Fracking fallout: 7.9 million at risk of man-made earthquakes

[Deleted member 159002]

http://money.cnn.com/2016/03/29/investing/earthquakes-fracking-usgs-oil-gas/

Government scientists believe these quakes are mostly triggered by human activity, primarily the disposal of waste water as part of oil and natural gas drilling.

About 7.9 million people are now at risk from these man-made earthquakes, including certain regions in Oklahoma, Kansas, Colorado, New Mexico, Texas and Arkansas, the U.S. Geological Survey said this week. Even though these earthquakes don't factor in building-code maps, they create a hazard to buildings, bridges, pipelines and other key structures, according to the government agency that studies the U.S. landscape.

War Roomers who are pro-fracking- how do the benefits of fracking outweigh the harms? How much do you think these companies should compensate people in the regions where they do business?

 

[Livermore]

I found this bit in the article to be rather interesting.

The agency also said its forecast did not explore the causes of increase seismic activity and stressed that more research is needed in that area. It noted that some natural quakes could be scattered throughout its list of suspected man-made ones.

How are you going to argue that seismic activity is up in certain areas and blame it on fracking when your research doesnt even bother to explore the causes of the seismic increases? I'm no statistician but I know correlation doesn't equal causation. Now that's not to say fracking isn't an issue and that the negative affects shouldn't be explored further, but you're going to need more evidence and less shotty research than the one done by this agency before anyone is going to take anti-fracking people seriously.

 

[Deleted member 159002]

I found this bit in the article to be rather interesting.



How are you going to argue that seismic activity is up in certain areas and blame it on fracking when your research doesnt even bother to explore the causes of the seismic increases? I'm no statistician but I know correlation doesn't equal causation. Now that's not to say fracking isn't an issue and that the negative affects shouldn't be explored further, but you're going to need more evidence and less shotty research than the one done by this agency before anyone is going to take anti-fracking people seriously.

The evidence is there. You sound willfully ignorant.

https://stateimpact.npr.org/texas/tag/earthquake/

 

[SODACKER]

Fracking doesn't cause earthquakes. The disposal of the hydraulic fluid does. It has been determined that the safest place to dispose of that stuff is deep in the ground. So the decision needs to be is it worthwhile to continue to dispose of this stuff this way if it cause small earthquakes. I say yes, because these aren't going to cause some California style devastating quake, and I think we need to fully vet out this strategy of disposal because I believe we will be doing it with Co2 before to long.

If it really is an issue we will find another way to dispose of the hydraulic fluid, fracking will get more expensive and we will lean on the middle east for more of our oil.

 

[JDragon]

I think this is really where it comes into play that the US is a huge country. Germany has 80 million living in an area the size of Montana. Fracking (or assoicated procedures) would have good chances to cause minor quakes that demolish private property and houses here. We already have issues with geothermics.

 

[44nutman]

They are trying to pass a law in Florida, that cities can not ban fracking. Of course we have the most corrupt governor in history of governors, so he must be getting a taste of all the fracking money.
The thing with Florida, is we have low water tables, which leads to our sink hole problem. Fracking would be a disaster in my county. My county has pass a law to ban fracking, but the State will not allow that to be enforced. I love when people crow about less government are usually from the same party that pass bullshit laws that do not allow cities to dictate their own policy.

 

[Uchi Mata]

How are you going to argue that seismic activity is up in certain areas and blame it on fracking when your research doesnt even bother to explore the causes of the seismic increases? I'm no statistician but I know correlation doesn't equal causation. Now that's not to say fracking isn't an issue and that the negative affects shouldn't be explored further, but you're going to need more evidence and less shotty research than the one done by this agency before anyone is going to take anti-fracking people seriously.

So I am a statistician, or more precisely a data scientist, and I can tell you that the trope that 'correlation doesn't equal causation' is highly abused. It is certainly true that correlation is not sufficient on its own to prove causation, but it's a pretty damn good indicator and working statisticians use it all the time when discussing causation. In fact, the most common statistic cited for the strength of regression relationships (models which seek to establish the nature and strength of a causal link between variables), R^2, is essentially a measure of correlation since it's based upon the size of residuals which are essentially no more than measures of correlation.

