Reaction to research into carbon dioxide leaks highlights growing tensions on carbon storage, as concerns are growing about stored carbon leaking into the water supply.
Late last year, US biologists from Duke University, North Carolina, unveiled results from a year-long experiment that indicated injecting carbon dioxide deep underground could contaminate fresh groundwater more than previously thought. US media pounced on the research, with the New York Times provocatively asking ‘What if Captured Carbon Makes a Getaway?’ Just as swiftly, the Scottish Carbon Capture and Storage Consortium issued a sharp rebuttal, labelling the work a ‘Jacuzzi experiment’.
While the researchers have since expressed surprise at the attention their work received, the reactions point to growing tension around carbon capture and storage; could it leak? From the beginning researchers have stated that if carbon dioxide were to leak from injection formations into groundwater supplies, chemical reactions could release harmful contaminants from sediment, possibly rendering the water undrinkable. However, the Duke University study hit a nerve.
Funded by the US Department of Energy, Professor Rob Jackson and Dr Mark Little from the Duke University Center on Global Change set out to discover how carbon dioxide leaks from deep saline storage sites, hundreds to thousands of metres underground, might affect water quality in overlying shallow drinking-water aquifers. To mimic a slow gas leak from a storage site, the researchers bubbled carbon dioxide through aquifer rock samples placed in water. The samples had been taken from four freshwater aquifers in the US, sited above potential carbon storage sites.
As expected the hypothetical groundwater became more acidic, dissolving minerals in the rock samples. Metal ion concentrations fluctuated throughout the trials but the researchers highlighted two key results. First, concentrations of transition metals including manganese and iron increased by more than 1,000 per cent compared to control samples, and second, levels of uranium and barium rose throughout experiments.
“I did not expect concentrations of important factors to go up by a factor of ten or more,” says Jackson, director of the Center on Global Change. “And the fact that some of these elements, including uranium, manganese and iron, are important from a health standpoint has been a note of caution for me.”
As he says, levels of iron and manganese exceeded the maximum contaminant levels recommended by the US Environment Protection Agency (EPA), while the time- dependent trend exhibited by uranium and barium is a worry. After a year, levels of these heavy metals were within EPA guidelines but given more time would this change?
This is just one of many questions the researchers hope future research will answer. Right now, however, both Jackson and Little are confident their research has at least identified markers that scientists could use to test for early warnings of gas leaks.
“Concentrations of manganese, iron and calcium could all be used as geochemical markers of a leak, as their concentrations increase within two weeks of exposure to carbon dioxide,” explains Jackson. “Such changes may even be detectable long before direct changes in carbon dioxide are observed.”
And as Little adds: “One of the most important take-homes from this study is to monitor groundwater. Even after you’ve selected the best site, done all the best engineering, it’s still important to monitor the groundwater [for contamination].”
But while the technique may prove useful, Professor Stuart Haszeldine from Edinburgh University, Scotland, does little to hide his incredulity at the Duke University research. Also a researcher for the Scottish Carbon Capture and Storage (SCCS) Consortium, and joint author of its response to Jackson and Little’s published research, he believes the biologists “immensely over-emphasised” the possible consequences of a carbon dioxide leak.
Disputing many aspects of the researcher’s experimental technique, he first points to their use of disaggregated rock sediment mixed with a large water to rock ratio. “These samples will always result in maximum chemical reaction,” he says. “By taking the samples into laboratory glassware, the whole content of the sediment is exposed for chemical reaction; old grain surfaces and new surfaces.”
He also believes the researchers used an unrealistic carbon dioxide flux when bubbling the gas through their samples that, if scaled up from their laboratory bottles to the real world, would equate to an “explosive leakage”. “If you have an unplanned leakage, it will be a seepage or a dribble of carbon dioxide, not this runaway leakage that Little and Jackson talk about,” he explains. “Our work on natural carbon dioxide deposits in Italy, the US and the North Sea, shows you do not get a catastrophic flow, but a widespread dispersal of carbon dioxide over an area of many square kilometres.”
Recent research at the Department of Environmental Engineering and Earth Sciences at Clemson University, South Carolina, supports Haszeldine’s comments on carbon dioxide seepage. Here Professor Ron Falta and colleagues are looking at the mobility of carbon dioxide after it is piped underground.
“We dissolve the gas in brine, which is what would happen down in the formation... and simulate what would happen if that brine were to move up a fault or if the formation was to become de-pressurised in some other way,” he explains. “Initial results appear that it would not be very mobile.”
