OGJ FOCUS: Fall may be imminent for Kansas Cherokee basin coalbed gas output

This constitutes about 13% of the annual gas production in the state, and for the last few years it has offset gas production declines elsewhere in Kansas.

The great majority of Cherokee basin gas production is now coalbed methane (CBM). The major producers are Quest Energy LLC (23.2 bcf in 2008); Dart Cherokee Basin Operating Co. LLC (10.5 bcf); and Layne Energy Operating LLC (6.1 bcf).

Production declines in CBM wells depend on the age of the well, with relatively steep initial declines that stabilize to low rates of decline in older wells. The average yearly production decline for all CBM wells in southeastern Kansas can be used to infer the number of wells that have to be drilled each year to maintain production.

If the last quarter of 2008 is an indication, present drilling for CBM in Kansas is about one third of the activity before the crash in gas prices in late 2008. Only about 200 to 300 new producing gas wells may have been drilled in the Cherokee basin in 2009. With the number of new wells being so low, CBM production in Kansas is probably at or near its historic peak.

Historical perspective

Gas production in Kansas is reported to the state by oil and gas operators for every lease, but the type of production (CBM, shale gas, or conventional) need not be specified.

Specification of the producing zone for some leases, however, is positive identification of CBM production, but even in most of these cases the identification of the producing zone is generalized ("Cherokee Group coals"), and individual coal beds producing gas are usually not reported. As a consequence, a precise figure for CBM production in the state is not possible, but historically, relatively little CBM was produced before 2000.

If southeastern Kansas annual gas production prior to 2000 (2.3 bcf/year) is considered as a baseline for conventional gas production (Fig. 1), approximately 165 bcf of CBM have been produced cumulatively since 2001, which is the year southeastern Kansas gas production started rising dramatically.

CBM production data for Kansas, and associated links, are given on the Kansas Geological Survey (KGS) web site.¹

2008 production snapshot

Production data for 2008 are almost complete with only a few records remaining to be submitted, but the most prolific CBM producers in Kansas in 2008 were (1,000,000 Mcf minimum cutoff [1 Mcf = 1,000 cu ft]):

• Quest Energy (23,245,227 Mcf, 2,247 producing wells).

• Dart Cherokee Basin Operating (10,471,873 Mcf, 1,084 producing wells).

• Layne Energy Operating (6,116,041 Mcf, 864 producing wells).

Most CBM in southeastern Kansas is from Middle and Upper Pennsylvanian high-volatile B and A rank bituminous coals at 800 to 1,200 ft depth.

To this writing in November 2009, approximately 6,800 wells have been drilled for CBM in eastern Kansas. The peak for drilling was in 2006. Successive declines in wells drilled were recorded in 2007 and 2008 (Fig. 1).

The compilation of drilling data for 2008 is nearing completion, and it is clear that the following companies are the most active according to wells spudded in 2008:

• Quest Energy (304 wells).

• Dart Cherokee Basin Operating (166 wells).

• Layne Energy Operating (134 wells).

• Cherokee Wells LLC (130 wells).

• All others (171, with no single company spudding more than 30).

Some of these 905 wells were dry holes, for to date 688 producing wells have been reported for 2008 in the 10 counties in Fig. 1. The number of wells spudded monthly for CBM commensurately dropped with the steep decline in gas prices starting August 2008 (Fig. 2).

Only about 30 wells/month were drilled in the last 3 months of 2008, whereas in 2007 and the first half of 2008, some 90 wells/month were being drilled. This low rate of drilling will likely continue in 2009, for gas prices in 2009 are even lower than they were in the last quarter of 2008—hovering between $3.25 and $5/Mcf, but periodically even dipping below $3/Mcf and sometimes rising above $5/Mcf.

All this raises the question: How will the decrease in price and drilling affect CBM production volumes in coming months? The answer may be inferred from analysis and projection of drilling and production statistics—specifically, with a calculation of the number of wells necessary to maintain production if present production declines at a predictable rate. This decline rate thus needs to be determined.

Analysis of past production

Production data for oil and gas leases are collected by the Kansas Corporation Commission and made available to the public on the web site of the KGS.¹

Monthly production is recorded for leases, not individual wells, and in the situation of CBM production, it is not usually tied any specific coal seam. Virtually all wells are vertical and are completed in several coal seams.

A large majority of leases are reported as single-well leases, thus decline statistics for the individual wells constituting these leases can be calculated. Nevertheless, the lease production database is imperfect.

