Shale Gas EROI: Preliminary Estimate Suggests 70 Or Greater

A valuable measure of a fuel’s usefulness and long-term viability is its energy return on (energy) investment (EROI). This is the ratio of the energy obtained from using that fuel to the energy invested to bring that fuel to its point of use.[1],[2],[3] Back in the early days of petroleum and natural gas production, when wells were shallower and readily accessible by land routes, EROIs were probably in the range of 100 – that is, a well would return 100 units of energy for each 1 unit of energy it took to drill it and bring the product to market.

To the extent that the growth of industrial society has been supported by readily available and cheap energy (i.e., fossil fuels with high EROI), industrial economies will be increasingly stressed as easily extracted fuels are used up and replaced by fuels with lower EROI. Some analyses suggest that an EROI greater than 5 to 10 is necessary for even a limited functioning of industrial civilization and indicate that many of the newer oil wells in difficult locations, e.g. deep seas, have EROIs in the range of 10.[4]

New methods of extracting natural gas from organic-rich shales, such as the Barnett in Texas and the Marcellus in Pennsylvania and nearby regions, appear to offer promise of a large new source of natural gas. Geologist Ken Deffeyes who several years ago made what increasingly appears to be an accurate prediction that global petroleum production would peak somewhere between 2004 and 2008, regards natural gas from shale as a game-changing opportunity.[5] He considers opposition to new horizontal drilling and hydrofracking procedures as “evidence of economic suicidal tendencies.”[6] Deffeyes is a knowledgeable geologist (long ago, he was one of the early proponents of a then-controversial theory – plate tectonics). Others share the optimism; huge amounts of capital are flowing into the new shale gas plays.

But some wonder if the potential of shale gas is overblown. Geologist Arthur Berman has argued that shale gas is not economic to produce unless the wholesale price of gas rises above $7 per million Btu.[7] Others have made similar arguments.[8] . A recent report argues that the level of effort to capture significant amounts of gas from shale is so large that it is unlikely to happen, especially if the price of natural gas remains at or below the break-even point, which this report indicates is likely in the range of $4.20 to $11.50 per million Btu.[9]

The key to the future of shale gas is its EROI. I’ve been unable to find estimates of the EROI of shale gas in the literature. However, I’ve made a preliminary first-order[10] estimate that the EROI of shale gas is in the range of 70 to greater than 100. This is probably significantly better than most other energy sources available today.

This estimate is based on my interpretations of analyses by the Environmental Defense Fund[11] and the New York Department of Environmental Conservation (NYDEC)[12] which focused on the carbon dioxide (CO2) emissions from shale gas drilling and compressing operations. CO2 emissions are directly related to fossil fuel combustion, so these studies in effect provide estimates of the energy used to extract shale gas and get it to market. Other studies provide estimates of the ultimate production of gas from an average well[13] and on the portion of the gas that must be used to process and compress it and send it through pipelines.[14] Also included were approximate estimates of the energy it took to make the steel used for well casings and a portion of the necessary pipelines, and the concrete used in the casing process, which were apportioned based on assumptions.

The NYDEC study looked at the main tasks involved in drilling and hydrofracking, including 1) site mobilization, construction and demobilization, 2) well drilling, 3) transportation of water, etc. necessary for hydrofracking, and 4), the hydrofracking process. Using activities on well drilling sites and estimated CO2 emissions based on equipment emission factors and times of operation, this study estimated a range of CO2 emissions for each of these and several smaller tasks, depending on whether the well was near or far from necessary materials, water, etc. I chose approximate average to high-end values for these tasks of 100, 95, 400, and 325 tons of CO2, respectively. Adding in several smaller tasks as well resulted in a total CO2 emission to drill and hydrofrack a well of approximately 940 tons. Since all of this work is typically powered by diesel engines, this emission can be converted with standard conversion factors[15] to Btu consumed in the form of diesel fuel. It translates to 11.6 billion Btu.

