The big question: How much CO2 can the Earth hold?

UTRECHT, the Netherlands -- The Dutch used to discover new worlds across unexplored seas. Now, they are beginning to trace the edges of a new undiscovered country, and it is right beneath their shores.

The Netherlands, a country that chose to build many of its cities below sea level, is famous for its pragmatic, long-term planning. So it should be no surprise that, when it comes to efforts to store carbon dioxide underground for a millennium or more, Holland has been leading the way, planning for years to turn declining natural gas fields off their shores into storage sites.

Initial estimates of the fields were promising. It seemed 40 years of emissions from eight large coal-fired power plants could be stored. Then scientists looked closer, probing each site's geology, to disturbing results.

Some fields were too small or perforated by drills to store CO2, they found. Others were stubborn, their rocks likely to resist the injection of the gas.

Soon enough, the Dutch had to cut their storage estimate in half.

It is a disappointing result that should be kept in mind as estimates of CO2 storage potential, which mostly exist on countrywide or regional levels, are refined and localized, said Filip Neele, a research geologist here at the geosciences branch of TNO, the Dutch national lab of applied sciences.

In some cases, Neele would not be surprised to see storage estimates fall by up to 95 percent compared with the original projection. Though even then, he added, the capacity would be still large thanks to the vast size of the available storage formations.

"This is likely to be true for any large-scale inventory of storage capacity," Neele said. "If you look at a country scale and try to assess the storage potential, you're very likely to grossly overestimate the storage potential."

Welcome to the new terra incognita. As politicians and businesses push forward with carbon capture and storage, or CCS, as their "bridge" to renewable energy, geologists are scrambling to properly estimate how much CO2 can be stored in deep, water-flush rock formations -- called saline aquifers -- that have long been ignored by, well, pretty much everyone. They are blank spaces below the map and are only beginning to be better understood.

Europe has recently concluded a wide effort to martial accurate storage estimates, with mixed results. And in the United States, $75 million in last year's economic stimulus package went to 11 projects investigating the actual storage available in rock formations across the country, from just outside New York City to off the coast of Los Angeles. It is hoped that these studies will begin to cull accurate predictions from the rough guesses of yesterday.

Also, the U.S. Geological Survey, the country's leading authority on all things subterranean, is revising its methods of assessing storage potential, a move that could swing existing estimates by the Energy Department up or, more likely, down. Among other problems, the survey does not yet understand the rate at which many reservoirs will trap CO2 for hundreds of years or more, said Robert Burruss, one of the agency's lead research geologists for CCS.

"The issue of estimates for saline reservoirs -- it's a very tricky number," Burruss said. "Because we really don't have, at least in my opinion, an adequate understanding of the trapping mechanism."

Like many, the survey has previously had limited interest in saline aquifers, which are typically made of porous rocks like sandstone and often found near oil or gas deposits. They were a sideshow in the race for mineral wealth. The formations attracted little attention for precisely the reason they now hold so much storage potential: their sheer vacancy. No oil. No gas. No gold. Just brackish water.

Given how little is known, it will be difficult to make broad generalities about these aquifers, and estimates, unless they are site-specific, should be taken with a large grain of salt, said Sarah Forbes, the CCS project manager at the World Resources Institute.

"You have to be really careful about storage capacity estimates," Forbes said. With conservative measures, the United States still projects to have substantial storage potential, but that does not mean individual sites will pan out, she warned.

"You're not actually going to know the capacity until you test it," she said.

Despite their decline in capacity, the Dutch have in fact been lucky. They are targeting used natural gas fields for their storage, which are better understood than the saline aquifers that many countries, including the United States, would have to use for CO2 storage. The variability of these aquifers means that for much of the world, the only surety about how much CO2 can be stored underground is its unpredictability.

"You do not know a lot about deep aquifers," Neele said. "There are no wells. Often there is no seismic data. ... We don't know if they're sufficiently permeable, if they have a good seal. There's a lot of uncertainty."

Tank models

The Netherlands is among just a few countries that have any certainty about how much CO2 they can store underground. This is entirely because of geological happenstance. The country is rich in natural gas deposits but has few saline aquifers. The gas deposits, except for one massive field in the north, are running dry. And their structure is well understood.

As evidence that gas-field storage works, TNO points to a project called K12-B, which it has helped run for the past six years at a natural gas platform 50 miles off the coast. The French gas firm GDF Suez SA operates the platform, along with 30 other locations in the North Sea. And it is eager to see if it can extend its profitable use of the sites with CO2 storage.

So far, CO2 storage has operated smoothly, with K12-B proving to be a model for Dutch depleted fields. Most of the gas deposits are 2 to 3 miles below the sea floor and are encased by vast layers of salt, and no better rock exists for trapping CO2. Salt is plastic -- from a long view, it behaves like a fluid -- and fills in every possible pore, preventing escape.

Salt also prevents water from rushing in to fill the void left by natural gas, meaning the depleted deposits sit at unnaturally low pressures and are eager to suck in CO2. It is typical of Dutch fields and provides the "safest case we can make" for storage, said Rob Arts, TNO's CO2 monitoring expert.

"It's a tank model," Arts said. "You produce [the gas] and, since everything is sealed off and compartmentalized, there is no water to go in it. So it remains at very low pressures."

Between its onshore and offshore gas fields, Holland will be able to store the emissions of eight coal-fired power plants, each about 1,000 megawatts, for 40 years, Neele said. That would be enough electricity to meet nearly half of the country's existing demand -- if only Dutch needs were being met.

