Well production profiles to assess Fayetteville gas potential revisited

May 7, 2012
This article builds upon the analyses presented in the 2011 OGJ article, "Well production profiles assess Fayetteville shale gas potential."

James Mason
Energy Consultant
Farmingdale, NY

This article builds upon the analyses presented in the 2011 OGJ article, "Well production profiles assess Fayetteville shale gas potential."1

In that article, the reported average well production profile had a 1.7 bcf estimated ultimate recovery and was based on averaging aggregate well production over a multiyear period, October 2005-September 2010. The author received criticism that the multiyear averaging understated well production gains realized in recent years.

Advances in well production technology are increasing well production rates. Up-to-date average well EURs are important because E&P companies use EUR expectations to book reserves. Hence, the criticism I received is valid, and in response this article reports average well production profiles for each of the years 2008 through 2011.

In addition, this article addresses the ongoing debate about average well production decline rates and the Energy Information Administration estimate of US shale gas technically recoverable resources (TRR).

Average well production decline rates are important in assessing the average well EURs used by E&P companies for reserve bookings. Also, the EIA estimate of the Fayetteville shale TRR is evaluated.

Natural gas production in the Fayetteville shale is analyzed with monthly well production data compiled by the Arkansas Oil and Gas Commission (AOGC).2 The analyses are augmented with information from E&P company presentations.

Main Fayetteville players

The Fayetteville shale lies in the Arkoma basin and covers 9,000 sq miles (Fig. 1). The play is segmented into a 4,000 sq mile eastern section and a 5,000 sq mile western section. Shale gas exploration and production is focused in the core area of the eastern section. The Fayetteville shale is a dry gas play and does not produce natural gas liquids or crude oil.

Three E&P companies operate over 99% of the shale gas wells in the Fayetteville shale: Southwestern Energy Co.'s subsidiary SEECO; ExxonMobil's subsidiary XTO Energy; and BHP Billiton Petroleum. SEECO is the first mover in the Fayetteville shale and reports lease rights to 800,000 net acres in the eastern section and 125,000 net acres in the western section.3

In 2010, ExxonMobil purchased XTO Energy and Petrohawk Energy's Fayetteville shale lease holdings to bring its total Fayetteville shale lease rights to 560,000 net acres.4 BHP Billiton Petroleum purchased Chesapeake Energy's Fayetteville shale holdings in 2011 and has lease rights to 500,000 net acres.5

The net leased acreage in the eastern portion of the Fayetteville shale is about 1.9 million acres or 3,000 sq miles. E&P companies report an average 80-acre well spacing for well development of the Fayetteville shale.

Fayetteville well spacing

Well spacing affects well drainage area, which in turn affects well production rates and well EURs. Therefore, an understanding of the relationship between well spacing, well drainage area, well production rates, and well EURs is important. The drainage area of a sample well with 80-acre well spacing is presented in Fig. 2.

A change in well spacing, in and of itself, does not change total gas recovery for a given development area. This is demonstrated by changing the assumed 80-acre well spacing to 70 acres. When well spacing is reduced to 70 acres, average well EUR is reduced to 2.8 bcf from 3.2 bcf but total gas recovery is 25.6 bcf/sq mile of development area in both cases, other factors being the same. In converse, when well spacing is increased, well production and well EUR increase but total gas recovery per unit of development area still remains the same.

Advances in horizontal well hydraulic fracturing technology are occurring. Two notable trends in the Fayetteville shale are: 1) a reduction in well drilling time, which decreases well drilling costs; and 2) drilling longer laterals, which enables more frac stages and increases well production. For example, from 2009 to 2011 Southwestern Energy reported a decrease in average drilling time from 11 days to 8 days, an increase in average well lateral length from 4,100 ft to 4,836 ft, and a decrease in average drilling costs from $2.9 million to $2.8 million.3

Well production profiles

The focus now turns to average well production profiles for each of the years 2008 to 2011.

The following well production analyses are based on monthly well production data. The initial production (IP) rate is the first "full" month of well production, which is the second month of reported well production. Well production for the first reported month is most likely not a full month of production since a well can begin production on any day of the month.

Well completion totals for 2008-11 are presented in Fig. 3. In 2011, the number of well completions in the Fayetteville shale was 848, or 41 less than in 2010. Well completions in 2011 compared with 2010 increased for SEECO and XTO and decreased for BHP. The decrease in BHP well completions is possibly a consequence of BHP assuming control of Chesapeake Energy's Fayetteville shale holdings in 2011 and not necessarily a planned reduction.