The EDF study also inventoried activities on well drilling sites and estimated CO2 emissions based on equipment emission factors and times of operation. Its estimate, provided on a daily basis for 1000 wells completed per year, translates to approximately 1450 tons per well completion. This figure translates to 18.4 billion Btu.

These values average 15 billion Btu. To this I added 2.8 billion Btu for the embodied energy in the steel used for the well casing,[16] 1.2 billion Btu for the embodied energy in the concrete used for the well casing,[17] 1.5 billion Btu for the embodied energy of the trucks, pumps, and other equipment used in the drilling and hydrofracking process,[18] and 10 billion Btu for the embodied energy of the steel used for a portion of the pipeline necessary to transport the gas.[19] All of these embodied energy estimates involve a number of assumptions and are subject to much uncertainty and variation from well to well, but I doubt the uncertainty of any of them is more than a factor of two. The total of all these energy costs to construct a shale gas well and get its production to market is approximately 30 billion Btu.

Another, quite different approach is to estimate the total cost of a shale gas well and then use the average amount of energy associated with a dollar of gross domestic product (GDP) to translate this cost to an energy value.[20] In 2010, U.S. GDP was about $14.5 trillion, and the nation used about 100 quadrillion Btu. This translates to about 7000 Btu of energy expended per dollar. Assuming that the energy expended in drilling and hydrofracking a shale gas well bears the same relative relationship to the dollar, the approximately five million dollar estimated cost of a well and associated infrastructure[21] translates to an energy cost of 35 billion Btu.

These energy cost values must be compared with the total energy expected to be produced by an average shale gas well. There are now enough data on wells from the major shale regions to provide such an estimate. A cumulative production estimate[22] for a typical Marcellus shale well for a 10-year period of 2.11 billion cubic feet was extrapolated to a 25-year period, yielding an estimate of approximately 2.9 billion cubic feet. This translates to approximately 2.9 trillion Btu. Other estimates suggest typical total production from Marcellus wells may in the range of 5 trillion Btu.[23] This ultimate production must be reduced by 8% to account for the approximate percentage of gas that is consumed to process and compress the gas and move it through pipelines to consumers. [24]

The estimated total energy cost of shale gas extraction is thus in the approximate range of 30 to 35 billion Btu while the estimated ultimate energy produced is in the range of 2.6 trillion to nearly 5 trillion Btu. The ratio of energy produced to energy expended for shale gas based on the approaches outlined above is thus at least 70 and perhaps well over 100. This is extremely good relative to the probable EROI values for other current energy sources.

This relatively high EROI of shale gas has several implications:

1) Shale gas is not a speculative bubble that will go away. Its favorable energy balance means that economics will inexorably drive the extraction of gas from shales, especially as supplies of petroleum grow tighter. The number of wells drilled will continue to grow, as will associated truck traffic and other activity. To the extent that appropriate regulations are not put in place and enforced and/or that voluntary best management practices are not followed, damages to the landscape and pollution events are inevitable.

2) Natural gas will be in more plentiful supply than petroleum in the years to come; businesses and infrastructure that are dependent on petroleum are likely to look for ways to convert to natural gas. However, the potential size of the shale gas play should not be overestimated. It is not likely big enough to replace dwindling and ever-more-expensive petroleum on a large scale.[25]

3) Dangers of excessive regulation threatening the development of the nascent shale gas industry are probably overblown. Shale gas companies should be able to afford to adopt and enforce best practices for all that they do. Unless it can be conclusively proven that the entire industry needs protection from unnecessarily stringent regulations, a gas company’s arguments for regulatory leniency should be considered as nothing more than advocacy for that company’s own profitability.

While the embodied energy of basic materials and machinery involved in the drilling and hydrofracking processes have been considered in this analysis, some energy costs have not been considered. These include the embodied energy of labor and associated support and infrastructure (e.g. workers’ vehicles, energy costs of housing and food for workers, etc.). Further, energy costs of remediating pollution and other problems that could result from shale gas extraction have not been considered. Also, costs of impacts to resources such as water supply, while not directly comparable to energy costs, are relevant to a deeper look at EROI.

