The amount of methane sequestered in gas hydrates is probably enormous, but estimates of the amounts are speculative and range over three orders-of-magnitude from about 3,114 to 7,634,000 trillion cubic meters (reviewed by Kvenvolden, 1993).  It is likely, however, that the amount of gas in the hydrate reservoirs of the world greatly exceeds the volume of known conventional gas reserves.  Before reviewing assessments of the world gas hydrate resources it is necessary to examine the quality and variability of gas hydrate assessments at the accumulation and reservoir scale.

Gas Hydrates at the Reservoir and Accumulation Scale

Estimates of the amount of gas hydrates and associated gas within a given gas hydrate accumulation can vary considerably.  For example, recent estimates of the volume of gas that may be contained in the gas hydrates and free-gas beneath the gas hydrates on the Blake Ridge range from about 70 trillion cubic meters of gas over an area of 26,000 km2 (Dickens et al., 1997) to about 80 trillion cubic meters of gas for an area of 100,000 km2 (Holbrook et al., 1996).  The difference between these two estimates has been attributed to the observation that the amount of free-gas directly measured within pressure-core samples (Dickens et al., 1997) from beneath the gas hydrates is significantly larger than that estimated from borehole vertical seismic profile data (Holbrook et al., 1996).  Other published studies indicate that the gas hydrates at the crest of the Blake Ridge alone (area of about 3,000 km2) may contain more than 18  trillion cubic meters of gas (Dillon and Paull, 1983).  The broad range of these estimates demonstrates the need for high-resolution measurements of the gas hydrate and associated free-gas volumes within any gas hydrate accumulation of interest.

It has been suggested that the volume of gas that may be contained in a gas hydrate accumulation depends on five "reservoir" parameters (modified from Collett, 1993): (1) areal extent of the gas-hydrate occurrence, (2) "reservoir" thickness, (3) sediment porosity, (4) degree of gas-hydrate saturation, and (5) the hydrate gas yield volumetric parameter which defines how much free-gas (at STP) is stored within a gas hydrate (also known as the hydrate number).  In the following section, the five "reservoir" parameters (Table 1) needed to calculate the volume of gas associated with the gas hydrates on the Blake Ridge (ODP Sites 994, 995, and 997; Shipboard Scientific Party, 1996), along the Cascadia continental margin (ODP Site 889; Shipboard Scientific Party, 1994), on the North Slope of Alaska (Northwest Eileen State-2 well; Collett, 1993), and in the Mackenzie River Delta of Canada (Mallik 2L-38 well; Dallimore et al., 1999) are assessed.

The following "resource" assessment (modified from Collett, 1998) has been conducted on a site-by-site basis; that is, for each site examined the volume of gas hydrate and associated gas within a one square kilometer area surrounding each drill-site have been individually calculated (Table 1).  For this "resource" assessment, I have defined the thickness of the gas-hydrate-bearing sedimentary section at both the marine and permafrost drill sites to be the total thickness of the downhole log inferred gas-hydrate accumulation (Table 1).  Average core derived sediment porosities for the gas-hydrate-bearing reservoirs at Sites 994, 995, 997 and 889 range from about 52 to 58 percent (Table 1).  The "corrected" density log derived sedimentary porosities for the three gas-hydrate-bearing units identified in the Northwest Eileen State-2 (Units C, D, and E) range from an average value of about 36 to 39 percent (Table 1).  The density log derived sediment porosities for the gas-hydrate-bearing interval in the Mallik 2L-38 well averages about 31% (Table 1).  Gas-hydrate saturations at the marine drill-sites (Sites 994, 995, 997 and 889), calculated from available downhole logs, range from an average value of about 3 to 6 percent (Table 1).  Gas-hydrate saturations in all three gas-hydrate-bearing units (Units C, D, and E) in the Northwest Eileen State-2 well, calculated from available downhole log data, range from an average value of about 33 to 61 percent (Table 1).  The resistivity well log derived gas-hydrate saturations in the Mallik 2L-38 well average about 44% (Table 1).  In this assessment, I have assumed a hydrate number of 6.325 (90% gas filled clathrate) which corresponds to a gas yield of 164 m3 of methane (at STP) for each cubic meter of gas hydrate.  The log inferred gas hydrates at Sites 994, 995, and 997 on the Blake Ridge contain between 670,000,000 and 1,450,000,000 cubic meters of gas per square kilometer (Table 1).  The volume of gas within the log inferred gas hydrates at Site 889 on the Cascadia continental margin is about 467,000,000 cubic meters of gas per square kilometer (Table 1).  Cumulatively, all three log inferred gas-hydrate-bearing stratigraphic units (Units C, D, and E) drilled and cored in the Northwest Eileen State-2 well may contain about 1,511,000,000 cubic meters of gas in the one square kilometer area surrounding this drill-site (Table 1).  It was also determined that the log inferred gas-hydrate-bearing stratigraphic interval drilled in the Mallik 2L-38 well contains about 4,750,000,000 cubic meters of gas in the one square kilometer area surrounding the Mallik drill-site (Table 1).

