Well performance prediction using DT-1's
This discussion/question is sort of related to the "Production flow data from early HS wells" discussion, which was more about production and production decline than pressure issues, but pressure issues were mentioned. So I have been trying to make the most out of available data, and it looks like the two best sources for telling what a well is currently doing, and will do are 1) the actual production data, until such point that there is more than 1 well in a unit, and 2) the DT-1 tests. I have sort of deciphered the DT-1 tests, assuming that "GAS DEL" is basically MCF/D production under test conditions, for the given CHOKE (which I would guess is the current production choke), that WATER is barrels of water a day, which I am presuming is most desirably around 0, but not uncommonly around 100, and finally FLOWING PRESSURE. Being naive, I would think that FP x CHOKE**2 would be proportional to daily production; this does not seem to be the case, but it does correlate somewhat. SO if the FP is really low, I am presuming that is bad, and at some point they start compressing the well (which costs money), and when it gets truly abysmal, the well maintnenance guys go hunting around for some sacks of concrete and P&A the poor sucker. So I have a well, reasonable first check given the IP was about 8MMCF/d (produced 160MMCF for first check); there clearly are some challenges over in the T12N-R9W and T12N-R8W area; with wells varying in IP between 3 MMCF/d and 25 MMCF/d (so in these two townships, there have been 12 completed wells; 4 between 3-10 MMCF/d IP, 3 between 10-20 MMCF/d IP, and 3 between 20-30 MMCF/d IP (more data available no doubt; this is just what is on sonris). I need to check back and see if some of these might have actually tapped the BS instead of the HS - that could make sense, but otherwise, it looks highly variable in this region, but trending down rather fast W to E (I am right on the W to E township line. So I got sidetracked, but this is something else to comment on / think about - variability between near wells - is the shale that variable, or is it simply random events in drilling, or what? But back to the main question - FP is I am presuming pretty useful, though actual gas delivered in a DT-1 is probably the MOST useful number ;-) But what does low FP really tell you? This "well of interest" was delivering over 4 MMCF/d (not great I know, but better than some...) on a DT-1 after a bit more than a month; the FP had already dropped to 1500. I can speculate on several things that would cause FP to drop (less gas, fines plugging the fracture cracks, water still slowing things down maybe); but does this prognosticate anything in particular? Seems that the actual gas production is above typical gas production for this sort of FP. Also, when do they move to compression to keep gas flowing? I noticed a compression fee on my first check - largest deduction that was made; I am presuming this is even further up-pipe, at the point where flow from many wells is going into a pipe, but of course don't really know. So apologies in advance for the all-over-the-place question, but I would appreciate clarification of any misconceptions and addition of any insights about reading this sonris data as it is starting to actually accumulate. Ultimately, I know that what arrives in the mailbox is the best indicator, but I am sure that I am not the only mineral owner out there plowing around in sonris, trying to predict what is in store for us, especially closer to the edges.
Views: 270
Replies to This Discussion
-
Permalink Reply by Les Bamburg on -
Robert, all the information I utilize is sourced from Sonris and then includes a certain amount of interpretation. Depth is not really a critical factor in variations in well performance. As you say we know little about the changes in porosity, clay content, etc as you move across the play.
Each company's structural map is highly interpretative (especially in the southeast edge) due to the limited data available. These maps can be as much art as science in some areas. Early information indicates their map is completely off in the Ashland Field area.
Regarding the Mid-Bossier Shale play, there has been clear indication if the T12N-R9W/R8W is prospective or not. To date the only Mid-Bossier Shale completions in Red River Parish were located in T13N-R9W & T13N-R10W.
-
Permalink Reply by Les Bamburg on
-
Robert, now for Item #1. Flowing casing pressure is just a parameter, together with flow rate and choke size, that is used to compare the initial performance of various Haynesville Shale wells on an apples to apples basis. Of course performance is impacted by many factors such as reservoir quality, lateral length, formation pressure, facture treatment design & execution, lateral location vertically, etc and it is difficult to fully assess variations without being in the operator's shoes. The well rate would be higher and the flowing pressure lower with a larger choke size for a given well. Of course the timing for the reported initial well test can impact since a well may be still "cleaning up" or already starting to decline when the official test is taken.
