Disclaimer: in a long note like this and with so many numbers, and no triple-checking and no
NYT fact-checker, there will be factual and typographical errors.
Updates
July 2, 2018: with regard to DUCs, my thesis is that shale operators are using DUCs to manage their assets. Someone suggested the reason there are so many DUCs is because of limited takeaway capacity. I was thinking the same thing. Then I read
this from last week (the article is a keeper; it's been archived).
“The indicator that I typically watch, the Brent Oil price minus the WTI
price, that spread was just over $10,” Kringstad said. “When that price
is higher than $5 per barrel, historically the industry tends to pull
barrels off the pipeline systems and put them on the rail network to
take advantage of that market situation.”
CBR: scalable.
Number to remember: any number higher than $5, Brent/WTI spread. If Brent is $10 higher than WTI, then it makes sense to send Bakken oil to east coast where they buy Brent, instead of to Cushing where they buy WTI.
And more from the article:
Rail transport hit a low last summer of around 120,000 barrels per day,
Kringstad said, but was back up to 260,000 barrels per day in March as
market conditions shifted to favor east and west coast crude by rail
markets.
Some of that shift is a result of oversupply in Cushing, Oklahoma,
Kringstad said. Permian oil is congesting that market hub, lowering
prices. Oil that can go to east and west coast refineries is thus
finding better prices there.
Pipeline capacity, though pushing the upper limits, is still adequate (barely?):
Long-term forecasts for North
Dakota oil production by the mid-2030s is for between 2 and 2.4 million
barrels of oil per day, about double of what it is now, Kringstad said.
“When
we look at the takeaway capacity options, it’s right around 1.4 million
bpd with what is now in service or slated to come online in the next
several years, so there is potentially a 600,000 to one million barrel
per day potential shortfall long term,” he said.
Original Post
Update regarding DUCs.
Bottom lines:
- Bakken wells are most productive after initial frack (and possibly after subsequent fracks); completing one or two more DUCs in any given month can literally mean the difference between setting a new month-over-month production record and not setting a record
- there was a temporary increase in the number of DUCs during the Saudi Surge (2015 - 2016)
- considering the price of oil over the past three months, I consider the number of DUCs (now over 900) to be more than expected
- I know there are folks who disagree with me
- there are articles in the mainstream press suggesting ND is simply working down the backlog that occurred during the Saudi Surge; this may be the best article (a Bloomberg article) suggesting ND is "working down that backlog"
- the second and third graphs below do not support that argument (although that's in the eyes of the beholder)
- if ND was "working down the backlog" it does not explain the upward trend for the past three months
- anecdotally, recently there has been a long stretch in which none to few DUCs have been reported as completed
- most analysts I have come across, who are much, much more knowledgeable than I am, do not consider DUCs to be an issue worth spending much time on
Bottom, bottom line: with regard to DUCs, there's something going on that isn't explained in the data. Time will tell. As it always does.
Note: the number of DUCs (drilled to depth but waiting for completion) and inactive well numbers (AB and IA) are all estimates, although the number of DUCs should be a fairly accurate number. In addition, it depends on the exact day that the data is collected.
I will post graphics first to bring us up to 2016 or so. These graphics were taken from various sources using a google search.
First, these three graphics.
From January, 2014 -August, 2015:
- range: slightly below 600 to a high of just under a 1,000; trend was moving up through all of 2015; this was during the Saudi Surge; the trillion-dollar mistake Saudi made trying to break the US shale operators
From January, 2014 - October 16, an overlap with above data, but takes us to October, 2016:
- it makes no sense to me why EIA numbers are lower than NDIC numbers, when the former includes Montana and North Dakota, whereas the latter includes just North Dakota
- range: slightly below 600 on a couple of occasions in 2014, before the Saudi Surge; to a high of almost 1,100; trend was
moving up through all of 2015; this was during the Saudi Surge; the
trillion-dollar mistake Saudi made trying to break the US shale
operators; after September, 2015, there was a slight but definite downward trend, to about 900 in August, 2016
From December, 2016 - April, 2017, an overlap with above data, but takes us to April, 2017:
- December, 2016: 790
- April, 2017: 821
From most recent data, May, 2018 - May, 2017, most recent to oldest, which brings us up to date:
- taken from my posts on the blog. In May, 2017, I started posting DUCs and inactive wells on a regular basis. It's too time consuming to go back and fill in the few months in
which I did not post data so I'm not going to fill in that bit of
missing data. I think the existing data tells the story.
- data below: month/year, DUCs, inactive wells. All data below, including the up/down change were taken directly from the NDIC Director's Cuts
- range:
- an outlier, March, 2017: 689
- otherwise, a low of 799 in February, 2017
- a high of 955 in the most recent month, May, 2018
- not graphed, but the trend for the past four months has been up
- up 48 in February, 2018, 901
- up 15 in March, 2018, 916
- up 26 in April, 2018, 942
- up 13 in May, 2018, 955
Inactive wells:
- in data presented, we have never gone above 1,700 inactive wells
- number of inactive wells in the past three months are among the highest numbers in the data presented
The raw data:
September, 2018:
928, up 3 from previous report
inactive:
1,340, down 87
August, 2018:
925, down 18 from previous report
inactive:
1,427, down 59
July, 2018:
943, down 50 from previous report
inactive:
1,486, up 28
June, 2018:
993, up 38 from previous report
inactive:
1,458, down 111
May, 2018:
955, up 13 from previous report
inactive:
1,569, down 48
April, 2018:
942, up 26 from previous report
inactive:
1,521, down 132
March, 2018:
916, up 15
inactive:
1,653, down 1
February, 2018:
901, up 48
inactive:
1,654, down 100
January, 2018:
853, down 24
inactive:
1,554, up 85
December, 2017:
877, down 6
inactive:
1,469, down 23
November, 2017:
October, 2017:
September, 2017:
853, down 10
inactive:
1,444, down 54
August, 2017:
863, down 26
inactive:
1,498, up 20
July, 2017:
889, up 34
inactive:
1,478
June, 2017:
865, up 35
inactive:
1,458, down 53
May, 2017:
830, unchanged
inactive:
1,511, up 45
April, 2017:
830, up 141
inactive,
1,466, up 167
March, 2017:
689, down 110
inactive:
1,299, down 312 (need to correct typo at that post)
February, 2017:
799, down 3
inactive:
1,611, down 67
January, 2017:
802, down 5
inactive:
1,678, up 105
December, 2017:
807, down 32
November, 2017:
839, down 21
inactive:
1,519, up 16
October, 2017:
860, unchanged
inactive:
1,500 (unchanged (AB and IA)
September, 2017:
August, 2017:
888, down 24
inactive:
1,514, up 44
July, 2017:
912, up 25
inactive:
1,486, down 98
June, 2017:
887
inactive:
1,584
May, 2017:
931, up 39
inactive:
1,584, down 6
Disclaimer: in a long note like this and with so many numbers, and no triple-checking and no
NYT fact-checker, there will be factual and typographical errors.
