Mark Anthony's  Instablog

Mark Anthony
Send Message
Mark Anthony, is an IT professional and who had a scientific research background before joining the information revolution. Visit his blog: Stockology (
My blog:
  • Did Haynesville Shale Production Increase In June? 4 comments
    Oct 5, 2012 5:09 PM | about stocks: ACIZQ, APA, BTUUQ, CHK, COG, EOGQ, KOL, SWN, UNG, WPX, ECA, APC

    People invested in the US coal (NYSEARCA:KOL) and natural gas (NYSEARCA:UNG) sectors watch natural gas (or NG) production level closely for signs that it has started to drop. Fellow SA author Richard Zeits suggested that natural gas (or NG) production from Haynesville increased in June, reversing the decline since January 2012, defying the slow drilling activity in the area. He was puzzled by the reversal and tried to explain why. I think I have a simpler explanation of the anomaly.

    I am not just doing an exercise to analyze some data. I think that investors should pick apart what they see at first sight and analyze what is under the hood. Arthur Berman often criticized the NG industry for over-estimating EURs (Estimated Ultimate Recovery) and profitability of shale gas wells. Investors should study the data themselves to avoid falling into salesman pitches.

    The Importance of Haynesville Shale

    Haynesville shale wells are the most productive among all US shale plays, although the well decline rates are also high. Its cost per well is also the highest as producers need to drill much deeper there. Haynesville was the top producing shale play last year. It was taken over by Marcellus only recently. So Haynesville is very important to US NG producers. See the EIA chart below:

    The well count and cumulative productions of major shale plays:

    (click to enlarge)

    Haynesville is one of 10 key plays of Chesapeake Energy (NYSE:CHK):

    (click to enlarge)

    Some NG producers active in Haynesville include:

    • CHK stopped most of drillings in Haynesville. For all dry gas plays, CHK started 2012 with 50 rigs, averaged 38 rigs in Q1. They will average 12 rigs in H2 of 2012, with 2 in Haynesville.
    • Exxon Mobil (NYSE:XOM) has a joint venture with EnCana in the area with 10,800 acreages. They also acquired XTO Energy (XTO).
    • EnCana (NYSE:ECA) drilled a 22,350 feet (record depth) Sabine well and two wells of 7500 and 8000 feet record lateral lengths.
    • Cabot Oil & Gas (NYSE:COG) was active there and acquired 24 wells recently. They sold some non-operated units in May.
    • Forrest Oil (NYSE:FST) owns 163,000 acres in the area with two rigs operating. It re-deployed the only Haynesville rig to work in oil-rich Cotton Valley interval in east Texas as well.
    • EXCO Resources (NYSE:XCO) has a gross production is 1.219 BCF per day. They had 22 rigs last year, cut to 7 in May 2012.
    • Comstock Resources (NYSE:CRK) completed 84 wells in 2011 and had 14 wells waiting for completion by year end.
    • PetroQuest Energy (NYSE:PQ) shifted its attention to Cotton Valley. A recently completed well had an IP of 3.8 MMCF/day of gas.
    • Goodrich Petroleum (GDP) participated in three non-operated new wells with impressive IP of 14.438 MMCF/day.
    • SM Energy (NYSE:SM) canceled drilling of four wells planned in 2012.
    • Kinder Morgan (NYSE:KMP) acquired $855M of Haynesville assets from PetroHawk Energy (NYSE:HK) in May 2011. In Q2-2012 they saw "below budget" production volume in the play.
    • Gastar Exploration (NYSEMKT:GST) stock lost -25.1% on Oct. 3 as CHK filed a lawsuit against the company.

    Richard's Discussion of Haynesville Shale

    Richard cited EIA data and this chart:

    (click to enlarge)

    The chart shows daily NG production rate of the Louisiana portion of the Haynesville Shale. The production peaked at the beginning of January 2012 and started to decline. But then it reversed to go up in June 2012. Why did it reverse?

    Richard thought that the relationship between active rig counts and the subsequent production volumes was broken. He explained the decline since January as due to producers shut in some gas wells due to low NG prices. He believed that by June, some producers decided to re-open some of the shut-in wells. Thus production increased again. If that was true, then we could see slow and gradual production decline in Haynesville in the near term.

    It is true that reduced drillings do not immediately lead to a drop in new well completions. But the correlation between well completions and production volume is so basic it could not have broken down, except for possible shut-ins as mentioned by Richard. But I think there is a simpler explanation of the June production anomaly.