The other thing to keep in mind is that when you're evaluating a process and trying to determine deviations from that process and if something has changed in a global way, one important thing to determine is how stable that process has been and for how long. Given the extensive seismic record the USGS keeps, the normal range of seismic activity is pretty well understood. When you see a major deviation from that distribution, it's completely sensible to assume that it's probably not natural since you have a really good notion of what the natural variation in activity looks like. And it's also perfectly reasonable to attribute the increase in seismic activity to a process of resource extraction that you know impacts the structure of the Earth's crust, explicitly weakening it (that is after all the whole point of fracking). You can certainly say that there's some doubt as to whether or not fracking is the true cause, but that doesn't make it shoddy science to say it's the most likely cause. You literally cannot remove all doubt from any scientific result, and pointing out that doubt still exists is the scientific example of the Cartesian cogito. Yeah, it does. So what? Doesn't mean we don't take drugs when we're sick even though there's some (theoretical) doubt as to whether or not they make any difference. The standard in science is/should be preponderance of evidence, combined from a practical decision making standpoint with the cost/benefit calculus of being right or wrong.

 

[Livermore]

The evidence is there. You sound willfully ignorant.

https://stateimpact.npr.org/texas/tag/earthquake/

I'm willfully ignorant because I question the veracity of a government agencies forecasts? You do realize that in my OP I never denied that fracking could lead to man-made earthquakes?

 

[Livermore]

So I am a statistician, or more precisely a data scientist, and I can tell you that the trope that 'correlation doesn't equal causation' is highly abused. It is certainly true that correlation is not sufficient on its own to prove causation, but it's a pretty damn good indicator and working statisticians use it all the time when discussing causation. In fact, the most common statistic cited for the strength of regression relationships (models which seek to establish the nature and strength of a causal link between variables), R^2, is essentially a measure of correlation since it's based upon the size of residuals which are essentially no more than measures of correlation.

The other thing to keep in mind is that when you're evaluating a process and trying to determine deviations from that process and if something has changed in a global way, one important thing to determine is how stable that process has been and for how long. Given the extensive seismic record the USGS keeps, the normal range of seismic activity is pretty well understood. When you see a major deviation from that distribution, it's completely sensible to assume that it's probably not natural since you have a really good notion of what the natural variation in activity looks like. And it's also perfectly reasonable to attribute the increase in seismic activity to a process of resource extraction that you know impacts the structure of the Earth's crust, explicitly weakening it (that is after all the whole point of fracking). You can certainly say that there's some doubt as to whether or not fracking is the true cause, but that doesn't make it shoddy science to say it's the most likely cause. You literally cannot remove all doubt from any scientific result, and pointing out that doubt still exists is the scientific example of the Cartesian cogito. Yeah, it does. So what? Doesn't mean we don't take drugs when we're sick even though there's some (theoretical) doubt as to whether or not they make any difference. The standard in science is/should be preponderance of evidence, combined from a practical decision making standpoint with the cost/benefit calculus of being right or wrong.

Thanks for your post. As a data scientist and statistician (I'm assuming you're not lying) I respect your opinion but surely you could see how someone who isn't a data scientist and or isn't a statistician could read the article in the OP, particular the part i quoted, and come away skeptical right?

I'm not denying that fracking probably causes man-made earthquakes, as you noted, that is a reasonable conclusion given the nature of fracking. However, shoddy stats are used all the time as evidence for the implementation of policy that's either unnecessary and or causes more harm than good. My point is that we should get more reliable data before we start looking for the government to step in and impose heavy-handed regulations. I feel the same way about climate change. While I don't deny that climate change is partly man-made, I do think a lot of scientists on both sides are doing less than respectable research to come up with conclusions that satisfy their ideological priors. These tainted results are then used to try and pass sweeping legislation that affects everyone and I just feel we need to be more cautious that's all.

