I have written about the shale gas controversy. I questioned whether shale gas will ever be profitable. I predicted a looming debt crisis in the natural gas (as NG) sector due to capital destruction. I believe that NG producers tend to over-estimate EURs (Estimated Ultimate Recovery) of shale gas wells. If critics like Arthur Berman were correct, we have much less shale gas left to be produced.
Once again, Seeking Alpa declined to publish my well researched article on shale gas. So I am posting here of the main point in a reduced form.
SA author Richard Zeits alerted me when he noticed that production of Haynesville shale probably increased in June. He tried to explain it. He referenced a BG Group presentation and cited a presentation by QEP Resources (NYSE:QEP) for support.
I disagreed with Richard's explanation.
Impressive Haynesville Shale Well Performance
QEP claims their Haynesville wells performed exceptionally well, with high IP and slow decline rate. See their Barclay presentation:
An EUR of 6 to 8 BCF?! Their Haynesville type curve below:
QEP used a low b = 0.4 and a very low initial decline rate D.
b-factor = 0.4
D = -ln(1-50.1%)/year = 0.695/year = 0.05793/month
Chesapeake Energy (NYSE:CHK) type curve has a high b value and a very steep initial decline D:
The Arps formula is explained below:
CHK used 500 months well life to calculate EUR, with b = 1.351 and D = 0.75/month. Very different from QEP's.
Who do you want to believe, CHK or QEP?
Are These Haynesville Wells Really So Good?
I decided to scrutinize QEP's actual production data to find out. I need production rates and well additions.
Here is my calculation of the total production rate, assuming the wells comply with the QEP defined type curve:
Oops, the result does not match the actual production! The wells did not follow QEP's type curves of IP=8.25MMCF/day and slow initial decline rate D (losing only 50% in a year).
QEP operated 126 producing wells in Haynesville in Q2, with an average production rate of 340 MMCF/day. The average per well was 2.7MMCF/day. Consider that half of the wells are less than one year old, the 2.7 MMCF/day average is inconsistent with an IP of 8.25 MMCF/Day and first year decline of only 50%.
Let me keep b = 0.4 but use higher D so first year decline = -81% as claimed by CHK, which is D=0.2/month. So I used these numbers to plot the chart again:
Bingo! I got a perfect match with the actual production!
I should explain that to obtain the above precise match, I did tweak the IPs a bit as following:
- From Q1, 2009 to Q4, 2010, the IP = 8.25MMCF/day, except:
- In Q1, 2010, the IP = 7 MMCF/day; In Q3, 2010 IP was only 6.
- In Q1-Q4 of 2011, the IPs were 10, 5, 9, 11 MMCF/day.
- In 2012 the IP reached 12 MMCF/day.
The above IP assignments are inline with QEP's claim that they have achieved higher IPs in the most recently drilled wells.
The Real Performance of QEP's Haynesville Wells
I have modeled QEP's well declines to match actual production data successfully with these numbers:
- Initial decline rate D = 0.2/month, or -90.93% annually.
- First year production decline is -81.4%. This agrees with what other NG producers have been saying about Haynesville.
- At IP = 8.25 MMCF/day, 300 month EUR is 2.046 BCF/well.
- At IP = 12 MMCF/day, the EUR is at 3.0 BCF/well. At higher IP, the EUR will be proportionally higher.
My EURs are far lower than predicted by QEP. They gave 6-8 MMCF per well. If my 2-3 BCF/well is correct, the $9M per well drilling cost suddenly looks prohibitively expensive!
Do you believe QEP or do you believe me? The math does not lie!
The statistics of the entire Haynesville also agrees with my EUR:
There were 1402 production wells in LA Haynesville. Accumulated production was 4086.233 BCF, or 2.9 BCF/well. This average agrees with my EUR of 3.0 BCF much better that QEP's 6-8 BCF.
I also calculated the combined decline rate of Haynesville based on production and well completion statistics. Here is the result:
I obtained average IP of 6.5 MMCF/day, and combined decline rate of the whole Haynesville at -0.2%/day. So EUR = IP/D = 3.25BCF.
Why the EUR Is Important to Both NG and Coal Companies
I attempted to obtain a realistic EUR of the Haynesville shale wells using different methods. All of my estimates suggest that EUR is at slightly above 3.0 BCF/well. This is far lower than NG producers have been projecting. I believe my result has much stronger data support.
This is important. Profitability of shale gas plays depends on not just the gas price, but also how much gas could be produced from a well. Whether the real EUR at Haynesville is 3 BCF or 6 BCF makes a huge difference to the producer bottom lines.
The same question is even more important to coal mining companies. When the NG price is cheaper than $3/mmBtu, NG competes with coal for electricity generation.
As long as NG producers can not make a profit at a NG price that is competitive against coal, the king coal is not dead. My research in different shale plays shows that the real EURs are far below what NG producers tried to make us believe. Here, once again, I showed that the producers own production data contradict their own projections.
Shale gas can NOT be produced profitable and compete against coal. Coal is not dead. NG prices must go a lot higher. Can gas price realistically go high enough to allow NG producers to make a profit? I do not believe so.
I believe investors should divest in the NG sector and invest in coal. If you hesitate, then consider these factors:
- Current coal prices are very close to the profitable levels.
- Current NG prices are already non-competitive against coal. Yet they still do not allow NG producers to make a profit.
- Almost the entire investment community made a big mistake thinking that there is a paradigm shift of NG replacing coal.
- Investors put 75 times more money in the NG sector than the coal sector, while the two generates the same amount of energy and are equally important to the US economy. This is probably the biggest portfolio imbalance in the energy investment history.
There are not many US coal companies to buy. My favorites are:
- Alpha Natural Resources (ANR)
- Arch Coal Company (NYSE:ACI)
- James River Coal Company (JRCC)
- Peabody Energy (NYSE:BTU)
The discussions are also relevant to these issues in the NG sector:
- EnCana (NYSE:ECA)
- Cabot Oil & Gas (NYSE:COG)
- EOG Resources(NYSE:EOG)
- Southwestern Energy (NYSE:SWN)
- Anadarko Petroleum Corp. (NYSE:APC)
- SandRidge Energy (NYSE:SD)
- Apache Corp. (NYSE:APA)
Disclosure: I am long ANR, ACI, JRCC, BTU.