Unfortunately, there are many countries in Europe, Germany especially, that burn far more coal than the Dutch and have far fewer storage options, particularly if public resistance to saline aquifer storage remains high (Greenwire, April 7).

The Germans are already eyeing locking up storage in Dutch fields, Neele said, and under European law, it would be illegal for the Dutch to block access to the storage.

It has become apparent that should CCS catch on in any way, a massive infrastructure project will be required for countries like Poland and the Czech Republic to ship their CO2 through pipelines to offshore sites in the Netherlands, Britain and Norway, said Henk Pagnier, the head of TNO's CO2 storage program.

"If storage is big in Europe, then we need the North Sea," Pagnier said. "Because we can see the Central European countries will have problems storing the quantities they need to store. And the Norwegian and U.K. gas fields are the main suspects, especially for the near future."

Unlike the United States, which has a substantial CO2 pipeline system along the Gulf Coast, Europe has no pipelines for handling the gas. Should CCS catch on, the continent would have to build large, backbone pipelines at a rate of 100 kilometers a month, Neele said -- a high but not impossible frequency.

"Our challenge will be to generate a completely new set of gas-transportation infrastructure that would run in reverse," said Christoph Heubeck, a geologist at Berlin's Free University. "Think about the huge infrastructure that we have. All the gas stations. All the pipelines. All these gigantic tankers. You'd have to construct a similar-sized infrastructure to put it all back."

Efficiency matters

Like in Holland, U.S. companies that first adopt CCS will most likely target CO2 storage in depleted oil and gas deposits. Their mechanisms for holding CO2 are easy to see, and storage estimates are "pretty straightforward," said Burruss, the USGS geologist, relating directly to the amount of oil and gas that has been previously produced from the field. Plus, CO2 injection is already used to enhance oil recovery in many Texan oil fields.

According to DOE estimates, the United States and Canada have 138 billion metric tons of potential CO2 storage in 10,000 oil and gas reservoirs scattered over 27 states and three provinces. For comparison, U.S. coal-fired power plants emit some 2 billion metric tons of CO2 each year, and CCS-equipped plants would likely operate for 40 years or more.

"Certainly, there is plenty of storage capacity in [these fields]," Burruss said. "But not if you're planning to hook up every coal-fired power plant in the world."

Because of this storage squeeze, saline aquifers will be needed, Burruss said. The potential capacity of these rock formations at first seems staggering. According to the most conservative estimate presented in DOE's Carbon Sequestration Atlas, a maximum of some 3,300 billion metric tons of storage is available, 23 times the capacity of depleted oil and gas reservoirs.

However, the DOE estimate for saline aquifers hinges on one poorly understood variable: the percentage of space in the rock formation that will hold CO2 for 1,000 years or more, a term known as storage efficiency. Given this uncertainty, DOE has been conservative in its estimates, using an efficiency rate of 1 percent to 4 percent, even though some natural gas formations, also often found in sandstone, have efficiencies as high as 80 percent.

The fickle nature of storage efficiency reflects a fundamental scientific question. Most saline aquifers are not the perfect, bubble-like formations that trap natural gas beneath a caprock. Instead, the CO2 plume is likely to have some space to slide around under the caprock. While it sashays, geologists expect bits of the gas to be trapped behind in the rock pores, a mechanism called capillary trapping.

With only a few CO2 injection projects operating, so far storage efficiency is a "number we don't have a lot of control over, in terms of actual measured behavior," Burruss said. At individual sites, it could vary from 0.01 to 10 percent.

"It's a difficult number," he said.

In the end, efficiency will have to be gauged at each storage site -- part of the reason for the advanced characterization efforts funded by the stimulus. And it will have one of the largest roles in whether a storage project goes forward, because it will control the space needed: say, whether a project is 2 square miles large, or 200 square miles.

'Learning as you go'

Beyond efficiency, there are a host of other factors that could drop the practical storage reserves. A large issue will be injectivity, whether the massive volume of CO2 produced by coal-fired plants can be injected underground without a burdensome number of wells. Some formations will have poor sealing rocks, unable to resist added pressure. And saltwater displacement could kill other potential projects (Greenwire, April 23).

Currently, USGS is revising the methods used by DOE to calculate large-scale storage estimates that will begin factoring in these risks, which are not now considered.

Water displacement could be particularly problematic. As Burruss pointed out in congressional testimony, the lifetime emissions from one large coal-fired power plant would displace water equal to the volume of about 4.1 billion oil barrels -- the size of a "giant" oil field. Few formations, he said, can easily handle such displacement.

The CCS estimate methods, which are available online in a draft form, are being revised. But given the risks being added to their calculations, USGS seems unlikely to increase reserve estimates.

In the end, perhaps, the conversation should not necessarily be about gross estimates, which will fluctuate. It is best to remain conservative and then see what you learn about the subsurface as projects develop, Burruss said. That is how geologists have long operated with oil, gas, uranium -- you name it. Every project will yield mistakes and knowledge, he said.

"In general, this is a process of learning as you go," Burruss said.

Even Neele, the Dutch geologist deeply skeptical of storage capacity, said the absolute amount of underground reservoirs could compensate for the inconstancy of storage estimates.

"The good news," Neele said, "is there's a lot of space. We can accommodate a lot of uncertainty."