外文翻譯--煤層氣完井一種世界的觀點（英文）

1、Coalbed methane completions: A world viewIan PalmerHiggs-Palmer Technologies, 10140 Arroyo Crest Drive NW, Albuquerque, NM, 87114, USAa b s t r a c t a r t i c l e i n f oArticle history:Received 1 July 2009Received in r

2、evised form 23 December 2009Accepted 24 December 2009Available online 7 January 2010Keywords:Coalbed methanePermeabilityWell completionsStimulationProductionCBM wells are cheap because they are usually shallow. Deeper we

3、lls tend to have lower coal permeabilityand productivity. Low permeability is the weak point (Achilles heel) for CBM ventures, and success willentail finding regions of enhanced permeability, utilizing horizontal wells w

4、ith the option of fracturing thesewells, and possibly in “creating” permeability by new-paradigm methods such as shear stimulation andbranch fracturing. This paper discusses the role of permeability in choosing CBM compl

5、etions, as well asexamples of those completions from around the world. The guidelines offered are new and practical(including perm bands and a decision tree), and should benefit any CBM completion strategy.© 2010 El

6、sevier B.V. All rights reserved.1. IntroductionWell completions and stimulations for coalbed methane (CBM)have generally been chosen by trial-and-error, because there has notbeen much consensus, unlike wells in conventio

7、nal sandstone. Agood example of this is the development of CBM well completionsand stimulations in Australia (Kubenk, 2006). However, this is anexpensive approach. Further, the role of permeability in CBM successand in C

8、BM completions has not been fully appreciated (Palmer,2008). Now, after over 30 years of CBM production in the USA, generalprinciples and decision trees have emerged to guide an operator inchoosing a well completion. Thi

9、s is the subject of this paper.The paper illustrates the central importance of permeability in thesuccess of CBM completions. Reviews have been published on dif-ferent CBM completions, but these are mainly focused on par

10、ticularbasins or plays (Holditch, 1990; Cramer, 1992; Palmer et al., 1993,1995; Robinson and Holditch, 1999). One paper has searched fortrends in CBM production (across basins) as a function of reservoirand completion pa

11、rameters (Palmer and Cameron, 2003). Althoughdiffering industry viewpoints are revealed in this paper, there is notmuch consensus.Geological predictions of permeability have not been discussed,because this is a large and

12、 (sometimes) controversial subject which isbeyond the scope of this paper. Further, predictions based on geologyare more qualitative than quantitative, whereas this paper focuses onquantitative measurement of permeabilit

13、y.The intention is not to be rigorous, because that would requirepresenting an analysis of 30 years of CBM data from the USA, as well asdata from other countries where CBM is now exploited. In place ofthis, the reader ma

14、y refer to the references above (all review articles),and particularly that of Palmer and Cameron (2003).Effective stress has an effect on coal permeability. However, inthis paper the principal message is to measure the

15、coal permeability ina prospective CBM play, and choose the well completion accordingly.The potential effect of stress and depth on permeability is not neededif one measures the permeability.After a background of CBM prod

16、uction worldwide, the paper dis-cusses the role of permeability in choosing CBM completions, as wellas examples of those completions. We define perm bands as ranges ofcoal permeability in which certain well completion ty

17、pes are pre-ferred. At another level, this approach leads to a decision tree whichincludes various completion options as a function of the coal per-meability. This approach is new and practical, and should benefit anyCBM

18、 completion strategy.2. Worldwide backgroundCoalbed methane (CBM) has become a resource of globalsignificance, with the emergence of active CBM plays in Canada,Australia, India, and China, for example (Palmer, 2008). At

19、the otherend of the spectrum, countries like England, France, Turkey, andColombia have much smaller resources, and so far have not been ableto get a CBM industry started, though they are actively trying. Theattraction is

20、 usually good gas content and coal thickness, but theroadblock is often poor permeability (less than 1 mD), and this iscritical for project economics.In the U.S., BP and ConocoPhillips are two of the biggest CBMproducers

21、 in the San Juan basin (see Fig. 1), where BP has over 1500wells, while ConocoPhillips has over 800 producing wells (Palmer,International Journal of Coal Geology 82 (2010) 184–195E-mail address: ian@higgs-palmer.com.0166

22、-5162/$ – see front matter © 2010 Elsevier B.V. All rights reserved.doi:10.1016/j.coal.2009.12.010Contents lists available at ScienceDirectInternational Journal of Coal Geologyjournal homepage: www.elsevier.com/loca

23、te/ijcoalgeoTo produce economic gas from low-perm coals requires a three-pronged strategy (Palmer, 2008):1. locate a sweetspot by using structural geology to predict where tofind one, or by drilling a lot of wells until

24、you encounter one,2. drill horizontal wells, and3. deploy novel stimulation techniques.An example of the first strategy is the Australian CBM play onthe Undulla Nose, which is the “fairway” of the Surat basin, withpermea