Additional mathematical simulations back these findings and also show that if a leak did take place, it could be easily stopped. One possible scenario modelled by the researchers looks at what happens if carbon dioxide was to leak through an abandoned bore-hole connecting a deep aquifer formation with a shallow one. As Falta points out, the carbon-dioxide containing brine is denser than fresh water and would only move up the borehole if the deep-water aquifer was under greater pressure than the shallower formation.
“If, for example, the deep formation was over-pressurised, fluid could be forced up the well-bore, and when this happens, carbon dioxide comes out of the brine solution so you have gas in the shallow formation,” he says. “Importantly, we have found that this process doesn’t run away. If you stop the pressure imbalance, this [gas dissolution] also stops, in other words it doesn’t feed on itself. This means if you detect a leak you can correct it by altering the pressure of the formations.”
Haszeldine also questions the way in which the Duke researchers present their results, suggesting they used the most extreme values, for example, the 1,000 per cent rise in transition metal concentrations and ongoing rise in uranium and barium concentrations. Among a host of other criticisms, he highlights how the researchers ‘deliberately’ chose aquifers which were already high in trace metals.
Taking it to the extreme...
“People do need to be aware of what the possible outcomes of leakage into groundwater are, and research needs to be done,” he adds. “But I am not convinced that the Little and Jackson paper is a strong contribution towards helping us with that.”
Amid the criticism, both Jackson and Little emphasise that the purpose of their work has been to shed light on what could happen to freshwater aquifers in the event of a carbon dioxide leak from a deep water storage site. As Jackson says: “Sometimes you’re interested in average responses and other times, the extremes and unusual cases. The extremes are important here as this is where landowners might have problems.”
Little also highlights that his aquifer samples were sandy in texture and naturally disaggregated, and were not directly disturbed by the carbon dioxide, which was bubbled into the water, not through the sediments. As for the high water to rock ratio, he agrees it was higher than in a real aquifer but asserts this would actually dilute the measured metal concentrations.
On the flux rate, Little says his choice reflected the desire to explore these possibilities, but maintains the amount of carbon dioxide still represented a very small leak compared to the size of a proposed carbon capture and storage project. “And yes, we selected sediments from aquifers where trace metals were already present,” he adds. “You might have a shallow aquifer of pure quartz sand and if you bubbled carbon dioxide through it then there would never be a problem with metal contamination, so we skipped these locations.”
“No-one is ever going to mimic an actual project perfectly, and other groups are working on ‘practice’ releases in the field,” adds Jackson. “Our approach is just one of many approaches we need.”
Ultimately, Jackson stresses that both researchers believe carbon capture and storage is needed “as a partial solution to greenhouse gases”. “Our results do not lead me to the conclusion that we shouldn’t be doing carbon capture and storage, which was the [message] reported by many in the press as well as some advocacy groups,” he says.
And so the research into carbon storage continues apace. The US Department of Energy continues to fund research into the risks associated with geologic storage through its Carbon Sequestration Research Program. The Duke team, for one, is now continuing experiments on more samples from different locations as well as repeating the experiments under different oxidising conditions. Many shallow aquifers are not exposed to the atmosphere at all, which will significantly alter the way in which reactions take place.
Across the Atlantic research also continues. One team at Heriot-Watt University has just developed a process that promises to automatically seal small fissures in underground carbon dioxides stores before gas leaks to the surface. Meanwhile, myriad EU-funded initiatives are underway to locate and test potential storage sites, and build carbon capture and storage plants (see ‘UK industry developments’).
But will the rapid progress answer the very real concerns about the ability to store carbon safely? Recent comments posted on a US Natural Resources Defense Council Staff blog read: ‘Optimistic engineering and promises from the oil and gas industry have already led to disastrous consequences in the contamination of groundwater through fracking,’ and: ‘After the BP blowout in the Gulf, nobody is going to accept the guarantees from industry and government that carbon sequestration is safe.’
Well-founded or not, public fear will always be a strong deterrent as researchers and industry players know only too well. “If I was a scheming coal baron I might want to point out the potential dis-benefits and complexities of carbon capture and storage to delay its implementation. Everyone has a motive,” reflects Haszeldine.
“With a pure science motive, you do the experiments because [drinking water contamination] is an issue that could happen and we need to know more about it. But in an unbiased way.”