Saturday, February 2, 2013

Idle Rambling On The Dreaded Bakken Decline Rate

Updates

February 3, 2012: Extremely important update. See comments below. It was pointed out that the BR Kummer well below is "takeaway constrained" and that's why the monthly production has not decreased over time.

That pretty much takes most of the wind out of my sail regarding thoughts on the Bakken decline rate but there may be some validity in some of what I wrote. Certainly the linked Oil & Gas Journal article is still relevant.

If the BR Kummer well is takeaway constrained, and that certainly appears accurate, a whole new story line opens up. Maybe more on that later. But if this well is producing 20,000 bbls/month and is takeaway constrained, it speaks volumes about some of the better sweet spots in the Bakken.

Original Post

A few weeks ago I mentioned in passing that operators might be getting a handle on the "dreaded Bakken decline rate."

One example I cited:
  • 22050, 2,806, BR, Kummer 41-30MBH, Blue Buttes, t5/12; cum 165K 12/12; 2-section spacing; completion report not seen (1/27/13); this particular well lies less than a thousand feet from the "Helis Grail." At the time of the original post, the data only through November; this time another month has been added, December, and still no decline. 

PoolDateDaysBBLS OilRunsBBLS WaterMCF ProdMCF SoldVent/Flare
BAKKEN12-20123125411254203311390803903244
BAKKEN11-201230228422281728923001529656355
BAKKEN10-201231259142600333333516434971193
BAKKEN9-2012302506225012365234151323981753
BAKKEN8-2012312572625693386935756329252831
BAKKEN7-2012312595925928372326356026356
BAKKEN6-20122077157576961259402544
BAKKEN5-2012276800675406700

In the process of looking for something else along the lines of decline rates, I happened to come across an old posting. I had forgotten I posted it, suggesting I was posting quickly that day and didn't have a chance to reflect on it.

I am surprised that I have not seen more in the chat rooms about this.

I have had no formal training in the oil and gas industry. I am a novice, and know relatively little, and what I "think" I know might be "way wrong." Readers remind me of my mistakes, misunderstandings, misreadings, and typos on a daily basis.

Be that as it may, I feel I have a very, very good understanding of the relationship of IPs and overpressurization, with or without natural gas.

I also have a "picture" of what a conventional pool of oil looks like and an unconventional reservoir, or tight oil, looks like. I have seen photographs of the shale from which "tight oil" is extracted, which helps. I have never seen a live video or a photograph of a conventional pool of oil; something tells me my "picture" of a conventional pool is not quite accurate, but probably not too far off the mark.

A thought experiment.
First: A one-liter bottle of Coca-Cola, unopened. Shake vigorously. Open. Let settle. Put in a very small-diameter straw (say, the size of a ball-point pen refill). Suck out the remaining cola. Conventional pool of oil.

Then: A two-liter bottle of Coca-Cola, previously opened and one liter of the liquid cola removed and replaced with flour. Let settle. Shake vigorously. Open. Settle. Put in a very small-diameter straw (say, the size of a ball-point pen refill). Suck out the liquid but leave the flour behind. Tight oil.
End of thought experiment.

Now, this "cut and paste" from an article in a trade journal written by folks who are trained, educated, and experienced in the oil and gas industry:
For continuous shale oil fields such as the North Dakota Bakken, the decline rate may not be as steep as those experienced in conventional reservoir oil fields. Upon well saturation of the development area with four wells/sq mile, E&P companies will continue to perform well refracs and drill infill wells as long as well economics are positive.
Read that again: "...the decline rate [in the Bakken] may not be as steep as those experienced in conventional reservoir oil fields."