Operators may have varying attention to detail, and human errors in reporting and entry of the data after submission can create anomalies. For example, upon direct consultation with operators, it was determined that some "single-well leases" with suspiciously high rates of production turned out to be multiwell leases, or they were actually conventional gas wells with initial flush production.

An additional problem in ascertaining annual decline rates for CBM wells in Kansas is that most wells so far have short production histories. For example, wells drilled and completed in 2007 probably hit their peak production in 2008, and their annual rates of decline have yet to be ascertained because their 2009 production is still pending.

The production obtained from CBM wells differs from conventional gas wells. Conventional gas wells typically have their best production 1-2 months after being brought on production, whereas production from CBM wells increases gradually with peak production occurring several months after initial reported production.

This production characteristic is actually useful for identifying CBM wells in the Kansas production database. Almost 12 years of production data for 22 CBM wells in eastern Kansas^{2 3} (Fig. 3) show drastic month-to-month variations, probably due to geological, engineering, and mechanical influences. When averaged though, these wells show a relatively well-behaved production curve typifying CBM production, with building, peak, and declining phases (Fig. 3).

A distribution of CBM maximum monthly production rates (Fig. 4A), based on 2,973 single-well leases in southeastern Kansas, indicates that the average CBM well produces 66.7 Mcfd (2,000 Mcf/month) at its peak rate. Median maximum-rate per well per day is 48.9 Mcfd (1,466 Mcf/month).

The median may better typify CBM production because the average is influenced by a statistical "wing" of relatively prolific gas wells (Fig. 4A), some of which may be conventional or unreported multiple-well leases.

By monthly production, the most productive confirmed CBM well is the Dart Cherokee Basin Operating 2-26 'D' Orr, in southern Wilson County, which recorded 18,461 Mcf in July, 2004. This well is completed in the Bluejacket coal but may have a component of conventional production due to a sandstone in stratigraphic proximity to the coal.⁴

The time necessary for a well to reach its maximum monthly production rate (Fig. 4B) shows a similar type of distribution to the production-rate diagram (Fig. 4A). CBM wells reach their maximum rate 14 to 15 months on average after reporting their initial production. The median value is 10 months.

These results roughly agree with data collected by Ebers⁵ on 96 wells in the Cherokee basin, where two thirds of them experienced peak production 7 months after the start of dewatering and the remaining wells took 8 months or longer.

Inferring future production

Future CBM production can be inferred by comparing the average production decline for producing gas wells drilled in previous years with the expected increase in production from wells drilled in the latest year.

Considering that month-to-month production data vary greatly for almost every well (Fig. 3), a yearly time-scale was chosen to attempt to "damp out" the noise inherent in monthly production rates. A database of 200 wells selected in the 10 counties in southeastern Kansas with considerable CBM production (Fig. 1) was compiled so that a reasonably accurate (but nevertheless generalized) decline estimate could be determined and then applied to the larger population of CBM wells in this region.

Superimposing the monthly peak production distribution for this 200-well database on the distribution for 2,973 wells for southeastern Kansas (Fig. 4A) shows that the cumulative distributions of maximum monthly production in both databases are virtually identical.

In addition, present production of wells in the 200-well database (184 of which are still active) is 20,528 cfd, which is virtually identical to the average daily production for all wells in 2008 (20,249 cfd; Fig. 1). The 200-well database can therefore serve as a reasonable proxy in determining production characteristics for the larger database.

Prediction of future CBM production is approached in two ways.

The first method is simpler than the second, but generalizations have to be made for both methods. The first way (Method 1) is to apply a likely decline rate to the latest annual production (49.1 bcf in 2008) so that the hypothetical drop in production in 2009 can be determined if no new wells were drilled in 2009. The volume of this production decline can be compared with the expected production from new wells drilled in 2009.

The second way (Method 2) is more complicated than the back-of-the-envelope calculation utilized in Method 1. Method 2 models the production history of southeastern Kansas using the number of wells drilled annually, their estimated annual production, and likely annual decline rates based on their age.

Method 1

The month of peak production is selected as the anchor-point for all subsequent production declines.

For purpose of simplification, a typical CBM well is assumed to reach its maximum production 1 year after its initial reported production. This 1-year assumption is close to reality, for data depicted in Fig. 4A indicate the median time to reach maximum production is 10 months, and the average is 14-15 months.