Some potential problems include:

1) Large amounts of gas may leak from extraction operations. Although there is much uncertainty, recent work suggests that such leakage of natural gas, because of its relatively high global warming potential, could be large enough to nullify the benefits of natural gas vs. coal and petroleum from a global warming perspective.[26]

2) Surface waters may be polluted by spilled or improperly treated flowback fluids and drinking water wells may be contaminated with chemicals used in the hydrofracking process or created in-situ as byproducts of this process. Several perceptive discussions of these potential impacts are available.[27],[28]

3) Gas itself, finding its way into aquifers from nearby wells, may contaminate drinking water. A recent study found levels of gas from nearby wells high enough to present explosion hazards.[29]

4) More drilling activity will generate more traffic on rural roads, resulting in more noise, air pollution, safety risks, and generating a need for road and other maintenance and improvements.

5) More drilling activity will fragment vast stretches of contiguous forest. (See photo of drilling sites in western PA.) Loss of contiguous forest is likely to accelerate the decline of many species of wildlife including neotropical migrant songbirds.


Serious efforts towards improving the efficiency of natural gas use could reduce the pressure to extract more gas and thereby reduce the incidence of negative impacts. (See, for example, earlier posts on this site on increasing home heating efficiency.)

Increased production of natural gas from shale is also likely to have positive impacts, including the creation of jobs and the flow of more money into rural areas.

If negative impacts can be controlled with best management practices, which will likely require appropriate and well-enforced regulations, shale gas could help maintain rural communities and ameliorate, to some degree, growing energy supply problems. Especially important, both from a global warming and from a safety perspective, appears to be minimizing gas leakage. It also seems critical that new supplies of natural gas not be squandered through wasteful usage; environmental costs resulting from shale gas will be lessened to the degree that less gas is used due to increased energy efficiency.