A slowdown in well completions is not surprising with low natural gas prices and well development demands in other shale gas plays. E&P companies are faced with a difficult balancing act in terms of acreage development priorities, dry natural gas production levels, and low natural gas prices.

It will be interesting to watch the progression of well completion rates 2012 forward, not just in the Fayetteville shale but all dry shale gas plays.

With expectations of low natural gas prices 2012-13 and corresponding reductions in hedge and contract prices, well development in the Fayetteville shale is likely to slow 2012 forward. For example, Southwestern Energy plans to reduce Fayetteville shale well completions in 2012 to 420-30 net operated wells, which is 25% fewer than their 2011 well completions.3 With the planned reduction in 2012 well completions, Southwestern Energy's projection of 2012 Fayetteville shale gas production is only about 6% greater than 2011 production.

In 2011, Fayetteville shale natural gas production totaled 943 bcf, which is 21% greater than 2010 production. Annual natural gas production totals for the Fayetteville shale are presented in Fig. 4. Even with curtailment of well development, it is likely that Fayetteville shale natural gas production will top the 1.0 tcf mark in 2012.

Advances in well production technology are increasing well production. In the 2011 OGJ article, the aggregate average well EUR with 40 years of well production is 1.7 bcf.1 With the addition of 15 months of well data, the aggregate average well EUR with 40 years of well production increases to 1.8 bcf.

As previously noted, an aggregate average well production profile generated by averaging multiyear well production totals does not convey accurate information about current expected average well production since multiyear averaging washes out the effects of most recent advances in well production technology.

To address this, average well production for the first complete 12 months of well production is calculated for each of the years 2008 through 2011 (Fig. 5). The first year average well production totals clearly show the increases accruing from well designs with longer laterals and more frac stages.

While there are significant increases in first-year average well production from 2008 to 2010, the annual incremental increases decline each year. Also, the 2011 first year average well performance is basically the same as that in 2010. This suggests that well design for the Fayetteville shale is approaching optimization.

It should be noted that the 2011 average well production totals are based on incomplete monthly well data and therefore should be interpreted as preliminary estimates. The annual well production cycle for wells completed in January of each year is February through January of the following year.

For example, wells completed in January 2011 will not have a 12th month production total until January 2012 since the first "full" month of production is the second reported month. This means that there are not 12 months of well data for each of the months in 2011, and each sequential month has a fewer number of wells than the previous month.

As reported in the 2011 OGJ article, there is wide variation in individual well production, which holds even in the case of contiguous wells. It goes without saying that shale gas drilling involves risk. While there is wide variation in production between individual wells, it is the average well production over a portfolio of wells that matters to E&P companies and investors.

Average well EURs are important because they are used to establish reserve bookings for specific well development acreage.

Average well EURs are estimated by fitting a hyperbolic curve to the actual first year average well production curves in Fig. 5. The average well EUR estimates are based on a 30-year well production life, which is more conservative than the 40-year well life used in the 2011 article.

The conventional hyperbolic curve formula used by E&P companies is:

Hyperbolic curve formula: qt = qi/(1 + b Di t)1/b,

where qt = production in month t
qip = initial production
b = decline exponent
Di = nominal decline rate
t = time in month of production

The decline exponent, b, and the nominal decline rate, Di, are unknowns and are estimated by minimizing the squared differences between the actual first-year monthly well production data points and the hyperbolic formula fitted data points. The minimization routine searches for the b and Di values that generate a best fitting curve. The results are then extrapolated to generate a 30-year average well production profile.

It should be noted that the average well production decline rate generated with a hyperbolic curve formula decreases continuously over a well's production life. The continuous decrease in the average well production decline rate has generated controversy since long-term average well production decline rates are not known with certainty.6

To address this concern, results are reported for average well production models with more conservative, "constant," terminal monthly average well production decline rates of 0.5% (6% annual) and 0.83% (10% annual). The average well production profile generated with the hyperbolic curve formula is the "baseline" model, and the two constant terminal average well production decline rate models are for sensitivity analyses.