The first year of production is typified by increasing monthly production. It is identified as "year 1," and it presumably culminates with the month of maximum production. The 12 months of production after the peak month constitutes production of "year 2," the following 12 months constitute "year 3," etc.

A production decline percentage can be calculated for year 3 by the conventional method of comparing its production to that of year 2 by the formula:

where V is yearly production volume.

This production decline percentage can be calculated for individual wells or for the summed production of various groups of wells, such as all wells drilled in a given year. For the 200 wells in the database, summed production in year 2 and year 3 was 3,341,152 and 2,457,273 Mcf, respectively, which calculates to a 26.5% decline.

Decline percentages for years 3 and beyond are relatively straightforward, as per equation (1). The calculation of production decline in year 2 is slightly different, for in this method, year 2 records the 12 months of production following the peak-production month.

The production decline for year 2 is calculated by:

where V_{max month} is the maximum monthly production. Although this value is determined for 2,973 wells in southeastern Kansas (median V_{max month} = 1,466 Mcf/month; Fig. 4), V_Year2 is not. However, it can be estimated with the 200-well test database.

The V_{max month} production sum for the 200 wells in the database is 425,070 Mcf/month. Inserting this into equation (2), with 3,341,152 Mcf being V_Year2 (production for year 2), a 34.5% decline is calculated. This indicates that CBM wells generally fall off in production by about one-third 1 year after their peak monthly production is recorded.

The histogram of yearly production declines for individual wells in the 200-well database, shown in Fig. 5, peaks around 32% for the first year, with declines ranging from 6% to 90%.

Rates of decline for the CBM wells generally decrease the longer a well produces (Fig. 5). The decline characteristics of CBM wells are different from those of most conventional wells. Most conventional gas wells have annual decline percentages that change little from year-to-year. This is termed "exponential decline."

Some CBM wells display this also, but the more usual ever-lessening percentages of decline for given time intervals is described as "hyperbolic" or "harmonic."⁶ This is important, for these characteristics indicate that determination of a collective rate of decline for all CBM wells has to take into account the age of the wells and the numbers of these wells that have been drilled each year.

A decline percentage for "year 1" is somewhat conjectural, for this is the year when production typically increases for a CBM well. Nevertheless, "year 1" needs to be accounted for in an estimation of a collective decline percentage for CBM production.

An estimate is possible by comparing "year 2" (declining) production with "year 1" (increasing) production. Year 1 total production (2,454,477 Mcf) in the 200-well database is 73.5% of the production in year 2 (3,341,152 Mcf); so the decline percentage in year 1 is estimated to be –36.1% (Equation 1). A negative decline percentage actually indicates an increase in production.

With this information, a collective decline rate for all the gas wells in 2009 in the 10-county area (Fig. 1) can be determined by averaging the decline percentages presented in Fig. 5, weighted by the number of wells drilled each year. The weighted average decline rate for all southeastern Kansas wells is thus determined to be 17.8% (Table 1).

A gentler collective decline of 13.8% is calculated by averaging the number of new producing wells in a given year with that of the previous year. This essentially estimates the number of wells at the midpoint of each year.

If a yearly decline of 17.8% is applied to the 49.1 bcf of production recorded in 2008, then production would hypothetically decline by 8.7 bcf in 2009 to a total of 40.4 bcf if no new wells were drilled in 2009.

Generalizing that each CBM well averages 7,381 Mcf/year (Fig. 1), then 1,178 wells (i.e., 8.7 bcf/0.007381 bcf per well = 1,178 wells) would have to be drilled in 2009 to maintain the 49.1 bcf annual production of 2008 into 2009. If the 17.8% overall decline is too harsh, the lesser overall decline rate of 13.8% (using midyear well numbers; Table 1) results in a production drop of 6.8 bcf, which corresponds to 918 compensatory wells.

By the calculations using the gentler overall 13.8% decline rate, if more than 918 successful CBM wells are drilled in 2009, then gas production will increase from 2008 to 2009. Fewer than 918 successful wells drilled will mean a decrease.

Data are still being compiled for 2008, but as of November 2009 only 688 additional producing wells were reported for 2008. Experience with reporting patterns indicates that this number will not significantly change in coming months.

The net number of producing wells for 2009 is yet to be determined. If only 360 successful new wells in 2009 are ultimately reported (based on the highly speculative rate of 30 wells/month for the last quarter of 2008, shown in Fig. 2), then the midyear well number for 2009 will be 524 (i.e., (688 + 360)/2).

