How Much Natural Gas Do We Have Left?

President Obama proclaimed that the US has 100 years of natural gas (UNG) supply.

Do we really have 100 years of natural gas (NG) supply? Why did Chris Nelder claim that we have only 11 years or less NG supply left? The Potential Gas Committee, the EIA and the USGS gave different estimates of US NG resources. People may interpret the numbers wrong. They may not understand the differences between resource, reserve, and economical reserve.

The PGC claimed we have 2192 TCF of discovered and undiscovered potential NG resources. Marketed NG production was 22 BCF in 2010. So 2192 TCF divided by 22 BCF/year is about 100 years of supply.

I will show that it is naive to jump to a conclusion based on that.

The Difference Between Resource and Reserve

In the oil & gas industry, resource means the amount of gas or oil that remains underground, and reserve means what could be produced from the resource.

Only a portion of the resources could be recovered technically.

Only a portion of the technically recoverable resources could be produced economically.

Only a portion of the economically producible resources could be produced into supply. That is called reserve.

This graph explains what are non-discovered and speculative resources, and what are discovered and proven reserves:

Please read Chris Nelder for details. We extracted 28.6 TCF of gas in 2011. So 273 TCF of proven gas reserve only lasts for ten years. How do NG industry experts and geologists estimate the resources and reserves? I found that these two groups calculate things differently. Investors should know why they calculate the estimates differently.

The Marcellus Shale counts for about half of US shale gas reserves. Recently USGS slashed reserve estimate of Marcellus from 410 TCF to 84 TCF. In response EIA also revised their number to 141 TCF.

Let me do a case study on Marcellus Shale.

The Top-Down Reserve Estimate of Marcellus

This is the preferred approach of NG producers:

• Drill some demonstration wells to get some production data.
• Use the Arps formula to fit a few months' production data.
• Extrapolate the type curve to 40-50 years of well lifespan to calculate an EUR (Estimated Ultimate Recovery).
• Extrapolate the results of the demo wells to the whole area.
• Calculate number of wells that could be drilled based on well spacing and area of the play. Multiplied it by the average EUR, then multiply it by an effective recovery percentage.

This EIA document explains how the estimates were obtained:

The typical well spacing is 2000 feet apart, or 80 acres per well. The unit conversions are:

• 1 mile = 5280 feet
• 1 square mile = 640 acres
• 1 acre = 43560 square feet
• 80 acres spacing = 8 wells/SQM

Here is how EIA came up with 410 TCF for Marcellus (page 21):

• The area of the Marcellus play is 94893 square miles.
• At 8 wells per square mile, 759144 wells can be drilled.
• The average EUR/well is 1.18 BCF. So total is 896 TCF.
• Discount it by an effective recovery factor of 45.8%.
• The result is 410 TCF technically recoverable gas.

I think this approach is flawed:

1. Fitting the Arps Type Curve based on merely the first few months of production data and extrapolating to 500 months of production is wrong. It leads to overestimated EURs. The Arps formula diverges to infinity with B>1 as I explained.
2. Producers drill demo wells with more laterals of longer length. The laterals could extend up to 9000 feet away. So they are draining gas from beneath 4 or 5 well sites away. In doing so, they can show higher production rates per well. But the data does not represent the production potential for the area fairly. When they drill the next well on the next 80 acres patch, they may discover that the gas was already drained.

Following the USGS revision, EIA slashed the Marcellus reserve from 410 TCF to 141 TCF. That invalidated the previous calculation. The USGS geologists have a more scientific approach.

The Bottom-Up Reserve Estimate of Marcellus

A geologist would estimate the reserve as below:

1. Calculate the Gas-in-Place (GIP) based on the geology: The thickness of the shale layer and the gas content per volume or per ton of the shale rocks.
2. Calculate percentage of recoverable GIP.

This is reasonable to me. I checked this overview with a chart on page 8 giving parameters of the Marcellus Shale:

• Total area of Marcellus is 95000 square miles
• The thickness of the shale layer is 50 to 200 feet.
• The gas content is 60-100 CF per ton of material.
• Geologists calculated the GIP to be 1500 TCF.

The density of shale rocks is 2.5 tons/M³ or 0.0708 tons/CF. So Average gas content is 5.664 times the shale volume. Using 100 feet average thickness, GIP = 5.664*95000*640*43560*100 = 1500.08 TCF. My math is good!

Halliburton (HAL) disclosed that only 10% of GIP could be recovered. Various sources put the recovery factor at 5% to 10% or less. But somehow Chesapeake (CHK) got a 30% recovery factor magically.

Assuming a 9.4% recovery factor, the 1500 TCF of GIP in Marcellus would produce 141 BCF technically recoverable gas. That's the new EIA estimate. Chesapeake's mileage may vary.