The baseline 30-year average well production profiles, 2008-11, are presented in Fig. 6 and Table 1. The average well EUR estimates increase from 1.8 bcf for wells completed in 2008 to 3.2 bcf for wells completed in 2011. The 3.2 bcf average well EUR for 2011 is consistent with BHP's reported Fayetteville shale average well EUR.5 Notice that there is only a 5% difference in the 2009 and 2011 average well EUR estimates, which is another indication that an optimized well design is being approached in terms of maximizing well production. Once again, the 2011 result is preliminary.

Maintaining 1 tcf/year

Since natural gas production for the Fayetteville shale is approaching 1.0 tcf/year, it is of interest to evaluate the well count necessary to maintain 1.0 tcf/year of natural gas production level for 30 years.

Well production decline rates require the continuous addition of new wells to maintain a constant annual production level. With the baseline 2010 average well production profile, it takes 1,869 wells to initiate an annual 1.0 tcf gas production level, and the cumulative number of wells to maintain this annual production level for 30 years is 12,188 wells. The annual and cumulative well counts are presented in Fig. 7.

There is controversy about E&P company presentation of average well EURs and long-term well production decline rates, which impact reserve bookings. Accurate determination of long-term well production decline rates is problematic because of the short history for horizontal wells with hydraulic fracturing. In 2011, the controversy resulted in investigations by the Securities & Exchange Commission and State of New York Attorney General Office.

Because of the short 5-year well production history for horizontal wells with hydraulic fracturing in the Fayetteville shale, it is impossible to know with certainty the long-term average well production decline rate.

To provide insight into what is occurring to date, a 5-year aggregate average well production history is presented in Fig. 9. The average annual decline rates for each of the 5 years of aggregate well production history are: Years 1-2 = 49%; Years 2-3 = 36%; Years 3-4 = 29%; and Years 4-5 = 16%. Another 5-8 years of well production history is needed to get an accurate sense of long-term well production decline rates.

Hence, it is important to evaluate the effect of different average well production decline rates on average well EURs. Average well EURs are reported for two constant, terminal well production decline rates: 10% annual (0.5% monthly) and 6% annual (0.83% monthly).

The timing of when the constant terminal decline rates take effect is determined by the decline rate trajectory of the baseline hyperbolic curve estimates. That is, the terminal decline rate is held constant from the month when the hyperbolic curve model equals to the assigned constant, terminal monthly decline rate.

Graphic representations of the 6% and 10% annual terminal decline rate models are presented in Fig. 8. Notice that the 10% terminal decline rate takes effect earlier in a well's production life than does the 6% terminal decline rate, which means that it has a greater effect on average well EUR. The impact of the two constant terminal decline rates on average well EURs is presented at the bottom of Table 1.

The 6% terminal decline rate decreases average well EUR by 5% compared with the hyperbolic curve average well EUR, and the 10% terminal decline rate decreases average well EUR by 16%.

It appears that E&P companies are applying conservative well production decline rates for their average well EUR estimates used to estimate their reserve bookings. For example, Southwestern Energy is currently booking reserves based on a 2.4 bcf well EUR, which is about 25% less than its average well EUR calculated with the hyperbolic curve formula.3

EIA's recoverable estimate

The final task is an assessment of the EIA's TRR estimate for the eastern section of the Fayetteville shale.

The EIA's Fayetteville shale TRR estimate and the underlying assumptions are presented in Table 2.7 The question is whether the EIA's 27 tcf TRR estimate and underlying assumptions are reasonable expectations.

First, the EIA's assumption of a 2.25 bcf average well EUR is evaluated with knowledge gained from the aggregate average well EUR derived from the 5-year well production history. Second, the well development area required is evaluated by inspection of well coverage area within the identified production areas of the core counties. The production area of core counties is determined by the distribution of well completions and average well production totals.

The 5-year aggregate average well production history presented in Fig. 9 has a 40-year 1.8 bcf aggregate average well EUR (the 30-year EUR is 1.7 bcf). The 1.8 bcf aggregate average well EUR is 0.1 bcf greater than the 1.7 bcf aggregate average well EUR reported in the in the 2011 article.1

Since aggregate average well EUR increased 0.1 bcf with the addition of just 15 months of well data, it is reasonable to expect that the aggregate average well EUR will at least equal the EIA's 2.25 bcf average well EUR when the high well production rates of recent years gain dominance in the averaging of aggregate well production.

The well development area for the eastern section of the Fayetteville shale is estimated by mapping wells by county location and well production. From this, the percentage of county area expected to be covered by well development is estimated.