Two thoughts come immediately to mind:
a) averaging
b) the long game
Taking them in order:

First, averaging. Go to the NDIC Basic Services module and go sequentially through the well files, starting with #1 through #22000.  I'll do the first 20 for you:
  • 1: DRY; 2: DRY; 3: DRY; 4: DRY; 5: DRY: 6: DRY; 7: DRY....okay, enough of that
Let's move ahead to the 10,000 series:
  • 10001: DRY, 10002: 0; 10003: 0; 10004: 0; 10005: 0; 10006: 0 ... well, not much better
Let's move ahead to the 14000 series:
  • 14001: 217K; 14002: 514K; 14003: DRY; 14004: 165K; 14005: 396K, 14006: PNC; 14007: PNC; 14008: PNC; 14009:  337K; 14010: 573K (mostly Red River wells from 1997 time frame)
Now the 16000 series:
  • 16000: 312K; 16001: 68K; 16002: 32K; 16003: PNC; 16004: DRY; 16005: PNC; 16006: PNC; 16007: 61K; 16008: 829K (Red River); 16009: PNC; 16010: 81K (mostly Madison wells from 2006)
So, early on: mostly dry holes; then in the early 80s, the 10000 series -- look at all the dry holes. Then the Red River wells from the 1997 time frame but after 15 years, very erratic -- some are still great wells; but many were PNCd.

The 16000 series, Madison wells were really erratic; some good ones, lots of poor one, and many dry or PNCd.

The Bakken wells are too new to get much meaningful data for total production to date, but a) there are "no" dry Bakken wells and a fair number (the majority?) of Bakken wells producing for four years have produced more than the majority of legacy wells producing more than 15 years.

It doesn't take many dry wells to ruin one's "average" production of wells from a given formation. The decline rate for a DRY well is meaningless, but one can do a thought experiment with decline rates for a group of wells if seen as one well.

So, that takes care of "averages" for the moment.

Now, the second point, the long game. Maybe it is more important to look at the decline rate over the entire life of a well rather than just the first few years.

Again, the cut and paste from the trade journal:
For continuous shale oil fields such as the North Dakota Bakken, the decline rate may not be as steep as those experienced in conventional reservoir oil fields. Upon well saturation of the development area with four wells/sq mile, E&P companies will continue to perform well refracs and drill infill wells as long as well economics are positive.
Averaging, and the long game. 

In that "cut and paste" there are two interesting points. The first one: re-fracs. Self-explanatory. The second one, sort of hidden, but this is it: "... upon well saturation of the development area with four wells/section, companies will ... drill infill wells as long as well economics are positive.

With conventional reservoirs, once that first well has been drilled and the development area saturated with wells, that's pretty much the end. However, in an unconventional reservoir, the writer of the linked article suggests that even when a development area is saturated with wells (four wells/section), operators will drill more infill wells as long as the economics are positive.  (By the way, "we" are now up to 14 wells in a spacing unit, and infill drilling really has not begun. Re-fracking on a routine basis certainly has not begun.)

Go back to the thought experiments involving the Coca-Cola bottles. Expand the thought experiment. After the first liter bottle is pretty much empty, lay it on its side and insert ten more straws.

Now, do the same with the second bottle. At any point, turn it on its side and insert ten more straws.

I apologize for mixing apples and oranges in this discussion regarding Bakken decline rates. My hunch is that most folks were not even aware of that. I started out talking about operators getting a handle on the Bakken decline rate and providing an example of a well with just eight months of production (the short game), and ended with talking about the decline rate as discussed in the linked article (the long game).

Even if I am way wrong on my narrative, and/or if I don't make sense, and/or if the thought experiments have flaws, ignore what I've written and do the following (go to the source).
  • read the linked article from the trade journal;
  • follow the monthly production of BR's Kummer well and the newer BR and EOG wells coming on-line

8 comments:

  1. Infill drilling is standard. If drilling a well between other wells nurses out enough to make a good profit, it will happen. If not, it won't.

    The formation doesn't matter.

    Profit does.

    The same with refracks.

    Some conventional verticle wells are 2acre spacing.