Calculating the EUR Per Well

Let's calculate the EUR per well, using the upper bound numbers: 100 cubic feet of gas per ton of rocks, and 200 feet shale layer.

GIP = 100CF/ton*0.0708ton/CF*80Acres*43560SQF/Acre*200FT

The result is GIP = 4.93 BCF per well. Assuming 10% recovery, the EUR would be 0.5 BCF per well.

Even CHK's optimistic 30% recovery factor gives EUR of 1.5 BCF. That is way below the 3.75 BCF in CHK's type curve.

In Barnett Shale, the GIP was estimated to be 327 TCF. The NG industry projected an EUR of 1.42 BCF and recoverable reserve of 44 TCF. So far, they drilled 16000 wells and developed 66% of the 3000 square miles core area. The cumulative production is only 10.8 TCF. That's 3.3% GIP recovery factor and 0.675 BCF per well. The GIP of the thickest core of Barnett is 52 BCF per well, BTW.

How could producers ever make a profit at that kind of EUR? They quietly retreated from Barnett and switched to other shale plays, without admitting defeat. The gas content in Barnett averaged 320 CF/ton. That is four times better than Marcellus, and the highest of all US shales. That must be the reason Barnett was developed first. Do they expect success in Marcellus with 1/4 the gas content?

How could they place wells only 2000 feet apart, when the wells were drilled with 9000 feet or longer laterals?

If the well spacing was increased, then fewer wells could be drilled. Thus the resource assets of producers would value less. Likewise, with a lower EUR, the asset values must be written down.

It is much worse! A low EUR means a shale play might not even be profitable at reasonable NG prices. When an asset could not bring in profits, should its value not be marked down to zero?!

Implications to Investors

Investors check balance sheets and quarterly operation reports to determine a company value. Knowing the real reserve estimates and the realistic EURs, what do you see when you look at the balance sheets of major NG producers?

What happens to the equity values when the assets on the books must be marked down by half or more, or reduced to near zero?

I see grave danger in the US NG industry. They told us the shale gas boom story for years. It is time for the investors and the NG industry to wake up to the reality. The industry has survived so far by keep begging banks for more money to keep drilling. For example, CHK earned $1.068B in gas and oil revenue in Q1 of 2012 while spent $4.4B in capital expenditures.

This is unsustainable! I foresee a looming collapse of the US NG sector. In my follow-up articles I will discuss in details. I urge people to scrutinize the financials of these names:

• Chesapeake Energy [CHK]
• Anadarko (APC)
• Devon Energy (DVN)
• EnCana (ECA)
• ConocoPhillips (COP)
• Southwestern Energy (SWN)
• Chevron (CVX)
• WPX Energy (WPX)
• EOG Resources (EOG)
• Occidental (OXY)
• Apache Corp. (APA)
• Ultra Petroleum (UPL)
• QEP Resources (QEP)
• Cabot Oil & Gas (COG)
• EQT Resources (EQT)
• Exco Resources (XCO)
• Range Resources (RRC)
• Newfield Exploration (NFX)
• Noble Energy (NBL)
• Pioneer Natural (PXD)
• Marathon (MRO)
• Quicksilver Resources (KWK)
• Forest Oil (FST)
• Linn Energy (LINE)
• Energen Resources Corp. (EGN)
• SandRidge Energy (SD)

Invest in Coal, Not Natural Gas

Mean while, deeply discounted US coal producers are poised for a strong rebound. There is a myth that natural gas is now dirt cheap and abundant, and coal is dirty and expensive and abandoned. Some thought the entire coal sector is doomed, and that the NG sector will thrive.

Nothing is further from the truth. Coal is not going away. The shale gas boom will turn to bust. Amid surging electricity demand, reversal of the coal-to-gas fuel switch and rapid growth of exports, US coal producers continue to curtail productions, which is bullish:

China's coal imports in the first half of 2012 reached 140M tons, up 65.9% Y-o-Y. That's bad news to Chinese coal producers but good news to the oversea producers. I predicted that the Chinese coal industry faced peak coal, peak water and peak labor. Recent data suggests that other than the top three producing provinces, productions in other provinces are collapsing. I will discuss the black lung in Chinese coal miners and more in my future articles.

I call to short certain NG players after the NG prices recovery, but not now. After the Patriot Coal (PCX) bankruptcy, I increased my positions in James River Coal (JRCC), Arch Coal Inc. (ACI), Alpha Natural Resource (ANR) and Peabody Energy (BTU). I encourage readers to consider other US coal producers as well:

• Cloud Peak Energy (CLD)
• Consol Energy (CNX)
• Black Hills Corp. (BKH)
• Walter Energy (WLT)
• Westmoreland Coal (WLB)
• Nacco Industries (NC)
• Alliance Resource Partners LP (ARLP)
• Market Vectors Coal ETF (KOL)

Disclosure: I am long JRCC, ACI, ANR, BTU.