A close examination of the gas hydrate saturations in Table 1 reveals a potential problem associated with production of gas from marine gas hydrates.  Even though vast portions of the world’s continental shelves appear to be underlain by gas hydrates, the concentration of hydrates within most marine accumulations appears to be very low.  Low gas hydrate concentrations may significantly affect the economic production potential of marine gas hydrates (gas hydrate production is discussed in more detail later in this paper).

Gas Hydrates at the World and National Scale

World estimates for the amount of natural gas in gas hydrate deposits range from 14 to 34,000 trillion cubic meters for permafrost areas and from 3,100 to 7,600,000 trillion cubic meters for oceanic sediments (modified from Kvenvolden, 1993).  The estimates in Table 2 show considerable variation, but oceanic sediments seem to be a much greater resource of natural gas than continental sediments.  Current estimates of the amount of methane in the world’s gas hydrate accumulations are in rough accord at about 20,000 trillion cubic meters (reviewed by Kvenvolden, 1993).  If these estimates are valid, the amount of methane in gas hydrates is almost two orders of magnitude larger than the estimated total remaining recoverable conventional methane resources, estimated to be about 250 trillion cubic meters (Masters et al., 1991).

The recently completed 1995 National Assessment of United States Oil and Gas Resources, conducted by the U.S. Geological Survey, focused on assessing the undiscovered conventional and unconventional resources of crude oil and natural gas in the United States (Gautier et al., 1995).  This assessment included for the first time a systematic resource appraisal of the in-place natural gas hydrate resources of the United States onshore and offshore regions (Collett, 1995).  In this assessment, 11 gas-hydrate plays were identified within four offshore and one onshore gas hydrate provinces.  The offshore gas hydrate provinces assessed lie within the U.S. Exclusive Economic Zone adjacent to the lower 48 States and Alaska.  The only onshore province assessed was the North Slope of Alaska.  In-place gas resources within the gas hydrates of the United States are estimated to range from about 3,200 to 19,000 trillion cubic meters of gas, at the 0.95 and 0.05 probability levels, respectively.  Although this wide range of values shows a high degree of uncertainty, it does indicate the potential for enormous quantities of gas stored as gas hydrates.  The mean in-place value for the entire United States is calculated to be about 9,000 trillion cubic meters of gas.

