Musings: Marcellus Shale: Good News Critique

In the last issue of the Musings, we wrote about good news and bad news for the development of the Marcellus gas shale deposit extending across New York, Pennsylvania, West Virginia and eastern Ohio.  This deposit with its multiple shales is considered to be potentially the largest gas deposit in the United States.  It’s economics are challenging as the area is hilly, the road access is less than ideal, the land holdings are fractured and the public is not necessarily enamored with oil and gas drilling activities, especially hydraulic fracturing, which is key to the successful development of gas shale deposits.  Low natural gas prices are potentially the biggest hurdle for Marcellus gas profitability. 

Our article discussed the recently released 12-month natural gas production data for wells in the Pennsylvania portion of the Marcellus through June.  The data showed average cumulative production for Marcellus horizontal wells in the 5-county core area of the North Central and Northeast part of Pennsylvania.  The new data shows solid production results, and in fact, the average well’s production slightly exceeded the expected production suggested by Chesapeake Energy (CHK-NYSE) in a 2008 investor presentation.  That chart was presented to show the company’s anticipated well economics for its foray into the region.  Pennsylvania has a long history of oil and gas having been the cradle of the U.S. oil business with the Drake well drilled in 1859.  Coal and oil and gas have a long history in the state and were key commodities that enabled the state to become a leader during the industrial revolution.

The better-than-expected gas shale production was the good news, the bad news was the Environmental Protection Agency’s (EPA) request for hydraulic fracturing data from nine oilfield service companies to support the Agency’s investigation into whether the procedure should come under greater regulation.  In the research for our article, we relied upon an article published in The Scranton Times-Tribune where the writer had interviewed Dr. Terry Engelder, professor of geosciences at Penn State University and a student of the Marcellus gas shale, about the significance of the production data.  After commenting about how much better the production data was compared to Chesapeake’s 2008 expectation, we pointed out that the professor compared the value of the gas production to the amount of investment gas producers made during the same 12-month period.  Costs exceeded all gas revenues.  We took the liberty to slide from this simple comparison of revenue to cost to point out that “It is these negative economics that are beginning to play havoc with the profitability of the E&P companies active in the gas shale formations.” 

Our article was re-published by (read full article) and it drew a response from Dr. Engelder pointing out the fact that well profitability is determined over an extended period of time and our broad-brush conclusion was inaccurate. (read full article) wrote a follow-up article correcting our misstated conclusion.  We want to make it clear that we fully understand the timing issue in determining the economics of oil and gas exploration and production efforts.  Our point, which unfortunately is more complex and deserves greater explanation than our one-line observation, is that current low gas prices are nowhere near the prices producers had plugged into the economic models when beginning their leasing and drilling efforts in the region.

For nearly two years, while natural gas prices were falling to below $4 per thousand cubic feet (mcf) of production, but have recently rallied to slightly above that level, drilling and production in the Marcellus was ramping up.  A land rush of leasing has been underway which has created extraordinary pressure on the industry to drill the leases in order to hold the acreage by production.  Once wells are drilled and begin producing, the producer can choke back the rate of production or possibly suspend the well in response to the low gas price.  As shown in a recent leasing report (although hard to read), in Tioga and Washington Counties in Pennsylvania, the most common lease bonus paid was $2,000 per acre with royalty rates (at the wellhead price) ranging between 12.5% and 18.0% in Tioga County and 12.5% and 20.0% in Washington County.  The lease terms were reported to be between three years on the low side and 10 years on the high.  A large number of leases have been 3 + 2 (three years in length with an option for a two-year extension).  Some producers have been trying to secure leases with 5 + 2 terms because of the long lead-times for drilling wells and getting them on production, especially in light of the pipeline bottlenecks for moving natural gas, and in some cases natural gas liquids, to market.

The Marcellus gas formation is an important new resource for this country.  Dr. Engelder’s estimate in 2009 was that the Marcellus formation contains 2,445 trillion cubic feet (Tcf) of gas in place and

Exhibit 1
view large

about 489 Tcf can be recovered with today’s technology.  That recoverability estimate represents about 20% of the in-place gas, which is likely low given the industry’s history of successful application of technology to gas shale production. 