References

[1] While EROI is the key to a resource’s energy usefulness, there are important aspects to a resource’s costs that are not considered. These include its renewability, environmental impact, its size, and the need for ancillary resources and materials. For a thorough discussion of EROI and net energy see Heinberg, 2009, referenced below. . Also see Mulder & Hagens, 2008, referenced below.
[2] Heinberg, Richard, 2009, Searching for a Miracle: “Net Energy” Limits and the Fate of Industrial Society, http://www.postcarbon.org/report/44377-searching-for-a-miracle
[3] Mulder, K., and N. Hagens, 2008, Energy return on investment: Toward a consistent framework, Ambio, 37, 74-79.
[4] Hall, Charles, 2008, Why EROI Matters, The Oil Drum, http://www.theoildrum.com/node/3786
[5] Deffeyes, K., 2010, When Oil Peaked, Hill and Wang, NY, p.107
[6] http://www.princeton.edu/hubbert/current-events.html
[7] http://petroleumtruthreport.blogspot.com/
[8] http://baobab2050.org/2010/10/28/shale-gas-miracle-or-mirage/
[9] Hughes, J. David, 2011, Will Natural Gas Fuel America in the 21st Century, Post Carbon Institute, www.postcarbon.org/report/331901-will-natural-gas-fuel-america-in
[10] See Mulder & Hagens, 2008, for a further discussion of this term and of the methodological issues in EROI determination
[11] Armendariz, A., 2009, Emissions from Natural Gas Production in the Barnett Shale Area and Opportunities for Cost-Effective Improvements, prepared for Alvarez, Ramon, Environmental Defense Fund, Austin, TX, January 26, 2009, http://www.edf.org/documents/9235_Barnett_Shale_Report.pdf
[12] NYDEC, 2009, DRAFT Supplemental Generic Environmental Impact Statement on the Oil, Gas and Solution Mining Regulatory Program, NY Department of Environmental Conservation, Albany, NY, http://www.dec.ny.gov/energy/58440.html
[13]Harper, John, and Jaime Kostelnik, PA Geological Survey, The Marcellus Shale Play in Pennsylvania, http://www.marcellus.psu.edu/resources/PDFs/DCNR.pdf, accessed 6/15/11
[14] U.S. Energy Information Administration (EIA), 2011, Natural gas consumption by end use, http://www.eia.gov/dnav/ng/ng_cons_sum_dcu_nus_a.htm. The quantities used for “lease and plant fuel” and “pipeline and distribution” in 2010 represented 8.3% of total consumption.
[15] EIA, 2011a, http://www.eia.gov/oiaf/1605/emission_factors.html
[16] Wikipedia, 2011, “Embodied Energy,” http://en.wikipedia.org/wiki/Embodied_energy, and references therein
[17] Wikipedia, 2011
[18] Stodolsky, F., A. Vyas, R. Cuenca, and L. Gaines, 1995, Life-Cycle Energy Savings Potential from Aluminum-Intensive Vehicles, Argonne National Laboratory, Argonne, IL 60439. See analysis at http://michaelaucott.blogspot.com/2010/08/still-in-service-why.html, which uses this report to estimate that the embodied energy of vehicles and other machinery represents approximately 10% or less of the energy needed to operate the unit over its lifetime. This same percent of the energy used during the drilling, etc. processes was assumed to represent the energy expended in the form of embodied energy of vehicles and other equipment.
[19] Value for steel is from Wikipedia, 2011; M. Aucott assumed for this analysis that 10 miles of 20” pipeline would be installed, but that this would serve 10 wells. Energy expended for construction of pipeline was ignored. Pipelines may serve many more than 10 wells, and could last for longer than the lifetime of one well, which would lower the apportioned energy expenditure. There is considerable uncertainty with this figure.
[20] Hall, C., and M. Lavine, 1979, "Efficiency of Energy Delivery Systems:1. An Economic and Energy Analysis", Environmental Management, vol 3, no 6, pp 493-504, 1979 (First part of a 3 part article), as referenced in “North American Natural Gas Production and EROI Decline” from http://www.theoildrum.com/node/3673.
[21] Harper and Kostelnik, PA Geological Survery, 2011; $5 million dollar figure is sum of estimated cost of a Marcellus well from this the figure “Comparisons of Four Major Shale Plays” from this reference ($3.5 million) plus additional $1.5 assumed by M. Aucott to approximate the cost of associated pipelines.
[22] Harper and Kostelnik, PA Geological Survey, 2011
[23] Vanderman, Kris, 2011, Penn State University webinar, 4/21/11.
[24] EIA, 2011
[25] Hughes, 2011
[26] Howarth, R., R. Santoro, and A. Ingraffea, 2011, Methane and greenhouse-gas footprint of natural gas from shale formations, Climatic Change, http://graphics8.nytimes.com/images/blogs/greeninc/Howarth2011.pdf
[27] Penningroth, Steven, 2010, http://yosemite.epa.gov/sab/sabproduct.nsf/A4105736E3A173AC85257703006B8648/$File/Pub+Comments+by+S+Penningroth+Ithaca+NY4-7-10+for+EEC+Apr+7-8+2010+Meeting.pdf
[28] http://www.energybulletin.net/stories/2011-06-20/forum-just-how-safe-%E2%80%98fracking%E2%80%99-natural-gas
[29] Osborn, S., A. Vengosh, N. Warner, and R. Jackson, 2011, Methane contamination of drinking water accompanying gas well drilling and hydrofracking, PNAS, http://www.pnas.org/content/early/2011/05/02/1100682108

VMT Trend Flat - the End of an Era?

Vehicle miles traveled (VMT) are strongly linked to the American way of life, and correlated with economic activity. From the 1930s until recently, with a few interruptions, VMT in the U.S. have grown consistently.

But around 2005, well before the recent economic downturn, an inflection point in the long-term trend appears to have occurred. Since then, the VMT trend has essentially been flat. This is a change of potentially historic proportions.