    Take EIA Data with a Grain of Salt

    When looking at EIA data releases, people would think those are accurate and authorial numbers carefully reported by producers and tallied up by EIA. That is not the case. Most EIA data is calculated using math models based on raw data that may not be complete or accurate. Some data comes from private research organizations like BenTech who does not have official data collection authority. You can often find inconsistencies between different EIA data sources.

    For example, EIA releases weekly and monthly coal productions. The 8 month total for 2012 was 684.777M tons. The period was one day short of 35 weeks. The 35 weekly productions totaled 673.184M. Subtracting one day of 2.862M tons, it results in 670.322M tons. That is lower than the monthly totals by 14.5M tons! This large discrepancy is equivalent to 210 BCF in terms of NG. Investors would freak out when the NG storage number was only 2 BCF off their consensus. Why nobody has noticed the difference between the EIA weekly and monthly coal productions?

    When look at the EIA chart of NG gross withdrawal from Louisiana, I noticed that the June number was marked with an R that stands for "revised". From time to time, EIA may decide to revise a particular month's numbers, without necessarily revising the previous month as well. This might have distorted the month-to-month comparison. The month to month change could seem like an increase when it was actually a decrease. This often happens when you revise one number but do not revise the other number next to it.

    I do not know whether my explanation is correct or not. I do not know why EIA revised the June number, or by how much they had revised it. But I do know the Occam's Razor. The production drop early in 2012 might be fully accounted for by the slow down in well completions. If so, there were probably no production shut-ins. Thus there was no production re-opens in June, either. Let's find out.

    New Well Additions and Production Growth

    Shale wells are known to have steep production declines once they are brought online. NG producers keep drilling new wells to replace lost production due to declines. When producers add wells faster than needed to maintain flat production, the production will grow. If they add wells too slow, then the production will drop. So:

    Production Growth = (New Productions) - (Well Declines)

    With a reasonable well decline rate D and a new well IP (Initial Production rate), and with known well counts, I constructed a spreadsheet to calculate the total daily production rate and compare with the data. Here is the comparison:

    (click to enlarge)

    Knowing only the weekly well count, my calculation matched the production data perfectly. It shows that the production rate was accounted for by only the well additions and natural declines. It did not need contributions from well shut-ins and re-opens.

    (click to enlarge)

    The above shows how production is correlated to well completions. Note the total production includes 1.2 BCF/day of conventional gas production that declines much slower.

    The two parameters I used are:

    • Per well IP = 6.5 MMCF/day
    • Well decline rate D = -0.2%/day

    I am confident that the two number I chose are correct. If these two numbers were a bit higher or lower, the peak production would not have reached nearly 7.0 BCF as it did. More over, the way that the curve turned from growth to drop would not match.

    Significance of Decline Rate and Average IP

    Analyzing the profitability model of the NG industry is difficult. The cost and the productivity are not directly correlated. NG producers spend up-front money to drill gas wells and bring them to produce gas for decades. How much the wells can produce will not be known until well into the future. Analysts have struggled to understand the profitability models of NG producers. It does not help that producers love to pitch exceptionally good wells drilled at sweet spots. We have to look at the whole picture and examine the production history of shale plays in order to avoid the salesman pitches.

    We see NG companies fly all sorts of numbers of EURs, IPs and well decline models around. You never know how much you can trust those numbers, or how well they represent the average wells in a shale play. My simple method relies on only the historic production rates and well completion rates of a shale play. It provides accurate results of average IP and average decline rate D. Then IP divided by D tells you the average EUR per well. I urge people to study it.

    I have written several pieces about shale gas. I explained why the EUR estimates given by NG producers tend to be much higher than they should be. I scrutinized EURs from several different angles:

    • The earliest shale gas play, Barnett Shale, was proven to perform way below the industry's initial expectations, yielding only 0.663 BCF/well so far, versus the expected 2.5 BCF.
    • The Arps formula, when given a b-factor higher than 1.0, diverges to infinity. Thus it is not suitable for projecting long term shale well productions. But NG producers love to fit well production data with Arps type curves of high b-factor, and extrapolate the formula 40-50 years into the future. They do so to obtain high estimates of EURs.
    • Based on reasonable recovery factors, Marcellus and other shale plays have lower EURs than the NG industry projected.
    • The production rates of a shale play, when correlated to the well completion rates, as discussed here, provide a more realistic metrics to calculate IP, decline rate, and EUR. Such results are much lower than projected by NG producers.

    I welcome more debates on the controversy on shale gas.

    Investment Implications

    The decline in Haynesville production was real. It was the result of slowing rig activity and subsequent slowing of well completion rate. Investors do not need to be pessimistic about the NG price outlook. I believe that even Marcellus Shale is beginning to decline soon.