To put into perspective the significance of the Marcellus, it spans a total area of about 95,000 square miles compared to the Barnett, the most successful of the shale plays, with only 5,000 square miles.  The Barnett, after 17 years of development, produced 4.8 trillion cubic feet of gas in 2009.  An additional positive for the Marcellus formation is that the gas is found at shallower depths than many of the other shale formations being exploited in the country now.  The Marcellus tends to be located at depths between 5,000-feet and 8,000-feet, which translates into lower drilling costs for wells and potentially cheaper hydraulic fracturing costs due to the need for less horsepower to break up the shale rock.

Exhibit 2

The challenge for Marcellus production (and many other gas shale basins) is the economics of drilling and completing wells.  Compared to the anticipated well results presented by Chesapeake back in 2008, the gas production data provided by the Pennsylvania Department of Environmental Protection of daily production of about 1.95 million cubic feet per day (mmcf/d) or cumulative production of 0.71 billion cubic feet (Bcf) per well exceeds the Chesapeake target.  Certainly the Pennsylvania data is good news.  Whether the production data supports increases in the economically ultimately recoverable (EUR) reserves from a well is probably premature to assume.  The shape of the well production decline has been a basis of disagreement among people involved in the industry.  Is the decline curve parabolic or linear?  Only the additional production history of these wells will clarify the decline curve shape.

Exhibit 3

Above we have shown the 2008 chart that Chesapeake Energy presented to investors.  As can be seen from the chart, Chesapeake anticipated cumulative production in the first year of 0.67 Bcf, so the current Pennsylvania production of 0.71 Bcf per well is certainly an over-achievement. 

We also found a chart from a 2009 presentation about the Marcellus that Chesapeake made to industry representatives and investors.  In that chart, Chesapeake anticipated cumulative production of closer to 1.25 Bcf, which is more than 50% ahead of the actual Pennsylvania data and nearly twice what the company showed barely a year before.  Possibly Chesapeake will say that this later year chart is based on “targeted” wells, so they may be focusing on the sweet spots in the Marcellus formation. 

More interesting, however, is to compare some of the data parameters associated with the two charts.  In 2008, the targeted EUR was 3.75 Bcf compared to the later estimate of 4.2 Bcf.  The initial production of 4.3 per mmcf/d compares to the later year’s 4.0


Exhibit 4

mmcf/d estimate.  The lower initial production rate in 2009 is associated with a greater 10-year cumulative production estimate (2.35 Bcf) versus the 2008 estimate (2.11 Bcf).  Despite the better EUR and 10-year cumulative production estimate, the finding cost estimate went from $1.12 per mcf/d to $1.28 in 2009.  That increase is probably a reflection of rising costs of doing business, one of which is drilling and completion costs.  As a result of oilfield inflation, the cost estimate for the 2008 well of $3.5 million increased to $4.5 million in 2009. 

Exhibit 5

The other side of the economic equation is the price of natural gas.  When companies were initially targeting the Marcellus, natural gas prices, as represented by the futures market, were in the $8 per mcf in mid 2007 and over $12 in mid 2008.  By mid 2009, the futures price had fallen to under $4 but rallied to nearly $4.50 in mid 2010.  Current gas futures prices are barely over $4 per mcf, which is a half or a third of the price when the Marcellus leasing rush was underway. 

A July report prepared by Dr. Timothy J. Considine of Natural Resource Economics, Inc. for the American Petroleum Institute entitled “The Economic Impacts of the Marcellus Shale: Implications for New York, Pennsylvania and West Virginia” sets forth an analysis based on a slightly different set of well-production assumptions.  The study presented its well decline curve and production assumptions, which we have shown below.

Exhibit 6

The report stated the following: “Industry experience using modern technology for shale gas production in the Barnett now extends 17 years. Companies have estimated production decline curves from actual well experience. While translating experience from one shale play to another may not be entirely accurate, most companies are finding their production decline curve models fit early actual production from Marcellus wells reasonably well. In any play, there is a great deal of variation in the rates of initial production and the rates of decline.”