The 2005 growth cessation coincided with the price of gasoline rising above $2.50 per gallon. As the chart below shows, there have been only two other periods since the 30s when the real price of gasoline rose above $2.50 per gallon, the World War II era and the mid-70s through the early 80s. These periods also saw flat or declining VMT.

The current period of flat VMT is the longest since WW II, when gasoline was rationed. With gasoline prices again nearing $4.00 per gallon, it is unlikely that VMT will increase in 2011. It would not be a surprise if prices increase, and if a decline in VMT sets in. We may be witnessing the end of an era.

Heating Index: How does your house perform compared to other houses? Part 2

How does your house compare with other houses in heating efficiency? As outlined in the previous post, figuring out how much energy your house uses for heating is relatively easy. An average house will use 10,000 to 15,000 Btus per heating degree day (HDD). But this measure alone doesn't reveal how your house ranks on the scale of heating efficiency, and doesn't help determine if there is lots of room for improvement or if you've got your house about as efficient as it's likely to get. A refinement is to determine the energy used per square foot of living space, and compare with the range of houses. The chart here, adapted from the book Residential Energy: Cost Savings and Comfort for Existing Buildings, by John Krigger and Chris Dorsi, shows this range.

Heating Index: How does your house perform compared to other houses?

It’s becoming increasingly clear that the industrial world’s massive consumption of energy cannot continue its present course into the world of the future. Easily recoverable and therefore inexpensive petroleum and natural gas are fading from the scene, and efforts to keep these conventional fossil fuels flowing including hydrofracking, deep water oil wells, and exploiting tar sands all are showing problems. Coal remains cheap, but large areas subject to mountain top removal mining are paying an environmental cost. Next generation nuclear power is promising, but expensive to build. And since the scary event in Japan, nuclear power will doubtless meet with renewed resistance. Renewables such as photovoltaics, wind and ethanol are still so expensive they cannot compete without subsidies and currently contribute relatively trivial amounts of energy. There is one approach left and it’s probably the best from many standpoints. This is energy conservation and energy efficiency. It’s something everybody can do. We could get much better at it.

A good place to start is with one’s house. Most of us use a third or more of the energy we’re directly responsible for in running our houses. In the northern half of the U.S., about half of that is for heating.

How does your house rate in heating efficiency compared to the U.S. average, and to other houses in your neighborhood? Hiring someone to do a complete energy audit is the best way to get a precise accounting of your home’s energy use, but you can easily do an estimate that’s likely to be reasonably accurate.

First, you need to know how much fuel you use for space heating in a typical year. To keep it simple, we’ll limit this exercise to natural gas and heating oil, the primary fuels used to heat houses. Fuels use will vary somewhat from year to year but any recent year should be OK for a decent estimate. Tally up your fuel use for a year by looking at your fuel bills. Natural gas will likely be expressed in therms. Look at how many therms you use during the months of June, July, August, and September. These totals represent non-space-heating uses such as water heating, clothes drying, and cooking. Take the average for each month, multiply by 12, and subtract this total from your yearly total. What’s left is the amount of heat you used for space heating. If you heat with oil, you’ll be measuring gallons of fuel used. If the units are therms, multiply the therm total by 100,000; this will express this energy use in Btus. If the units are gallons, multiply by 135,000, which is the approximate number of Btus per gallon of heating oil.

Now, look on the map above and estimate how many heating degree days (HDD) (1) there are where you live. Here in central New Jersey, this is about 5000. Divide your space heating total by this amount and you’ll have a heating intensity value for your house expressed as Btus per HDD. The average value for the U.S. appears to be in the range of 10,000 to 15,000 Btu per HDD. How does your house stack up?

If your value is much less than this, you are one of the reasons that we don’t have a worse energy problem. Congratulations!

