As we transition away from fossil fuels, one way for us to have our cake and eat it too is to capture the CO₂ before it reaches the atmosphere and stick it back down in the ground. That can be done by pumping it into the same reservoirs that once held oil and gas or into deep, saline aquifers. While that CO₂ will gradually dissolve and eventually form carbonate minerals, in the meantime, you're relying on the integrity of the rocks to provide the container that keeps the CO₂ locked away.

Injecting CO₂ beneath the seafloor is also an attractive option, but questions have remained about the ecological effects of a CO₂ leak on the ocean floor. A newly published study created an artificial leak off Scotland's western coast to measure its impact; the work was done by a large group of researchers led by Plymouth Marine Laboratory's Jerry Blackford, the Scottish Association for Marine Science's Henrik Stahl, and the University of Southampton's Jonathan Bull.

They drilled a horizontal borehole out to a point 11 meters (slightly more than 36 feet) below the seafloor, beneath twelve meters of water. They monitored and sampled that area while injecting CO₂ for about five weeks. The injection started out slow, increasing over time. Without a barrier to keep it below the seafloor, some of the CO₂ escaped upward.

About 15 percent of the injected CO₂ bubbled up from the seafloor as gas. The rest dissolved into the water in the sediments below the seafloor. By emitting waves of sound energy, the researchers were able to see some of what was going on below the surface. At lower injection rates, the gas passed through a network of fractures in fine-grained mud before diffusing less violently through the sandier sediment near the surface. When the rate of injection increases, the 5- to 10-meter-wide "chimney" of fractures punched through the sand, reaching the surface.

Measured within the top 25 centimeters of sediment, pH dropped a bit in the final week of the experiment. It was also clear that calcium carbonate was being dissolved, buffering the pH change from being larger. The chemical measurements returned to their normal values within a couple of weeks after the CO₂ injection ceased, partly because the chemical equilibrium was restored and partly because the movement of water flushed the affected area.

Was that enough to mess with the organisms living there? Toward the end of the experiment, there was a shift in the community living in the mud, which they compared with another community away from the test site. But like the chemical changes, the community returned to normal in a couple of weeks. The researchers also measured a shift in microbial activity that took a little longer to return to normal—about three months.

The experimental CO₂ "leak" was small, but it was within the (very broad) range of possible leakage scenarios proposed for CO₂ storage projects. On the low end of these potential scenarios, you have the possibility for leakage from old exploratory oil and gas boreholes that could be imperfectly plugged—this could be considerably smaller than the experiment. At the high end, there's the potential for mechanical failure of the injection well, which would at least be fixable. Serious leaks through fractures in the rock are the wildcard, but this isn't considered likely.

The researchers conclude, "This emerging understanding synthesizing dispersion, impact, and recovery suggests that small-scale leakage from [CO₂ storage], should it reach the sea floor, is highly unlikely to have a regionally significant environmental impact."

The practical question is whether we could detect these leaks should they occur. The researchers say that monitoring for small leaks is doable, but it takes some work. Background conditions—including seasonal changes—would have to be well-studied in order for anomalous events to be identified. And they suggest that small roving underwater vehicles sniffing chemistry and patrolling for bubbles would be the best way to monitor these projects, ensuring that we're aware if any CO₂ finds its way out of storage.

Nature Climate Change, 2014. DOI: 10.1038/NCLIMATE2381  (About DOIs).