The report went on to say: “This [decline] curve is on the low end of publicly available decline curve information released by five major Marcellus Shale operators during the second half of 2009. The estimated production over the first 30 years is 2.8 billion cubic feet, after 50 years the yield is 3.5 Bcf. With this decline curve, average annual production from a Marcellus horizontal well is over 500 million cubic feet (mmcf) during the first year and about 250 mmcf during the second. After 8 years annual production is about 100 mmcf, and roughly 30 mmcf per year after 30 years of production. Vertical wells have similarly shaped decline curve but substantially lower output. This study assumes that annual production from a vertical Marcellus well is slightly less than 30 percent of the output from a horizontal well.”  What was not commented on was the impact of the discounted value of production beyond 10 years which has little value on the economics of wells.

As can be seen, the assumed production profile used in the study is considerably below the assumptions used by Chesapeake and is lower than the recently released production data for Pennsylvania Marcellus wells.  What can we conclude about the economics of the wells?  What we have seen, based on data collected by National Resource Economics, Inc. (NRE), is that in 2008 total spending in Pennsylvania was $3.2 billion and increased to $4.5 billion in 2009.  This compares with the $720 million of revenue from gas production.  Within that spending total, lease expenditures actually declined to $1.7 billion in 2009 from $1.8 billion the prior year.  Drilling and completion expenditures nearly doubled in 2009 to $1.7 billion, up from barely $0.9 billion spent in 2008.  Pipeline and processing investments increased significantly from $0.3 billion in 2008 to $0.7 billion in 2009, and are likely to go higher in the next few years as NGL processing and pipelines and new gas pipelines are built.

An interesting aspect of the spending data is that Pennsylvania expenditures are climbing while those in West Virginia are declining.  The author believes this spending disparity may reflect the absence of a severance tax in Pennsylvania, but one cannot rule out the impact of low gas prices and higher liquids content in Pennsylvania wells making their economics more attractive.  Pennsylvania is working to put a severance tax in place by October 1.

Dr. Considine’s report addressed the profitability of Marcellus gas production activity.  “It is very difficult to empirically estimate average and marginal extraction costs for the Marcellus industry when the companies have negative cash flow during the early phase of development as wells gradually get connected to pipeline systems and produce marketable gas. Nevertheless, discounted cash flow analyses of individual Marcellus wells suggest strong rates of return given drilling and gatherings costs and, of course, market price, which is a key factor affecting the development of the Marcellus. Since natural gas prices are volatile, gas drillers may lock in a price with a futures contract.”  We didn’t see any mention of leasing and overhead costs in the economic analysis statement.

Because of the potentially large Marcellus gas shale resource certain factors offer potentially significant economic advantages relative to other supply sources.  For example, the reserves have an important location advantage in providing access to a large natural gas consuming market that could grow significantly depending upon government environmental actions in the future.  Pennsylvania and its five bordering states currently consume 9.5 Bcf/d of natural gas. 

These states also have a large number of coal-fired electric power plants.  If all the coal-fired generating capacity were replaced with natural gas-fired capacity, daily consumption would grow by 7 Bcf/d.  Thus, there is the potential for a 16 Bcf/d gas consuming market within 200 miles of Marcellus production. 

Exhibit 7

The proximity to a large gas consuming market also translates into better gas prices.  The distance the gas needs to be transported is considerably less than competing supplies from the Gulf Coast or Southwest, let alone gas from the Rocky Mountains.  Over 2002-2009, Pennsylvania city gate prices have averaged 14.5% more than the national average.

Exhibit 8

An additional aspect of Marcellus natural gas pricing is its liquid content.  Natural gas liquids (NGLs) are priced at a discount to crude oil, but given oil’s high price, presently, the value of the gas stream is boosted improving production economics significantly.  On the flip side, these liquids require processing facilities and separate pipelines to reach markets.  There needs to be investments in these facilities and the timing could impact future production.  Likewise, as we add this stream of NGLs into the national market, there are questions about what might happen to the pricing of certain liquids and whether it might erode Marcellus gas producer economics.

Given the long-term attractiveness of the Marcellus formation, Dr. Considine’s study projected the amount of drilling and production that might occur in this region.  The study begins with the recognition of the activity that has recently occurred.  The following series of exhibits show the trend in wells drilled in Pennsylvania and West Virginia over the 2005-2009 period and the growth in natural gas production.  All the trends are positive.

Exhibit 9
Exhibit 10

Employing the well production profile outlined above, the professor produced a forecast for 2010-2020 for drilling and gas production in the Marcellus formation.  In the study, he also prepared an analysis of the potential for activity in New York State that currently has significant restrictions on activity until studies are completed on the risks from hydraulic fracturing activity.  The forecast for wells drilled and daily gas production are contained in the exhibit below. 