The county distribution of well completions is presented in Fig. 10, county level average first full month well production totals are presented in Fig. 11, and the spatial distribution of wells is presented in Fig. 12. From the information in the figures, it is clear that the highest quality core production areas are in Van Buren, White, Conway, Cleburne, and Faulkner counties.

These counties contain 96% of total Fayetteville shale well completions. The highest well production is in Van Buren and Conway counties, and the lowest well production is in White, Cleburne, and Faulkner counties. The lowest production totals are only about 20 less than the highest totals in Van Buren and Conway counties.

While annual well completions are increasing in Independence County, average well production is sharply lower than that in the core counties. Hence, Independence County is defined as a noncore production area. Wells are located in other counties, but the number of wells is few and their production rate is low.

The percentage of the county area targeted for well development is estimated by inspection of the well location map in Fig. 12. Total county land area and estimates of percentage of the county land area targeted for well development are presented in Table 3.

The result is a gross well development area of about 1,943 sq miles. If 80% of this area is suitable for well development, then the net well development area is 1,555 sq miles, which is a little greater than the EIA derived 1,518 sq mile well development area in Table 1.

To estimate total gas recovery, the number of wells with 80-acre well spacing for the 1,555 sq miles of net well development area is multiplied by the EIA's 2.25 bcf average well EUR. The result is 28 tcf of total gas recovery, which is 3% greater than the EIA's 27 tcf TRR estimate. Based on these results, the EIA's 27 tcf TRR estimate appears reasonable.

Core area well saturation

With knowledge about the size of the core production area and the average well production profile, it is possible to forecast when well saturation of the highest quality core area of the eastern section of the Fayetteville shale is likely to occur.

Well saturation of the highest quality core area of a shale gas play is an important event because it results in a significant decrease in average well production. A decline in average well production requires an increase in wellhead gas prices to cover well costs. Also, a decline in average well production increases the number of wells required to maintain a constant level of annual production, which in turn requires higher wellhead gas prices to cover the increase in total well development costs.

Therefore, when well saturation in the highest quality core area of a shale gas play occurs, E&P companies will have to evaluate the costs/benefits of continuing or delaying well development in the play. E&P company decisions as to whether to expand well development once well saturation of the core area occurs will largely be dependent on well development opportunities in other shale gas plays.

From the well count estimates to maintain a 1.0 tcf annual gas production rate with the 3.1 bcf EUR average well production profile, a net well development area of 1,555 sq miles, and an average 8 wells/sq mile, well saturation of the highest quality core area of the eastern section of the Fayetteville shale is likely to occur around 2035. If the annual production rate is increased to 1.5 tcf by 2020, then well saturation of the core area occurs around 2025.

Well spacing is not a factor since a change in well spacing only changes the number of wells required to maintain a particular annual natural gas production level, and the well saturation date remains the same regardless.

Two caveats regarding the well saturation projections need to be considered.

First, the well saturation projection is highly dependent on the net well development area. If a new high quality core area in the eastern section of the Fayetteville shale were discovered, then the well saturation date is extended. Second, the projected dates for well saturation presented above are likely outside estimates since they are based on a high-end, 3.1 bcf EUR average well production profile.

References

1. Mason, James E., "Well production profiles assess Fayetteville shale gas potential," OGJ, Apr. 4, 2011, p. 76.

2. Arkansas Oil & Gas Commission, "Fayetteville Shale Gas Sales Information," Little Rock.

3. Southwestern Energy Co., Form 10-K for the fiscal year ended Dec. 31, 2011, commission file number 1-08246, Southwestern Energy, Houston, Feb. 28, 2012.

4. Williams, Jack, president, XTO Energy, "Shale Gas: The Keys to Unlocking Its Full Potential," SPE Unconventional Gas Conference, Houston, June 14, 2011.

5. Yeager, J. Michael, BHP group executive and chief executive, petroleum, "BHP Billiton Petroleum Onshore US shale briefing," investor presentation, Nov. 14, 2011.

6. Berman, A.E., "Facts are stubborn things," World Oil, November 2009.

7. Energy Information Administration, "Review of Emerging Resources: US Shale Gas and Shale Oil Plays," US EIA, Department of Energy, July 2011.

The author

James Mason ([email protected]) is an energy consultant in Farmingdale, NY. He has published numerous articles on energy issues in peer-reviewed scientific journals. He has a PhD in economic sociology from Cornell University and a master's in environmental sociology from the University of New Orleans.

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