Robert L Kleinberg & Peter G Brewer. American Scientist. Volume 89, Issue 3. May/Jun 2001.
In the novel Cat’s Cradle, author Kurt Vonnegut described a form of ice, “ice-nine,” that is more stable than water at room temperature. According to his story, the only thing that prevents liquid water from immediately converting to ice-nine is the absence of a seed crystal to initiate freezing. Throwing a small chunk of this peculiar form of ice into a lake, for example, would cause the entire body of water to solidify. This phenomenon makes the ice-nine in Vonnegut’s tale a rather dangerous invention, one that ultimately becomes a doomsday weapon. This literary creation has an interesting origin: It was the author’s brother, Bernard Vonnegut, who realized that the water vapor in clouds could be made to form ice (which would fall as rain) by seeding the cloud with crystals of silver iodide.
The ice-nine of Cat’s Cradle is of course a fantasy But surprisingly, some forms of ice can indeed be more stable than water at room temperature-the gas hydrates. These substances contain guest molecules, of methane for instance, trapped within crystalline water. The small nonpolar molecules that are hydrate formers favor the formation of solids called clathrates. Like the ice-nine of Vonnegut’s story, nucleation of the clathrate phase is difficult because water and gas molecules must assemble themselves into a rather complex crystallographic form.
Clathrate compounds of many compositions exist, including some with interesting semiconducting properties (see “Thermoelectric Clathrates” by George S. Nolas and Glen A. Slack, March-April). The gas hydrates in particular have long been known. The pioneering British chemist Humphrey Davy discovered the first example, chlorine hydrate, in his chilly laboratory in 1810. Our focus is on methane hydrate that forms within the earth. These curious ices are worthy of scientific study in their own right and may one day provide an important source of natural gas.
Frozen Fields of Gas
Hydrates of natural gas first came to the attention of the petroleum industry during the 1930s, when these solids proved to be the cause of plugged natural-gas pipelines. Since the early 1970s geologists have come to realize that gas hydrates also exist within the earth, under many regions of permafrost and beneath certain parts of the ocean.
Methane hydrate forms in these places because it is stable only at high pressure or low temperature. The deep waters of almost all the world’s oceans are cold enough to stabilize hydrate at depths greater than about 500 meters. Warm seas such as the Mediterranean are an exception. On the other hand, gas hydrates are found close to the surface in permafrost regions, because of the low prevailing temperatures.
Hydrates are prevalent in sediments where methane is available either from the biological breakdown of organic matter or from thermal decomposition of an underlying petroleum reservoir. These deposits are not, however, very thick: Despite higher pressure, elevated temperature at depth destabilizes methane clathrate. Thus deposits of gas hydrate normally exist only over a narrow range of depths, typically just a few hundred meters.
If the temperature of a hydrate deposit increases, it will decompose and liberate methane. Because methane is 25 times more effective than carbon dioxide as greenhouse gas, its addition to the atmosphere will tend to reinforce the original temperature rise. Many geologists believe this mechanism accounts for the warming the planet experienced in the late Paleocene, 55 million years ago. On the other hand, a drop in sea level, caused by enlargement of continental glaciers during an ice age, will reduce seafloor pressures. This can also destabilize hydrates, cause a large release of methane, and bring the ice age to an end.
Although gas hydrates threaten profound global climate change, they may also offer an important environmental benefit: a vast supply of clean energy. Even if the growth of fossil-fuel consumption continues to slow in Europe, North America and Japan as conservation and alternative energy sources come to play larger roles, the rest of the world inevitably will require much more energy than it now uses. Absent surprising developments in nuclear power or other sectors, only fossil fuels will be able to meet this demand.
The Global Gas Market
Natural gas is increasingly becoming the fossil fuel of choice. It is normally delivered to consumers at very high purity and so is almost completely nonpolluting-the only products of combustion are carbon dioxide and water. Carbon dioxide is a greenhouse gas, but to generate the same amount of energy, burning oil produces one and a half times as much of it as does gas, and burning coal generates twice as much. Fortunately, the use of natural gas has increased from less than 5 percent of the world’s primary energy consumption in the 1930s to more than 25 percent today.
The world presently consumes about 85 trillion standard cubic feet (tcf) of natural gas per year. (A standard cubic foot as defined by petroleum engineers is the quantity of gas that would fill a cubic foot at one atmosphere pressure and 60 degrees Fahrenheit.) Total proven reserves that can be recovered with existing technology at current prices, what is called conventional gas in the energy industry, amounts to approximately 5,300 tcf.
