Shakin’ Up the Plains, A Tale of Three States
Its been a busy week or so for earthquakes and the oil and gas industry. In Colorado, a wastewater well was shut down after the Colorado Oil and Gas Commission requested that they do so while the cause of a series of small earthquakes is investigated. In Kansas, in contrast, a strong increase in seismicity in an area with extensive oil and gas activity remains controversial as progress on a commission set up by the governor last February remains slow. And in Oklahoma, things are hopping as residents of Edmond, Oklahoma, met to consider banning industry activity producing earthquakes in the area. [We’ll leave Texas to the side for now; it is a separate challenge].
There are some priceless comments justifying the continued use of the injection wells coming from the Oklahoma meeting. GG’s fav:
[Oklahoma Geological Survey seismologist Austin] Holland said stopping the use of injection wells, which pump water deep underground, would not be recommended from a scientific standpoint because that would rob researchers of valuation data that could help them figure out how to prevent earthquakes.
One can only hope he said “valuable data”. Anyways, this is silly as stopping injection would also produce valuable data (arguably, seeing the seismicity die off or continue would be pretty diagnostic–this was pretty clear in the case of the seismicity from the Rocky Mountain Arsenal in the 1960s). Pretending that continued injection is a valuable scientific experiment instead of an expensive commercial activity is ridiculous and casts real doubt on the professionalism of the Oklahoma Geological Survey. One can hope the media quote was sufficiently out of context that this wasn’t as silly as it sounds.
(And the industry apologists are out in force, marshaling bad statistics and extreme cherry-picking to claim that this seismicity is all from a global increase in earthquakes. Please, VP of exploration, go back to your day job… oh, and by the way, there was a lot of oil and gas activity in the 1950s, too, so I wouldn’t be too quick to cast that episode of seismicity as purely natural).
Why the different responses? In Colorado, a CU group immediately put out sensors, suspecting that an unusual event like the Greeley earthquake might be associated with oil and gas work. They informed the industry and government overseers of their progress, and the relatively sparse number of injection wells and association of seismicity with a fairly new well made it unusually clear that that well was a very likely suspect. We’ll see in coming months what happens…
In Kansas, small earthquakes have occurred in the oil patch for a long time (many of the older ones might actually be due to withdrawal of oil rather than injection of fluid–you can generate earthquakes both ways); the oil plays are, by and large, well away from population centers. So by the time the state got interested things were already a bit complex.
In Oklahoma, though, you can see in action the old adage that you can’t get somebody to understand something when their paycheck relies on them not understanding it. Although the increase in seismicity has been noticeable for some time and the association with increased injection activity was pretty strong, the chorus of denial from local academics, industry members, and the geological survey bordered on delusional, though the recent about-face from the geological survey (which joined the USGS in making a public warning about the possibility of a damaging earthquake) shows that reason (and a lot of felt earthquakes) can have an impact. But the shear size of industry in Oklahoma and the lengthy history of injection activity mean it is awfully hard to point the finger at individual wells.
Well, that was until this past week, when a former OU researcher (now at Cornell…can’t imagine why she’d move…) led a research group (including CU hydrologists) that argues that a lot of the Oklahoma earthquakes could be the result of injection at four particular high-volume wells (Science has a nice summary of events associated with this work; the paper itself is presumably behind a paywall). The nice things about this work is that the hydrologic modeling shows how earthquakes distant from a well can be caused by actions at the well (this is not really a news flash for those in the know; injection wells in southwestern Colorado that pump saline water into deep layers have produced significant seismicity quite some distance from the injection wells; because the area is remote and the Bureau of Reclamation has not typically made the raw data public, this isn’t so well known). Of course the corporation running the wells isn’t convinced (though I would not want to be in their shoes in a civil suit should anybody get hurt).
Now the argument is made that the earthquakes could be natural and we need more study. (Sound familiar? This has a close parallel in the argument that global warming could be natural). Yes, there are earthquakes in the interior of the U.S.; what this means is that there is stress in the crust at levels capable of producing earthquakes. So if you do something that reduces the resistance of a fault to stress, it is likely to rupture because the forces are already present. Does this mean that the fault would rupture at some later time anyway? It is hard to say; it could be that stress level remain quite high for very long times with little or no seismicity. But when you see strong increases in seismicity across the country associated with new or increased use of injection wells, it seems imprudent (and frequently self-serving) to argue that this is a coincidence.
Does this mean all injection wells should be shut down? The group publishing in Science says no; there are a lot of injection wells that seem to operate with no effect on seismicity. It means there are some situations–and they might not be obvious beforehand–where injection wells will cause earthquakes. It means that regulators need to have information that they can share on well volumes and pressures. It means that operators need to include in their business plan the possibility that a well will need to be run at lower volumes or even shut down. It probably means we need to have better seismic networks where disposal wells are being used.
Are there alternatives to injection wells? The problem is that fracking does produce a lot of contaminated water, water with heavy metals and other evils that most treatment plants cannot deal with (and some of the reservoirs being exploited now also have a lot of naturally briny waters associated with them). The two main alternatives are to develop treatment plants that can deal with this or to leave the water in lined ponds to evaporate, then disposing of the salts left behind. The first is expensive and could make many fields subeconomic. The second is hazardous as such lined ponds have frequently leaked or failed, contaminating shallow drinking-water aquifers or even surface waters.