As a HS well produces and declines both the flow rate and flowing pressure decline. On typical choke management plan the flowing pressure would drop to line pressure (~ 1000 psi) in 1.5 to 2 years. More restrictive choke plans may allow a well to produce 3-4 years before reaching line pressure. Most gas gathering networks have 2-3 pressure systems to a well can be switched from high pressure to intermediate or low pressure one it reaches line pressure.
-
Permalink Reply by Robert Duke on
-
Hi Les,
Okay, sort of as expected, but here's the more detailed question. You have a well (well, I do) that has had the first FP turn up as 1500# on a 22/64 choke, about 35 days after an IP completion of 8 MMCF/d on a 20/64 choke. Is this well likely running into problems that will limit its EUR to below what one would expect for an 8 MMCF/d well? That is my guess, but the thing has been burping out somewhat reasonable gas in the first two months, based on checks and reports about MMCF/d around Nov 1st. This very low FP so early in the game seems likely to be problematic to me, but maybe the well is still cleaning up? I would expect the character of the shale - how it fracked, what kind of fines (plugging particles) it produced most influence both FP decline and well longevity.
-
Permalink Reply by Les Bamburg on
-
Robert, in absence of full information it is difficult to properly assess. Assuming the DT-1 report is correct and the flowing casing pressure has dropped to 1500 psi that would be a concern.
The "fines" issue has not been confirmed or documented by any of the operators so is more conjecture at this point.
-
Permalink Reply by FXEF on
-
Maybe the well is liquid loading. Most likely the operator is on top of this, because liquid loading can kill a well.
-
Permalink Reply by Robert Duke on
-
Hi Les, FXEF,
Thanks for useful comments. The "fines" issue is an extrapolation from my laboratory background, where one can observe problems with, eg., filter columns plugging up due to junk clogging the pores in the medium; I would think it has to happen, may be a reason clay is more of a factor, etc. But I also understand that at the types of pressures we are working with down around 12000 ft, bench intuition may not be exactly on target ;-) And fractured shale is a medium I probably don't understand. I wonder how much "benchtop" work, if any, has been done with shale in the laboratory, but at appropriate pressures and temperatures? On water loading, I have inferred that Les thinks that in shale, if you have a water-loading problem, then you wandered outside the shale somewhere into another material. I would love it if this is a controllable water-loading problem, caused by it being harder to stay strictly in the shale. Certainly, the wells in this area have had higher water production values than on average, and they don't seem to go bone-dry in later DT-1's - I need to get good at using sonris "classic" so I can more intelligently query the existing data. The other wells right around here, to the extent that there is production data and FP data, look worse than 240843 so far; I am hoping it is more a function of "getting things right" (I think 204843 is the latest well) than that the shale goes from great to hopeless in 3 miles; I suppose the next 5 wells will give us more data. Thanks again to you all for your input.
-
Permalink Reply by Tiredpharm on
-
Is the company doing the drilling a factor; i.e, do Petrohawk and Devon wells tend to turn out better on the average than Chesapeake or XTO? (Just randomly assigning specific names - no preconception intended!).
Probably still feeling the effects of eating too much yesterday, but it might be interesting to prospectively test one's hypothesis by using the model to predict the outcome of a well, where drilling had just started, and then see how close the prediction was when the actual production data came in a few months later. Could even make, for example, 3 or 4 "best case" scenarios, each one with a different mix of variables as the predominant influence on the final result, and see if one of the approaches gave a production estimate more consistent with measured outcome. Could even do this on a group of wells already producing, if one could "blind" themselves and not be influenced by already knowing the correct answer.