    Gulf of Mexico wells often decline very fast. Oil comes fast with 25 % porosity. Think Slurpy.

    Anon 1

    ReplyDelete
    Replies
    1. Agree. I don't know enough about the oil and gas industry to comment one way or the other. I just find it interesting that at least one source suggests we may be looking at the "dreaded Bakken decline rate" the wrong way. Or at least that's how I understood the Oil & Gas Journal article that I linked.

      I am an eternal optimist. A lot has been written by folks wringing their hands over the Bakken decline rate. I've never worried about the decline rate, at least not to the extent that others seemed to be concerned. I was more interested in rate of return / how fast the wells would pay for themselves; what the EURs would be; what the cumulative production would be at one year, three years, five years, etc.

      When Slawson pays $19,500/acre, I just get the feeling that some folks are too worried about the decline rate. For Slawson, and the other Bakken operators, the decline rate "is what it is" and they deal with it.

      The above post had two conjectures in it (if that's the right word): a) operators were getting a handle on the Bakken decline rate; and, b) the decline rate might not be as bad if measured "correctly."

      Only time will tell if either or both of these "conjectures" are accurate. For me, that's what makes following the Bakken story so interesting -- watching these events unfold.

      Two years ago I didn't even know what an infill well was.

      Delete
  2. That well is takeaway constrained. When that happens you will have pressures declining, but that isn't posted publicly. There is always some sort of thing declining. Think propane tank and regulator. Take the regulator off and you have a fast, quick decline at a constant rate until the tank equalizes with atmospheric pressure. If you have a regulator on, the production will be flat...

    ReplyDelete
    Replies
    1. Thank you! That explains a lot. That data point, of course, takes the wind out of the entire post, I suppose, but it helps me understand the Bakken.

      That data point - that the well used as an example is takeaway constrained -- is very important; thank you for taking time to comment.

      Delete
  3. Not takeaway constrained at the surface, at the subsurface. Tight reservoirs. We call this the weeping tile effect. For visualization a 1000bopd well is really bleeding 1 teaspoon/foot/minute per a 4000' lateral. Not including fraction wings. Huh. - Slawson wizard

    ReplyDelete
    Replies
    1. Counter-intuitive, but makes sense.

      I suppose one can suggest that no matter the reason for flat production month-over-month, if production remains flat over time, the decline rate is diminished.

      Delete
  4. I think you should edit out the PNC because the IP, decline rate and EUR of wells that were never drilled are not really relevant, especially when a permit costs about $100 or $0.08 per acre in a 1280 acre spacing.

    ReplyDelete
    Replies
    1. I agree completely; I just didn't want to start "throwing out" wells for whatever reason...some folks might think I'm cherry picking.

      Again, this is a not statistical analysis. I could be way off base. I'm simply rambling out loud, giving folks something to think about, talk about. I now a lot of folks think I have no clue; that's fine.

      The only point I was trying to make was this, I guess: early on the decline rate took "us" by surprise. But now, three, four, five, six years into the boom, "we" are all aware of the decline rate. From my perspective, "it is what it is" and I no longer "worry about it" as a negative.

      If anything, I see it as a "positive": we now know about it, and the operators are working to improve on it.

      I tend to ramble. I guess the decline rate for me is a lot like the discussion we used to have regarding IPs. Folks argued that a) IPs don't mean anything; and, b) some companies "hype" their IPs. But now that "we" recognize that IPs "are what they are" and that IPs are simply a single data point, "we" have sort of moved on to other discussions. I don't hear much about IPs any more.

      [Don't take this as arguing, or starting another thread: a permit is much more than $100. A lot of work went into getting that permit. Exploration, seismography, study, discussion, working up the permits, etc. I don't know whether it's more common to get the lease first, and then the permit, or vice versa, but a lease would have been additional cost. And, of course, while doing the exploration, seismography, etc., the operator was using resources on a permit that was eventually PNC when they could have been working on another site.]

      Delete