25、bilities up to 500 mD (Scott et al., 2004). The Undulla Nose is astructural high, and the coal measures are draped across the nose. Thisdraping caused cleats and fractures to develop during coalification.Strategies 2 and

26、 3 will be discussed below.An assignment of CBM completions based on permeability isshown in Fig. 2. We will elaborate more on this ranking shortly. Butfirst, to follow this perm-based completion strategy requires a full

27、permeability picture of the prospective reservoir. This meansperforming lots of permeability measurements both areally (～50measurements desirable across the CBM play) and within differentseams. This can be expensive, but

28、 is important to optimize well com-pletions. Too many operators have recognized as important (andmeasured) gas content and net coal thickness in deciding to develop aCBM play, but have under-estimated (or even ignored) t

29、he role ofpermeability. Sometimes a few token measurements have been madeof permeability, but nothing like a statistical representation of thewhole field. Permeability, varies a lot due to the inherent variabilityof clea

30、ts and joints (natural fractures), but the more permeabilitymeasurements, the less is the uncertainty. A good review of fracturesystems in coal is available (Laubach et al., 1998).Fig. 3 illustrates the results of such a

31、 comprehensive approach tomeasuring permeability, which has actually been done by at least onecompany, Reliance in India2. The obvious advantage of such a pictureis that we can see where the sweetspots and fairways are,

32、which iswhere we want to drill the first wells and show that we can make gas.At the other extreme, the picture tells us where the coals are tight(permeability b3mD), which requires special attention as describedlater. Wi

33、thout the full permeability picture, an operator may drill thefirst wells in tight coal, which will very likely not produce commercialgas, and the project may fail prematurely.Of course, creating permeability pictures li

34、ke Fig. 3 is easier if thereis one main seam. If there are several main seams, ideally we wouldcreate a permeability contour for each of the principal seams, whichof course will be even more expensive. To clarify this si

35、tuation, apermeability measurement in each seam of a vertical well, or at leastin each of the main seams, should be mandatory.How are permeability measurements best done? There are severalmethods, and injection fall-offt

36、ests (IFOTs), diagnostic fractureinjectiontests (DFITs), and drill-stem tests (DSTs) appear to be equally good.Since there have been some reports that DFITs over-estimate perme-ability (but the discrepancy depends on per

37、meability), it would be goodto first do a DFIT and an IFOT in the same seam of the same well tocompare them. Note that permeability from history matching of gas rateis not immediately comparable with well test measuremen

38、ts because(1) there will often be several seams contributing to the gas flow, and(2) other influential input parameters such as relative permeability areoften not well known.4. Perm-based completion strategyA simplified

39、completion strategy is provided in Fig. 4, which coversthe entire spectrum, and we will discuss each band separately.4.1. Perm band b3mD — multi-lateral wellsThis is the arena for multi-lateral wells, which can be trilat

40、eral(pitchfork), quadrilateral, or pinnate. Fig. 5 shows a pattern of pinnatewells (Spafford, 2007). The wells are designed for full coverage of thereservoir, to initiate gas flow faster, and to recover more of the gasmo

41、re quickly. The laterals are unlined, which may cause them tocollapse in some situations. Sometimes the downdip quadrant is leftundrilled because it is more difficult for gas to lift water in an updipdirection. Pinnate w

42、ells have been drilled extensively in theAppalachian basin (209 since 1989), where permeability is low andthe terrain is rugged (it provides a smaller footprint), but also in theArkoma basin (51 since 2000) and the Cahab

43、a basin.In the Arkoma basin, pinnates have performed well as Fig. 6 reveals(Spafford, 2007). They outperformed single-laterals in two ways: (1)cum gas ratio is 2.3, and (2) they continue to produce gas after single-later

44、als died. The production advantage of 2.3× must of course beevaluated along with the additional cost of a pinnate well. The single-laterals die after 18–24 months, which is likely due to the mediumbuild radius (～400

45、 ft) for these wells, making it harder for gas to liftthe water in late-well life (Cameron et al., 2007). In contrast, thepinnates have zero-radius for their build, since the laterals are drilledthrough the mother produc

46、ing well at right angles.However, pinnates do not always do so well. In the Cahaba basin,gas rates are no different from vertical wells which were hydraulicallyfractured. One explanation is that the pinnates may have col

47、lapsed,since the region is chopped up by faults.Another geometry for multi-lateral wells is the quadrilateral (orjust quad) shown in Fig. 7. This is a CNX Gas design for the PittsburghFig. 2. Permeability bands for CBM c

48、ompletions.Fig. 3. Idealized map view of permeability contouring a hypothetical CBM field aftermeasuring permeabilities in many different vertical coreholes. The contours are per-meabilities in one particular seam. The h

49、ighest perm areas may be candidates for cavitycompletions, while the lowest perm areas may require horizontal or multi-lateral wells.2 Prem Sawhney, Reliance Industries, 2008.Fig. 4. Perm-based completion bands for CBM.