In effect, 2009 will be 394 wells (i.e., 918 – 524) short of maintaining the record 49.1 bcf production achieved in 2008. The approximate 2009 production decline will thus be 7,381 Mcf/well × 394 wells, or 2,908,114 Mcf (~2.9 bcf). Thus 46.2 bcf of production for 2009 is predicted by this method.

Method 2

A second check (Method 2) is to calculate production using the number of wells drilled each year, the expected production of each well, and the inferred annual production decline percentages.

Method 2 relies on the assumption that the median maximum monthly production (1,466 Mcf/month, Fig. 4) typifies all CBM wells. The decline percentages (Fig. 5) that were used in Method 1 for calculating the weighted average production decline (Table 1) can also be used in Method 2, but with the respective decline percentages applied throughout the history of a well.

Each CBM well is thus generalized to produce 11,523 Mcf in its second year of production (i.e., 1,466 Mcf/month × 12 months × (1 – 0.345)). The first year of production is 8,469 Mcf (i.e., 0.735 × 11,523 Mcf). Production for the third year, coincidently, is 8,469 Mcf (i.e., 11,523 Mcf × (1 – 0.265)). Production for the fourth year is 7,008 Mcf (i.e., 8,469 Mcf × (1 – 0.1725)), etc.

Smoothed decline percentages (Fig. 6) based on the raw data, however, were used in the modeling. The production history for the region can thus be constructed based on the number of CBM wells drilled every year and their expected production declines.

The number of new CBM wells drilled per year can either be taken as the actual number compiled for a given year, or as an average with the previous year so as to approximate the number of wells at the midpoint of each year (Table 1).

In circumstances where there is drastic variation in well numbers from year to year, the midyear number probably better characterizes the number of wells that are behaving in a similar manner with increasing or decreasing production.

An additional consideration is conventional gas production, which is not differentiated from CBM production in the production database. Not many CBM wells were drilled prior to 2001, so production from 1995 to 2000, which averages 2,332,175 Mcf/year, is considered for modeling purposes as a constant baseline for conventional production in southeastern Kansas.

By this method, modeled and actual production reasonably compare with each other (Fig. 6), with 2008 modeled production of 50.2 bcf being close to the 49.1 bcf actually recorded (Fig. 6).

By the model, approximately 420 new producing wells have to be added in 2009 maintain steady production. This causes the 2009 midyear well number to be 554 wells [(688 + 420)/2)]. Due to the composite production declines of all previous producing wells, approximately 1,020 new wells would be needed in 2010 to maintain steady production. Present drilling is far short of attaining these requisite numbers.

Individual well performance

The attempts above to describe the central tendency of CBM wells in southeastern Kansas and their collective behavior can also be utilized to typify what production history can be expected for individual wells.

Production behavior based on peak production of individual wells (Fig. 4) is thus used to infer the likely production history of a well (Fig. 7).

Just as each CBM well is different, production economics are also different, but for purpose of simplification, 5 Mcfd (1,825 Mcf/year) is considered the level at which a well will be shut down. By this scenario the median CBM well will produce 71.6 MMcf over a 15-year life. Not surprisingly, more prolific CBM wells will produce more and longer; less prolific wells will produce less and will be plugged earlier.

Gas prices will affect the payout time and the longevity of a well. Approximately $125,000 is necessary to drill and complete a CBM well in this region.⁷

Under these conditions, the median CBM well will recoup $125,000 in drilling and completion costs in slightly over 5 years with gas priced at $3/Mcf, whereas a less productive 25th-percentile well will not able to pay back its costs (Fig. 7). However, this 25th-percentile well will recover its costs in about 5 years if gas is priced at $6/Mcf; the median well will achieve this level of production in just over 2 years.

Some wells (certainly the lowest 10th-percentile) cannot pay for themselves even if gas prices are sustained at the relatively high level of $9/Mcf.

Feedback and caveats

The statistical hinge on which much of the analysis depends is the average annual decline rate of CBM wells and its empirical relationship a central (median or average) maximum monthly production rate reported for CBM wells in the Cherokee basin.

Basic production data inherently have considerable scatter, and the numbers on which the modeling is based can also carry uncertainty. No two CBM wells are alike. Perhaps production data summed on a quarterly basis would reduce some of the scatter in the data, but a test of this speculation is best the subject of another academic exercise.

Any predicting of future production and drilling is at best an educated guess because commodity prices, well costs, and technology will certainly change with time. The slow economic situation today may drastically change for worse or better, and all this will affect the number of wells drilled.