The more uncontrolled variables there are in a calculation, the greater the population needed to formulate a model with any statistical validity. Sounds like there are a lot of unknowns and guesses here, and it might take data from thousands of wells to say, "there's a 90% probability that this particular well being drilled will come in at 22 Mmcf/d." But if one did derive such a formula with that level of accuracy, they could sell it! (There is a 90% chance that the money from the sale would be equivalent to a 1 Tmcf/day well. :)
-
Permalink Reply by Robert Duke on
-
Encana is drilling all of the wells that have completed on this line running NE to SW, and my general impression is that Encana is just great all-around. I am undoubtedly attempting to overanalyze the data available on sonris in this area, as I have a financial interest, and overanalysis is an occupational hazzard for me anyway ;-) My gut feel is that there is a heck of a lot more art in doing this drilling stuff than cold hard science, so that makes things a bit harder to predict; despite my natural tendencies, I am of the "too many unknown variables" mindset on being able to really come up with a general equation. But overall, over a lot of the play, the wells have seemed pretty darn consistent (if you call factor of 2 performance differential consistent) to me, based on nonscientific sonris browsing. I am mostly puzzled by why the wells in this area drop from 20+ MMCF/d to 3-8 MMCF/d in something like 3 miles - seems pretty abrupt, and it is not explained by the shale depth contours in this area (or at least that seems the case to me). No doubt, all sorts of things could explain it, both in terms of what is actually down there, and in terms of the driller perhaps not knowing how to get optimal production in this area; there are at least 5 more wells in progress (drilling or completion) right on this NE to SW trend, so there is a heck of a lot of money being invested in getting better than 0.5 BCF EUR out of these wells so far!
-
Permalink Reply by Tiredpharm on
-
That's an interesting observation. Almost an order of magnitude difference in production only 3 miles away is a little worrisome from the view of a mineral owner. If a producer drilled 4 holes in an area and got three 25 MMCF/d wells and one 3 MMCF/d well, he would still average almost 20 MMCF/d for the four wells. However, from the view of a mineral owner with only one of these wells being drilled on his tract, he has a 25% chance of missing it. That would be an even greater concern if someone were considering buying the mineral rights and trying to figure how long it would take to get an investment return -- might get it back in 3 years or it might be 30 years.
-
Permalink Reply by Robert Duke on
-
Yes, indeed. Trying to do financial planning based on info coming out of the field, combined with the rather volatile nature of natural gas prices, is, shall we say, a challenge... So there appears to be a more predictable core, though there is a sprinkling of low performers scattered in there. Where the core boundaries are, and how rapidly things drop, and whether current tech can be applied to get sufficient EUR's with a bit of practice on the edges - who knows. It is a risk for the driller too, though, in that he is trying to make leasing and drilling decisions, and every hole is at least a $9 million gamble.
-
Permalink Reply by Greedy Landowner on
-
I don't think that someone can reach the conclusion you are looking for based on location, tvd and production alone. Operators have different views on how much they want to produce at current prices and sometimes completion issues alter production
-
Permalink Reply by Robert Duke on
-
Yes, the odd well drilling / completion disaster. Here, of the 4 wells drilled so far, one had problems in the horizontal and they had to back up, plug, and do-over. No apparent other problems from logs.
- ‹ Previous
- 1
- 2
- Next ›
Support GoHaynesvilleShale.com
Top Content
Blog Posts
The Lithium Connection to Shale Drilling
Shale drilling and lithium extraction are seemingly distinct activities, but there is a growing connection between the two as the world moves towards cleaner energy solutions. While shale drilling primarily targets…
ContinuePosted by Keith Mauck (Site Publisher) on November 20, 2024 at 12:40
Not a member? Get our email.
Groups
-
Franklin County, Texas
9 members
-
Cass County, TX
120 members
-
Rusk County
97 members
-
Livingston Parish, LA
34 members
-
San Augustine, TX
386 members
-
Miller County, Arkansas
27 members
-
Sabine Parish, LA
455 members
-
South Caddo Parish Group
440 members
-
Natchitoches Parish
400 members
-
Bienville Parish, LA
244 members
© 2024 Created by Keith Mauck (Site Publisher).
Powered by
