/month.

Doing the high school algebra and solving for rigs as the unknown:

rigs = 67,000/1400 = 47.8 rigs

Note, however, that when you lose rigs, you tend to pull the LESS capable rigs first (mechanically, crew experience, and more importantly rigs running in the less desirable areas like Divide, etc.). So there is always an effect of rigs getting more efficient as the numbers get lower.

It's like a class's average height getting larger, when you eliminate the shorter third of the students. You didn't actually make any students taller. But the average got higher because of cutting in a non-random manner.

So given all that, it would be something less than 48. I'm saying 40 as a guess, but it might be slightly higher or lower. Don't shoot me if it ends up being 45 or 35.

There is also another effect that takes place during downturns or plateaus. That is that the decline rate gets gentler as the percentage of new wells drops. (This happens during downturns or flat-turns, since less completions are done per month.) This has no immediate effect the first month of a big downturn, but over a year or two it can be substantial...it can lead to "less Red Queen" which means less rigs are needed just to fight shale decline. [The Rystad presentations on the video show this very nicely for the Permian...it's base decline went down from 2014 to 2016.] This also occurred in the Haynesville where 30 rigs were holding production flat (down from 200 at the peak, during growth).

Other effects that can happen over time include technology improvement (practice, efficiency, completion improvement, knowing where better land is), but they also fight depletion of the best rocks (look at EOG using up the Parshall). In any case, I think both of these effects are more long term and tend to offset each other.

Net/net: sure, dropping a rig or two is no big deal. But if we get into the 40s, we will be hurting for growth and in the 30s probably shrink. 
Original Post 

When rig counts drop in North Dakota later this year, that will be a headline story for some. But rig counts are irrelevant.

From Lynn Helms on the April, 2019, Director's Cut, see video here:

Natural gas capture:
  • new all-time record
  • oil production stayed flat; even so, natural gas production went up almost 1% 
  • even more interesting, natural gas capture increased -- not so much as a percentage change as a raw number change
  • due more to large temperature swings in North Dakota than any other factor --this is fairly unique -- not something seen in the rest of the world -- the wide seasonal changes that affect natural gas gathering seen in North Dakota (-40 degrees in the winter to +100 degrees in the summer)
  • summer infrastructure season has not yet begun; that is not the reason for increase natural gas gathering
Producing wells:
  • all-time high in number of producing wells; reduced the DUCs a little bit and reduce the inactive well count even more 
  • inactive wells being put back on stream
  • permitting: pretty much steady
Price of oil:
  • price of oil in June: problematic; for March, April, and May, above forecast
  • state revenue forecast: results above forecasts in March, April, and, May by as much as $8/bbl
  • June: revenue only 15 cents/bbl above forecasts
Operators:
  • Helms visited a dozen operators in Denver last month
  • among the top 25 operators in North Dakota, half are located in Denver
  • Helms spent about an hour with each one
  • operators are very concerned about WTI price and over-supply
  • but, despite that concern, will stay with about the same number of rigs/frack crews through 2020: 
  • rigs: 20% improvement in drilling efficiency in last year; one operator will reduce number of rigs from five to four
    • each rig: three wells per month (30 days/3 wells = 10 days from moving, rig set-up, drilling to depth, moving rig off site;
  • 30% improvement in frack efficiency; that same operator will reduce frack crews from three to two
Rig counts -- beating a dead horse:
  • The weekly change in the number of rigs in North Dakota is a relatively meaningless metric. Wouldn't it be great to maintain production of 1 million bopd with one rig vs 218 rigs? Wouldn't it be great to be able to put in pipeline more efficiently? More meaningful metrics:
    • monthly production 
    • price of oil
    • takeaway capacity
    • number of DUCs
    • number of inactive wells

8 comments:

  1. Rigs matter eventually. Even if they have gotten better. Even if there is some DUC to work off.

    I calculate we need about 40 rigs to maintain production.

    DUCs are fine, but they are inventory. If you use them up, it's like draining your savings account. If can float you for a while. But eventually, you need to be bringing in as much income as you spend. So, when you look at long term sustainability, I would just ignore DUCs.

    Looking at EIA Drilling Productivity Report for Bakken, I calculate about 40 rigs needed to maintain production near current (1.4 MM bopd) level.

    The way to do the calculation is explained by EIA and is pretty simple:

    Rigs*(new oil/rig) + base decline = 0

    Looking at the EIA charts:

    https://www.eia.gov/petroleum/drilling/pdf/bakken.pdf

    First panel (upper right) shows the new oil/rig is ~1400 bopd/month/rig. Next two panels directly below show the base decline is running about 67,000 bopd/month.

    Doing the high school algebra and solving for rigs as the unknown:

    rigs = 67,000/1400 = 47.8 rigs

    Note, however, that when you lose rigs, you tend to pull the LESS capable rigs first (mechanically, crew experience, and more importantly rigs running in the less desirable areas like Divide, etc.). So there is always an effect of rigs getting more efficient as the numbers get lower.

    It's like a class's average height getting larger, when you eliminate the shorter third of the students. You didn't actually make any students taller. But the average got higher because of cutting in a non-random manner.

    So given all that, it would be something less than 48. I'm saying 40 as a guess, but it might be slightly higher or lower. Don't shoot me if it ends up being 45 or 35.

    There is also another effect that takes place during downturns or plateaus. That is that the decline rate gets gentler as the percentage of new wells drops. (This happens during downturns or flat-turns, since less completions are done per month.) This has no immediate effect the first month of a big downturn, but over a year or two it can be substantial...it can lead to "less Red Queen" which means less rigs are needed just to fight shale decline. [The Rystad presentations on the video show this very nicely for the Permian...it's base decline went down from 2014 to 2016.] This also occurred in the Haynesville where 30 rigs were holding production flat (down from 200 at the peak, during growth).