Gas Hydrate Production Technology

Even though gas hydrates are known to occur in numerous marine and Arctic settings, little is known about the technology necessary to produce gas hydrates.  Most of the existing gas hydrate "resource" assessments do not address the problem of gas hydrate recoverability.  Proposed methods of gas recovery from hydrates (Figure 18) usually deal with dissociating or "melting" in-situ gas hydrates by (1) heating the reservoir beyond hydrate formation temperatures, (2) decreasing the reservoir pressure below hydrate equilibrium, or (3) injecting an inhibitor, such as methanol or glycol, into the reservoir to decrease hydrate stability conditions.  Gas recovery from hydrates is hindered because the gas is in a solid form and because hydrates are usually widely dispersed in hostile Arctic and deep marine environments.  First order thermal stimulation computer models (incorporating heat and mass balance) have been developed to evaluate hydrate gas production from hot water and steam floods, which have shown that gas can be produced from hydrates at sufficient rates to make gas hydrates a technically recoverable resource (Sloan, 1998).  However, the economic cost associated with these types of enhanced gas recovery techniques would be prohibitive.  Similarly, the use of gas hydrate inhibitors in the production of gas from hydrates has been shown to be technically feasible (Sloan, 1998), however, the use of large volumes of chemicals such as methanol comes with a high economic and environmental cost.  Among the various techniques for production of natural gas from in-situ gas hydrates, the most economically promising method is considered to be the depressurization technique.  However, the extraction of gas from a gas hydrate accumulation by depressurization may be hampered by the formation of ice and/or the reformation of gas hydrate due to the endothermic nature of gas hydrate dissociation.

The Messoyakha gas field in the northern part of the West Siberian Basin is often used as an example of a hydrocarbon accumulation from which gas has been produced from in-situ natural gas hydrates.  Production data and other pertinent geologic information have been used to document the presence of gas hydrates within the upper part of the Messoyakha field (Makogon, 1981).  It has also been suggested that the production history of the Messoyakha field demonstrates that gas hydrates are an immediate producible source of natural gas, and that production can be started and maintained by conventional methods.  Long-term production from the gas-hydrate part of the Messoyakha field is presumed to have been achieved by the simple depressurization scheme.  As production began from the lower free-gas portion of the Messoyakha field in 1969, the measured reservoir-pressures followed predicted decline relations; however, by 1971 the reservoir pressures began to deviate from expected values.  This deviation has been attributed to the liberation of free-gas from dissociating gas hydrates.  Throughout the production history of the Messoyakha field, it is estimated that about 36% (about 5 billion cubic meters) of the gas withdrawn from the field has come from the gas hydrates (Makogon, 1981).  Recently, however, several studies suggest that gas hydrates may not be significantly contributing to gas production in the Messoyakha field (reviewed by Collett and Ginsburg, 1998).

It should be noted, that our current assessment of proposed methods for gas hydrate production do not consider some of the more recently developed advanced oil and gas production schemes.  For example, the usefulness of downhole heating methods such as in-situ combustion, electromagnetic heating, or downhole electrical heating have not been evaluated.  In addition, advanced drilling techniques and complex downhole completions, including horizontal wells and multiple laterals, have not been considered in any comprehensive gas hydrate production scheme.  Gas hydrate provinces with existing conventional oil and gas production may also provide us with the opportunity to test relatively more advanced gas hydrate production methods.  For example, in northern Alaska existing "watered-out” production wells are being evaluated as potential sources for hot geopressured brines that will be used to thermally stimulate gas hydrate production.

As previously noted, the low concentration of hydrates in most of the world’s marine gas hydrate occurrences raises a concern over the production technology required to produce gas from highly disseminated gas hydrate accumulations.  In addition, the host-sediments also represent a significant technical challenge to potential gas hydrate production.  In most cases, marine gas hydrates have been found in clay-rich unconsolidated sedimentary sections that exhibit little or no permeability.  Most of the existing gas hydrate production models require the establishment of reliable flow paths within the formation to allow the movement of produced gas to the wellbore and injected fluids into the gas-hydrate-bearing sediments.  It is unlikely, however, that most marine sediments possess the mechanical strength to allow the generation of significant flow paths.  It is possible that in basins with significant input of coarse-grained clastic sediments, such as the Gulf of Mexico or along the eastern margin of India, gas hydrates may be reservoired at high concentrations in more conventional clastic reservoirs; which is more analogous to the nature of gas hydrate occurrences in onshore permafrost environments (Collett, 1993; Dallimore et al., 1999).

Natural Gas Hydrates: Resource of the 21st Century?