    My calculation shows we need 2 new wells per day to maintain flat production at Haynesville. Using my method, the EUR roughly equals to IP divided by decline rate D. EUR = 6.5 / 0.2% = 3.25 BCF/well. Of course that is only my take. CHK projected a much higher EUR of 6.5 BCF. That is a huge difference that could decide if Haynesville shale is profitable at $4 or $8 per mmBtu.

    The debate on shale gas EURs and profitability is important. There is a widely accepted belief that the so-called shale gas revolution is a paradigm shift that kills coal. People believe coal is being replaced by cheap and abundant shale gas. This false myth is the reason why the investment community dedicated 75 times more money in the NG sector than in coal. The NG industry has accumulated more than half a trillion dollars of debts since it first ventured into shale gas development. I believe that there is a debt crisis looming in the NG industry. I believe people have invested in the wrong energy sector.

    I emphasized many time in the past that natural gas is bullish, as is coal. NG supply and demand is rebalancing just as I predicted. I also said that the coal sector is the better sector to leverage on rising prices of coal and NG. The reason is that coal producers are not far away from seeing profitable coal prices, but shale gas is still deeply non-profitable for NG producers. I will stick to my coal stocks.

    But I do not think that shale gas will be dead. It is an important part of our energy supply. But NG prices must go much higher to allow shale gas to be produced profitably. NG producers with low debt loads and thrift well drilling costs would survive and thrive.

    I like Southwestern Energy (NYSE:SWN) after looking at its balance sheet. Its debt was $3.988B in Q2-2012. That is low by industry standard. It did incur high capital spending each quarter. But that did not result in high debt load. I don't know how they operated so efficient without borrowing a lot of debts. I would study this company more when I have time, and write about it.

    COG is another NG producer I would like to exam. Its Q2-2012 revenue was $265.657M on $66.695M cost. How could its cost be so low relative to its revenue? I wish to study more on that.

    But again, now is time to stick to coal companies, not NG producers. Opportunity in the NG sector comes after the looming debt crisis in the sector unfolds and takes its preys. We are not there yet.

    The above discussions are also relevant to these equity issues:

    • NG players: Southwestern Energy [SWN]
    • Anadarko Petroleum Corp. (NYSE:APC)
    • SandRidge Energy (NYSE:SD)
    • Apache Corp. (NYSE:APA)
    • EOG Resources (NYSE:EOG)
    • WPX Energy Inc.(NYSE:WPX)
    • Coal players: Arch Coal Inc. (ACI)
    • Alpha Natural Resources (ANR)
    • James River Coal (JRCC)
    • Peabody Energy (BTU)
    • ETFs: United States Natural Gas [UNG]
    • Market Vector Coal ETF [KOL]

    Disclosure: I am long ANR, ACI, JRCC, BTU. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Back To Mark Anthony's Instablog HomePage »

Instablogs are blogs which are instantly set up and networked within the Seeking Alpha community. Instablog posts are not selected, edited or screened by Seeking Alpha editors, in contrast to contributors' articles.

Comments (4)
Track new comments
  • Jarinek84
    , contributor
    Comments (10) | Send Message
    Nice article Mark, thanks. Looking forward to your take on above mentioned NG companies.
    6 Oct 2012, 05:08 AM Reply Like
  • Mark Anthony
    , contributor
    Comments (3595) | Send Message
    Author’s reply » Thanks. This was meant to be a formal SA article. They did not accept it for publication so I have to post it as an instablog post.


    I have another article speficially discuss QEP's Haynesville wells. Hope they publish that one. Stay tuned.
    6 Oct 2012, 06:09 AM Reply Like
  • pete123
    , contributor
    Comments (877) | Send Message
    Interesting! Appreciate all your work, even though I disagree a lot :-)
    Especially like the focus on Haynesville and Marcellus.
    6 Oct 2012, 09:57 AM Reply Like
  • Mark Anthony
    , contributor
    Comments (3595) | Send Message
    Author’s reply » I may revise the projections with a more accurate set of well completion counts from this source:



    I stand correct on my projection of Haynesville production.
    4 Dec 2012, 02:07 PM Reply Like
Full index of posts »
Latest Followers


  • Jim Rogers favors water investment! California is in a water crisis. Cardiz (CDZI) is up almost 50% today on news:
    Jun 5, 2009
  • China is still buying US T-Bonds and Yuan is pegged to dollar. WHY? Read my shocking discovery: Enjoy!
    Jun 4, 2009
  • PEIX, a once hot ethanol play, goes Ch.11 today, as expected. Recent unusual price moves must lured speculators in. Good lesson learned.
    May 18, 2009
More »

Latest Comments

Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.