Although the world as a whole has at least a 60-year supply of gas at present production rates, these reserves are distributed very unevenly. Russia accounts for 1,700 tcf of the total, and the countries surrounding the Persian Gulf another 1,700 tcf. Proven gas reserves in the United States, Canada and Mexico total 260 tcf; production is 28 tcf per year. Europe currently buys much of the gas it consumes from Russia. Japan has reserves of only 1 tcf; it imports almost all the gas it uses from Indonesia and the Middle East at considerable expense. India and China, which will be major energy-consuming nations in the future, also have few domestic gas reserves.
Assessing Hydrate Reservoirs
Global gas markets would change dramatically if clathrate deposits began supplying gas. Gas hydrates are found off the shores of many nations and are also prevalent under permafrost in the arctic regions of Alaska, Canada and Russia. The U.S. alone is estimated to have 300,000 tcf of methane sequestered in gas hydrates. In a single hydrate deposit located in the Nankai Trough 50 kilometers south of Hamamatsu, there may be 2,000 times more gas than in all of Japan’s conventional reserves.
Such estimates have inspired dreams of energy independence among some gas-importing countries. Not only are the amounts of methane huge, but many potential reservoirs are within the exclusive economic zones of the consuming nations, and in some cases the fields lie quite close to end users. However, major hurdles must be overcome before this immense supply can be exploited.
The difficulty and expense of producing gas depend less on the gross quantity of the resource than on its concentration. The density of methane in some hydrate deposits compares poorly with what is found in conventional gas reservoirs. As an example, consider the Blake Ridge formation, located 300 kilometers off the coast of South Carolina.
A recent assessment showed that Blake Ridge contains about 1,000 tcf of methane, six times the total proven U.S. reserves of conventional gas. Unfortunately, this immense store of methane is spread over an area of about 26,000 square kilometers. Thus the hydrate is quite dilute, occupying on average only 2 percent of the sediment volume.
One might think that methane is so densely packed in solid hydrate that Blake Ridge nevertheless constitutes a rich deposit. After all, when a cubic meter of gas hydrate decomposes, the products are 0.79 cubic meter of water and 172 cubic meters of methane at 60 degrees F and 1 atmosphere pressure. While that seems impressive, a petroleum engineer would note that the density of methane in the solid hydrate is only 0.12 gram per cubic centimeter. A gas well drilled into a formation with 2 percent methane at this density is considered a “dry hole.”
Even if the concentration were higher, producing gas from a hydrate formation would not be a simple matter. One must raise the temperature, reduce the pressure or introduce chemicals to destabilize the clathrate. Theoretical calculations confirm that, in the most favorable circumstances, less heat is needed to dissociate hydrate than can be produced from the gas generated. Such calculations are, however, sensitive to the relative amounts of liquid water and hydrate that fill the pore space of the rock, because the energy spent heating liquid water is wasted. Similarly, the chemical treatments under consideration lose their effectiveness when diluted by too much water.
Depressurization might seem even more difficult to achieve, but, in fact, this strategy is fairly straightforward. Natural gas often collects beneath gas hydrates, where the temperature is too high for the solid phase to be stable. Extracting this gas lowers the pressure at the base of the hydrate deposit, releasing methane.
This process has been demonstrated at the only hydrate reservoir yet exploited commercially-one associated with the Messoyakha gas field in the west Siberian arctic. As the conventional gas field there was depleted, an overlying hydrate formation dissociated, contributing an estimated onethird of the total amount of gas extracted between 1969 and 1986-or so most geologists had believed: Serious questions have emerged recently as to whether hydrate actually contributed to production.
No matter what production method is used, drilling on continental shelves and slopes, where oceanic gas hydrate deposits are found, is technically challenging. Solid hydrate frequently cements the sediments in which it exists, and this semi-consolidated mass overlies highly fluidized sands or muds below. The setting is reminiscent of avalanche conditions on a snow-covered mountain. Indeed, the analogy is quite apt: The flanks of Blake Ridge have experienced massive subsea slumps, attributed to the instability of hydrates that have accumulated there. Gas production might well trigger more submarine landslides that could endanger equipment and personnel.
Another barrier is the considerable expense of operations at the edge of the continental shelf. Although drilling at these water depths has become possible in the past decade, the massive rigs needed cost in the range of a billion dollars. It is probably safe to say that there will be windmills on the roofs of homes before Blake Ridge is exploited for its natural gas.
A consideration of the practicalities certainly paints a gloomy picture of the prospects for ever exploiting Blake Ridge, but not all hydrate reservoirs are alike. A more promising one resides beneath the permafrost of Prudhoe Bay, the oil-producing region on the arctic coast of Alaska. Six reservoirs, with a total area of 1,643 square kilometers and an average thickness of 15 meters, contain hydrate, which constitutes 33 percent of the sediment volume. This deposit of some 40 trillion cubic feet of gas is much smaller, but much richer, than that on Blake Ridge.