If it’s much more than this, it may be because your house is bigger than average, you keep it warmer than average, it’s less efficient at retaining heat, or your heater is less efficient than most. Perhaps all of these factors are involved. There are a number of steps you can take. Two of these steps are free. One is simply to close off some rooms. Are there unused upstairs rooms, for example? If you can shut off the flow of heat to them, you’ll save a lot of energy. The other free step is turning the thermostat down, and it is tremendously effective; a one degree F drop will reduce your space heating fuel use by up to 5%.(2) And, as a recent article shows, dressing warmly can make a cooler house quite tolerable.(3) A third step is nearly free – installing a programmable thermostat that turns down the heat when you are sleeping or not at home. This can save 15% or more of your space heating fuel.

Next in difficulty and expense but often quite cost-effective are a variety of steps including installing better insulation, eliminating air leaks, and getting a more efficient heater. Even if it’s big, a tight, well-insulated house with an efficient heater can use less energy than a small, inefficient house. The heating efficiency of a house per square foot of living space is a useful metric to use if you want to look more closely at your house’s heating efficiency. I’ll discuss this metric in more detail in a subsequent post.

(1) A heating degree day represents the difference between 65 F and the average outside temperature for that day. For example, if a day’s high is 50 F and its low is 30 F, the average temperature is 40 F and there are 65-40 = 25 heating degree days that day.

(2) http://www.eia.doe.gov/emeu/consumptionbriefs/recs/thermostat_settings/thermostat.html

(3) http://www.lowtechmagazine.com/2011/02/body-insulation-thermal-underwear.html

Oil Production; Fuel Cost as a Percentage of Gross Domestic Product; An Update

There are two widely-used metrics of petroleum production, crude oil (including gas from oil wells that condenses and is mingled with crude oil), and “total liquids.” The latter includes “natural gas liquids” (NGLs), which are liquid fuels derived from natural gas, “other liquids,” which are primarily biofuels, and “refinery processing gain.” The latter is liquids made from crude oil that are less dense than crude oil and so represent an increase in volume over the original crude.
Although it will likely take several more years of data to be sure, it appears that global production of crude oil reached its peak annual production in 2005. Total liquids production, however, still seems to be rising, with the 2010 value about 1.5% higher than 2005. A closer look reveals that the total energy content of these liquids has risen somewhat less, about 1% since 2005. This is because the increase in total liquids is due to larger quantities of the non-crude oil components, and these components contain less energy per volume than crude oil (1).
This view of the data still isn’t the whole story. These data reflect the gross energy in the fuel consumed, not the net energy, sometimes expressed as an energy returned over energy invested (EROI) ratio. It is likely that production of oil and other liquids requires more energy today than formerly, because many of the easily extracted fuels have already been produced. Some estimates in the literature suggest that the EROI of crude oil produced today is in the range of 15 to 1 or less, whereas formerly it had been 50 to 1 or higher. If the EROI of the fuels that are the components of total liquids is declining, it is possible that the trend of the net energy of the total liquids is closer to flat, or even declining.

If market signals reflect reality, as more energy is expended in acquiring a fuel, the price of that fuel should rise. Plotting the cost of energy as a fraction of GDP is an indirect way of gauging the net energy of the fossil fuel resource. Pictured is an update of a plot of fossil fuel cost as a percent of GDP. Although the 2010 value is preliminary, the percent appears to again be above its range during recent periods of prosperity. This probably signals continued economic difficulties, since our way of life is heavily dependent on fossil fuels, and paying more for fuel means less money for other things. The percent would be higher were it not for a big drop in the price of natural gas. If, as some believe, there are vast new reserves of natural gas now available from shale, this fuel may offset to some degree flat or declining quantities and net energy of liquid fuels.

(1) NGLs probably are more or less similar to LP gas, which has about 70% of the energy per volume as crude oil. Ethanol, the primary biofuel, has only about 60% of the energy of crude. The volume of crude plus refinery processing gain has an energy density somewhere in-between the two chief refinery products, gasoline and diesel fuel and similar fuels, and this is about 5% less than the energy density of the original crude. This difference represents the energy consumed in producing the refined products.