The number of producing CBM wells added for 2008 (688) is largely compiled by now, but the number of new CBM wells for 2009 was difficult to predict even at this writing in the last quarter of the year, for there is several months lag time before these data are finalized.

The crash in gas prices from the high recorded in mid-2008 also likely changes the way companies marshal their capital. Company resources do not necessarily dry up for new drilling during hard times, but they can be redirected in other ways. For example, instead of contracting rigs for drilling new wells, companies may decide it is more politic to employ workover rigs to maximize production in their existing wells.

Similarly, aggressive economizing may result in shutting-in or abandonment of marginal wells, thus the average and median well productivity could actually increase in coming years. This may be occurring for many leases in 2008 that were producing less than 8 Mcfd, with declining production, and have not had any production reported in 2009.

Are 360 and 200 new producing wells respectively projected for 2009 and 2010 realistic? Anemic rates of drilling in late 2009 indicate that fewer gas wells will be completed, but the final number depends on economic, technologic, and even political events.

Upon query by the author, geologists from the three primary CBM companies producing in southeastern Kansas (Quest, Dart, and Layne) stated that their companies were drastically scaling back developmental drilling in 2009, so the educated guess that 360 new producing wells will have been drilled in 2009 may be even wildly optimistic.

Some feedback on what can be expected for 2009 is available though, for 20.6 bcf of gas production is reported through May 2009 for the counties shown in Fig. 1. This is proportional to 49.4 bcf for the entire year. This hints that the two methods presented for predicting future production are pessimistic, but the point still stands that the impressive production increases of previous years due to development of CBM in this region cannot be expected even in the near future.

The predicted drop in production does not mean the CBM resource in this region is being depleted. More accurately, the resource simply cannot be offered to the market given low gas prices in late 2009.

Many drilling locations in southeastern Kansas are still available. For example, Quest Energy reports on its website that it had in 2008 an inventory of more than 2,100 drilling locations on its leased acreage, and that its CBM wells have a life of about 15 years.⁸

If a sufficient number of remaining locations are drilled and put on production in coming years by Quest and other Cherokee basin CBM operators, then annual gas production can conceivably rise by several billion cubic feet from the 49.1 bcf recorded in 2008.

The prediction of declining production is not to say that "the bloom is off the rose" (i.e., what was once attractive is no longer), but rather, even if production economics and technology do not radically change, CBM production will still be a vital part of the regional economy for many years.

The production-by-year graph will not be a symmetric bell-shaped curve with a decline as steep as its rise; it will likely resemble the positive-skewed curve characteristic of a CBM well (Fig. 3). Fortunes can still be made, and the economic impact of CBM in Kansas and the Midcontinent is and will be substantial.

Acknowledgments

The author thanks colleagues Larry Brady, Rex Buchanan, John Doveton, and Dan Merriam; and Shane Huffman (Dart Energy), Ken Recoy (Quest Resource), Jim Stegeman (Colt Energy), and Rolland Yoakum (Layne Energy) for critically reading the manuscript and giving helpful suggestions for its improvement.

References

1. Kansas Geological Survey web site (http://www.kgs.ku.edu/PRS/petroDB.html ).

2. Newell, K.D., Brady, L.L., Lange, J.P., and Carr, T.R., "Coalbed gas play emerges in eastern Kansas basins," OGJ, Dec. 23, 2002, pp. 36-41.

3. Newell, K.D., Johnson, T.A., Brown, W.M., Lange, J.P., and Carr, T.R., "Geological and geochemical factors influencing the emerging coalbed gas play in the Cherokee and Forest City basins in eastern Kansas," Kansas Geological Survey, Open-file Report, No. 2004-17, 2004 (http://www.kgs.ku.edu/PRS/publication/2004/AAPG/Coalbed/index.html).

4. Personal communication, Shane Huffman, Dart Energy, 2009.

5. Ebers, M.L., "Kansas CBM flow rates correlate to coal gas content," OGJ, June 22, 2009, pp. 34-40.

6. Okuszko, K., and Gault, B., "Analyze CBM decline performance," Hart's E&P, Vol. 80, No. 6, 2007, pp. 99-101.

7. Personal communication, Jim Stegeman, Colt Energy, 2009.

8. Quest Energy web site (http://qrcp.publishpath.com/company-info).

The author

More Oil & Gas Journal Current Issue Articles
More Oil & Gas Journal Archives Issue Articles
View Oil and Gas Articles on PennEnergy.com