Although drilling operations in the arctic are always challenging and expensive, oil companies have considerable infrastructure at Prudhoe Bay, which makes that prospect a rather accessible one. A bigger problem is transporting this natural gas to markets located thousands of kilometers away. However, gas pipelines from the arctic to North American markets are on the drawing boards. Although these pipelines will be built primarily to transport gas from conventional reservoirs, methane from associated hydrate deposits may contribute to the supply.
As these examples illustrate, knowledge of the location and volume of a hydrate reservoir is not sufficient to determine its economic potential. The methane concentration is equally important. The same information is needed by oil and gas companies in their assessments of conventional hydrocarbon reserves.
One might imagine that the solution is simply to bring up core samples while drilling. Indeed, this technique is used to evaluate exploratory wells in conventional oil and gas fields and hydrate deposits. However, the process is too tedious and costly for routine drilling operations, and the situation is worse for hydrate-bearing formations, because clathrate decomposes and loses gas as it is brought to the surface. Even when cores are successfully extracted, their analysis is sufficiently complex that, typically, only a very small set of samples from each well is evaluated. Because the earth is heterogeneous on all length scales, depth– continuous measurements are required. Therefore earth scientists get most of their information about underground formations by sending instruments down boreholes, a procedure called well logging.
Well logging typically involves several sensors, each used for a specific purpose. Borehole measurements of sound speed, electrical resistivity and gamma-ray scattering are commonly made in hydrate-bearing sediments.
Logs of gamma-ray scattering are relatively easy to interpret. When a radioactive source irradiates the earth around the instrument, the proportion of gamma rays that returns to the detector depends on the density of the surrounding rock. This property in turn reflects the volume-weighted average of the densities of the various components. Water and methane hydrate have similar densities, but there is considerable contrast between these pore-filling constituents and typical rock-forming minerals such as silica. Thus the results give a measure of what geologists call porosity, the volume fraction of the rock occupied by water and hydrate. However, gamma-ray measurements say nothing about whether the pores contain gas hydrate or water or some combination of the two.
The task of distinguishing water from hydrate typically relies on measurements of electrical resistivity. Hydrate and most solid minerals are good electrical insulators, whereas saltwater is an excellent conductor. Thus comparing the electrical resistivity with porosity yields, in principle, the hydrate content of the sediment. In practice, a problem arises: Interpretation requires knowledge of the resistivity of the pore water itself, which is usually not equal to that of seawater, in part because the creation and dissociation of hydrate alters the constitution of the pore fluids. Clay minerals, which are usually abundant in sediments, also influence the resistivity measurements.
The third common measurement, the velocity of sound, varies with the acoustic properties of the constituent materials and their proportions. Sound speed also depends on the way these components are assembled, and therein lies an ambiguity. Hydrate may cement the grain contacts in what had been unconsolidated sediment. If so, the presence of a small amount of hydrate will increase the speed of sound in the sediment considerably. It is also possible for the hydrate simply to replace water in the pore space without contributing rigidity, in which case a larger amount of hydrate is needed to explain a measured increase in sound speed. Thus, as with resistivity logging, the proportion of hydrate cannot be uniquely determined.
Nuclear Magnetic Resonance
These difficulties prompted us to investigate yet another approach: nuclear magnetic resonance, or NMR. The most familiar application of NMR is in medical diagnosis: magnetic resonance imaging. The use of NMR to investigate geologic materials is quite different and has primarily focused on quantitative measurements of the fluids that reside in the pore space of rocks. Because typical pore-filling fluids-water, oil and natural gas-are rich in hydrogen, most of these applications are based on the magnetic resonance of hydrogen nuclei, that is, of protons.
For typical laboratory investigations, one places the sample inside an apparatus consisting of a magnet and a coil. The protons in the substance under study, which all have the same intrinsic magnetic moment, are aligned by the field generated by the magnet. The coil then transmits a series of radio-frequency pulses tuned to the appropriate resonant frequency for protons in the applied field. These pulses reorient the magnetic moments of the nuclei, which emit radio signals of their own. These are picked up by the coil acting as a receiver. The amplitude of the signal reflects the amount of hydrogen present in the sample. Further information is obtained by tracking the rate at which this signal decays.
NMR signals from solids disappear rapidly, and one can easily avoid detecting them. NMR signals from liquids decay more slowly and are easily detected. Proton NMR measurements are thus valuable for probing the fluids in sedimentary formations, which can be 40 percent or more pore space. The results can indicate the sizes of the pores and the quantity of liquid held in them.
NMR signals from a fluid-filled rock decrease faster than if the same liquid were measured in a test tube. The decay rate is enhanced by magnetic interactions between the fluid and solid grains, and depends on the surface-to-volume ratio of the pores. Thus the spins of water nuclei in small pores decay, or relax, rapidly, while water in large pores relaxes more slowly.