Exhibit 11

The study sees total Marcellus wells drilled growing from 1,121 wells in only two states in 2009 to 3,216 wells in three states by 2020.  New York drilling is estimated to account for roughly 10.5% of wells drilled in 2015 and 2020.  The gas production ramp up is much steeper as it is forecast to climb from 520 mmcf/d to 9,519 mmcf/d in 2020.  By 2020, New York gas production is estimated to account for about 10% of the total for the region.  The forecast suggests an attractive long-term outlook for the Marcellus.  Given the resource potential, it is an attractive prospective resource.

Exhibit 12

Should we be concerned about the economics of the Marcellus gas shale given its long-term outlook?  For those who have been around the energy investment business for a long time, we have seen highly touted exploration plays destroyed by poor economics in the past.  These plays such as the Deep Anadarko, the Austin Chalk and Bright Spots all ran into similar problems – optimistic production assumptions, expectations of higher oil and gas prices, well-managed drilling and producing costs and unlimited capital.  If anyone of these assumptions proves wrong, even for only a few years, the economics underlying the projects can be destroyed. 

The gas shale plays in this country have prompted leasing rushes with accompanying drilling booms.  That phenomenon has driven not only the domestic gas rig count higher than expected in light of low gas prices, but it has produced an increase in gas production.  Without a commensurate growth in natural gas consumption, ignoring the volatile temperature related demand, storage inventories continue to grow putting further downward pressure on gas prices.  These conditions have forced gas shale producers to resort to extreme measures such as borrowing substantial sums, selling non-shale assets (suddenly considered to be non-core), forming joint ventures with well-capitalized companies seeking representation in the gas shale plays and who generally have different investment rate of return criteria, and even selling parts of their plays.  We even see that a handful of gas shale producers are for sale. 

Are Marcellus gas producers all lemmings looking for a cliff?  A number of people are beginning to seriously question the gas shale phenomenon given the continuing, and projected to continue, low gas price outlook.  These critics recognize that many producers, especially the publicly-traded companies, are being pressured to engage in group-think by institutional investors.  In fact, several friends have compiled a list of reasons why this group-think exists and we list them below.  The authors of the list suggest it does not include all rationales and welcome any additions.

1. Bravely defend the leases
2. Add reserves
3. Grow production and be a good employee
4. Grow production to prolong the illusion that this is profitable
5. Grow your bonus and the value of your stock options
6. They made me do it (the investment bankers)
7. An investment in the future when gas prices are higher
8. Technology will save the day
9. Optionality
10. The land is the play
11. Greater fool theory: live another day to flip the company
12. Cash flow to pay debt service
13. Playing God: I think I’m flying
14. I can’t admit that I was wrong
15. No better ideas
16. Charles Prince at the Dance
17. Market share
18. Relatively low rate of dry holes (makes field operators look good)
19. Fear of litigation when the whole game ends
20. Continued access to capital
21. Peer pressure
22. Vast number of enablers (bankers, analysts, accountants, service companies, etc.). This is similar to “they made me do it”, but different in my mind
23. They have not yet run out of scenes for the moving circus (still more shales to declare “great”)
24. Government encouragement (allowance); seem as source of tax revenue, jobs
25. “I am an E&P company”; what else would you expect?”

Maybe the gas shale phenomenon will “muddle through” as we seem to be doing with our economy.  Unfortunately, the oil and gas business has a record of booms and busts – not always driven by poor investment decisions as economic downturns have often undercut demand just when long-lead-time supplies arrive at the market destroying prices and producer economics.  The troubling longer term economic outlook doesn’t hold how much hope for any sharp or near-term recovery in natural gas prices.  We go back to the chart on natural gas prices and focus on the level of pricing when the gas shale phenomenon got underway.  If gas prices stay at current levels, a fraction of the levels that existed when gas shale investments were accelerating, something has to give.  Will technology bail out poor economics?  Will gas shale producers become fodder for the largest oil and gas companies in the industry?  Maybe producers will finally decide to stop drilling.  We could easily spend many days spelling out various scenarios of how the gas shale phenomenon ends – many of them bad but some of them good!  As an investment analyst and trained to be critical, we can only offer words of caution at the present time.