    Other effects that can happen over time include technology improvement (practice, efficiency, completion improvement, knowing where better land is), but they also fight depletion of the best rocks (look at EOG using up the Parshall). In any case, I think both of these effects are more long term and tend to offset each other.

    Net/net: sure, dropping a rig or two is no big deal. But if we get into the 40s, we will be hurting for growth and in the 30s probably shrink.

    ReplyDelete
    Replies
    1. Agree completely. That's why I try to include "don't take out of context" whenever I say "rig counts are meaningless."

      But this weekly obsession with a loss or gain of two or three rigs is insanity. We don't even get the percentage change, just the raw number. And the headline never tells us where the change in rig numbers is occurring: oil? natural gas? Bakken? Marcellus? Permian? Haynesville? That's pretty basic stuff and yet most folks don't go beyond the weekly rig count.

      Speaking of which: most folks don't even know the issue of heavy oil vs light oil.

      On another note, if we could afford a science experiment, it would be interesting to see what would happen if we stopped all new drilling and went back to:
      a) completing the DUCs (about a thousand)
      b) putting all inactive wells back into production (about two thousand)
      c) mini-re-fracks where such make most sense on wells drilled after 2014
      d) re-fracks of all Bakken wells drilled before 2014

      Delete
    2. 1. Agreed, week to week numbers are silly. Even during a trend up, there will be individual weeks down (and visa versa). It's like evaluating a baseball player. Impossible to do on a particular game, need more evaluation time. That said, I still watch it, since if you see week after week same direction, it matters.

      2. There are many reasons why there is some statristical noise in the weekly numbers but one is just related to how they measure number of rigs. Baker Hughes is very strict and only counts rigs that are "turning" (drilling). Rigs that are getting moved to another pad or being set up on a pad are not counted as rigs (considered same as a rig in storage). So just random differences in number of moves can make the weekly numbers go up one week and down the next.

      It's a little like crude imports, where an individual VLCC is big enough that if you have a week with a few more vessels or less it makes the numbers move. But it's really just random scheduling...not a systemic change in level--need a longer time horizon to filter out the week to week noise.

      Heck, even month to month ND production growth can vary a lot. Not even the seasonal differences (huge themselves), just random month to month differences.

      3. Rig count has been pretty stable in the Bakken, but nationally, it's clearly down and down noticeably. Not catastrophic. But still noticeable. Down to 788 from 885 at recent last peak in NOV. This is a natural reaction to price decline. It's still enough rigs to grow US. But 780 ain't 880. And we should expect slower growth than in 2018 (which was colossal).

      4. Agreed the headlines are silly...but didn't you see Broadcast News? Journalists are not rocket scientists. Just go to the weekly BH results (like you do with looking at EIA or NDIC news).

      https://bakerhughesrigcount.gcs-web.com/na-rig-count/

      The "rig count summary" pdf has most of the info you want (basin specifics, oil versus gas, etc.) in an easy to read format.

      If you want even more detail, look at the "North America Rotary Rig Count Pivot Table (Feb 2011 - Current)" which will tell you oil/gas differences in basins, maybe even counties...huge detail. Plus you can see trends.

      It's all public and free...

      5. Your a-d are awesome points. They all tend to cushion a bust and to increase rig efficiency (a temporarily, b-d systemically). Just nitpicking, but it is not possible to do c/d (or even a) simultaneous with b. Nearby wells have to be off line temporarily while fracking other ones.

      Delete
    3. Thank you. I will post tonight and get back to it tomorrow, if necessary. I'm just too "Bakken'ed out" to do much more with this tonight. I really appreciate you taking the time to add so much to the conversation.

      Delete
  2. Of course, the effects of increased rig efficiency and lowered base decline are reversed during a boom. So if we double rigs, we won't double new oil. And if we grow fast, base decline (Red Queen) gets tougher also. So sure, doubling rig count would make growth increase...but less than double. The same things that cushion the crashes, also make the treadmill tougher during a boom. Still, of course, more rigs would increase growth. Just not one for one.

    ReplyDelete
    Replies
    1. By the way, this also explains why US operators can withstand a short-term (6 to 18 months) crash in oil prices. Lots of options for shale operators.

      Shares in publicly held companies will plummet, but the operators in the Bakken should survive. I do think some operators in the Permian could have more difficulty simply because of the high price they paid to get into the Permian.

      Delete
    2. It sounds cruel, but I don't care about the operators per se. I follow them because there is so much good info if you follow them. But my interest is more in the US growth. So if WLL goes BK, I don't care. Rig count (over time, of course) is more meaningful. But if it is WLL or some PE company drilling the land, not so important to supply/demand, US economy. And the joy of pouring salt into peak oiler wounds...paging Art Berman!

      Delete
    3. Wow, wow, wow -- in a general way, I agree with you completely. This is not an investment site. I am more interested in the technology, the science, ... and I wrote that in the original posts when I first started blogging ten years ago. I follow the money because the money trail explains a lot of things. I follow the operators because they have lots of great information, as you noted. Interestingly, over time, I have developed a spot in my heart for several operators and for their founders. A lot of these guys have put their hearts and souls into discovering and developing the Bakken. I don't like to see any company fail; I like to see folks gainfully employed and not be worried about whether they will have a job next week.

      By the way, even if some of these companies reorganize, get bought out, merge, etc, the blue collar workers continue to find work in the Bakken.

      So much could be written, but I'm "Bakken'ed" out tonight.

      Delete

Note: Only a member of this blog may post a comment.