Gas hydrate is likely to occupy the largest pore spaces available in a rock, but because this substance is itself a solid, the NMR signal from it decays too rapidly to be detected. How then can this technique be helpful in probing sedimentary formations for hydrates? Alone, NMR would not be able to do the job, but recall that gammaray scattering indicates the amount of water plus hydrate present. An NMR log from the same formation detects just the pore water, and by subtracting the two measurements one can determine the amount of hydrate present at any particular depth.
In the usual NMR measurement the sample, which could be a person, is placed inside the measurement apparatus. In well logging, it is the apparatus that is placed inside the “sample,” in this case the earth around the borehole. Thus the equipment must project the appropriate magnetic fields outward, into the surrounding formation.
One of us (Kleinberg) and colleagues at Schlumberger have developed just this sort of device, and this equipment is now used routinely to log oil and gas wells. The apparatus contains two large permanent magnets, magnetized perpendicular to the axis of the borehole. The field these magnets generate is reasonably uniform only over a small volume of investigation, about 2 by 2 by 15 centimeters in size centered about 2.5 centimeters inside the earth formation. The antenna, used as both transmitter and receiver, resides in a semicircular cavity, covered by a nonmetallic wear plate.
This borehole instrument further differs from laboratory NMR devices in that it is built to withstand hostile environments: It is designed to survive shocks of 100 g and is rated to operate at temperatures from -25 to 175 degrees C and at pressures as high as 1,400 bars. The conditions in hydrate formations never approach these limits. Thus this NMR instrument would appear to be suitable for examining hydrate deposits, but until we began working together, this application of NMR logging had not been validated.
To investigate the NMR properties of methane hydrate, we used artificially prepared samples of hydrate-loaded sediment and rock. Such samples are difficult to produce in the laboratory because a layer of solid hydrate forms at the interface between liquid water and gas, effectively stopping the reaction. This barrier can be broken by vigorous agitation, and some investigators have resorted to using grinding apparatus to hasten the completion of the reaction. Even employing these strenuous measures, it can take several days to fill a small pressure vessel with hydrate, and, of course, such means are not suitable for forming hydrate in the pore space of rocks.
Aware of these problems, one of us (Brewer) led a group from the Monterey Bay Aquarium Research Institute to attempt the artificial creation of gas hydrates in a novel manner-on the seabed. The ocean floor not only provides the appropriate temperature and pressure conditions for clathrates to exist, it also offers the ideal setting in which to examine how hydrates might arise naturally. These investigations, starting in 1996, used clear plastic tubes filled with seawater or with mixtures of sand and seawater, which a remotely operated vehicle carried to the ocean floor. Valves controlled by commands from the surface allowed methane from a pressurized tank to flow into each cylinder.
Knowing that hydrates form exceedingly slowly in the laboratory, the members of this group expected little to happen in the few hours of “bottom time” available for the experiment. Surprisingly, gas hydrate appeared within minutes. In the tubes filled with sand, newly created hydrate quickly turned the loose sediment into a solidified block. An underwater video camera mounted on the vehicle documented these changes, but the investigators had no means for making quantitative measurements.
The difficulty in determining the amount of hydrate formed in that experiment prompted our collaboration last year on another research cruise, employing the Schlumberger NMR instrument. We mounted this sensor on the front of another subsea vehicle and used it to investigate artificially prepared hydrate samples left on the seafloor several months before. Unlike in the previous tests, we arranged this time to monitor the amount of methane released when the remotely operated vehicle transported the samples into shallow water, where hydrate is not stable. Knowing the temperature, pressure and volume of methane gas evolved, we could then work out how much hydrate must have been in each sample originally. The values compared quite well with what we estimated from the NMR measurements.
Although hydrates contain an immense amount of natural gas, quantitative measurements of individual reservoirs will be required to determine where this resource can be exploited economically. Even where natural gas derived from hydrates will not be able to compete economically with conventional fossil fuels, these deposits could have value as strategic reserves. After all, some nations subsidize domestic production of food as a matter of national security; creating the means to exploit gas hydrates could serve the same purpose. This is why Japan has long devoted a substantial effort to studying natural-gas hydrates, including the drilling and analysis of a number of test wells. Strategic planning also accounts for the rising interest in hydrates within the U.S. Department of Energy.
Considerations of fossil-fuel supplies aside, there are important reasons to study these curious ices. There is mounting evidence that massive hydrate dissociation events are synchronous with climate shifts, although the interpretation of the geologic data is a matter of vigorous debate. Greater knowledge of these deposits will allow earth scientists to better gauge their contributions to episodes of global warming. Gas hydrate may not be Vonnegut’s fictional ice-nine, but there could be unsettling analogies to its impact on the planetary environment.