 

[Uchi Mata]

Thanks for your post. As a data scientist and statistician (I'm assuming you're not lying) I respect your opinion but surely you could see how someone who isn't a data scientist and or isn't a statistician could read the article in the OP, particular the part i quoted, and come away skeptical right?

I'm not denying that fracking probably causes man-made earthquakes, as you noted, that is a reasonable conclusion given the nature of fracking. However, shoddy stats are used all the time as evidence for the implementation of policy that's either unnecessary and or causes more harm than good. My point is that we should get more reliable data before we start looking for the government to step in and impose heavy-handed regulations. I feel the same way about climate change. While I don't deny that climate change is partly man-made, I do think a lot of scientists on both sides are doing less than respectable research to come up with conclusions that satisfy their ideological priors. These tainted results are then used to try and pass sweeping legislation that affects everyone and I just feel we need to be more cautious that's all.

I think caution is always warranted when you're talking about massive, expensive government intervention. And certainly statistics can be heavily abused, often in ways that are non-obvious to laymen (for example, look at the reproducability crisis in psychology, which is largely a function of poor statistical practice on the part of the original researchers). That being said, you have to weigh not only the costs of action, but also the costs of inaction alongside the likelihood that your hypothesis is correct. You take climate change, yeah it's incredibly expensive to address, and there is some disagreement over the severity of likely effects (though in the serious climate research community there's no real disagreement that humans are causing the Earth to warm and that it's not a good thing), but even at the low end of predicted severity it's a lot more expensive to ignore than address. Chalking the accumulated research up to people satisfying their ideological priors strikes me as both extremely cynical and as, ironically, a case of denying the evidence based primarily on ideology. I suspect fracking and seismic activity will end up in the same boat.

I'll note too that it's important when you're looking at science with political/economic implications to try and go beyond news stories, because news reported by laymen (in terms of not being scientists) is almost always going to sensationalize and distort the actual scientific results, often in ways the researchers themselves are probably uncomfortable with. In my experience I find most scientists are pretty cagey about how conclusive their results are and generally draw very limited conclusions at most from any specific study, whereas the media does the opposite. I spend an inordinate amount of time in my job downplaying the significance of my findings and trying to get folks not to jump off the deep end with my team's results, because the nature of data interpretation involves inherent uncertainty which many people are very uncomfortable with (and thus they act as if the data is more certain than it is).

 

[Livermore]

I think caution is always warranted when you're talking about massive, expensive government intervention. And certainly statistics can be heavily abused, often in ways that are non-obvious to laymen (for example, look at the reproducability crisis in psychology, which is largely a function of poor statistical practice on the part of the original researchers). That being said, you have to weigh not only the costs of action, but also the costs of inaction alongside the likelihood that your hypothesis is correct. You take climate change, yeah it's incredibly expensive to address, and there is some disagreement over the severity of likely effects (though in the serious climate research community there's no real disagreement that humans are causing the Earth to warm and that it's not a good thing), but even at the low end of predicted severity it's a lot more expensive to ignore than address. Chalking the accumulated research up to people satisfying their ideological priors strikes me as both extremely cynical and as, ironically, a case of denying the evidence based primarily on ideology. I suspect fracking and seismic activity will end up in the same boat.

I'll note too that it's important when you're looking at science with political/economic implications to try and go beyond news stories, because news reported by laymen (in terms of not being scientists) is almost always going to sensationalize and distort the actual scientific results, often in ways the researchers themselves are probably uncomfortable with. In my experience I find most scientists are pretty cagey about how conclusive their results are and generally draw very limited conclusions at most from any specific study, whereas the media does the opposite. I spend an inordinate amount of time in my job downplaying the significance of my findings and trying to get folks not to jump off the deep end with my team's results, because the nature of data interpretation involves inherent uncertainty which many people are very uncomfortable with (and thus they act as if the data is more certain than it is).

Good post. I think you make some good points especially the one about weighing not only the cost of action, but also the cost of inaction. And again here I want to reiterate the fact that I'm not opposed to fracking legislation and I do not deny that fracking probably is responsible for higher levels of seismic activity in and around the areas where fracking takes place (you're last sentence in your 1st paragraph seems to imply that I've taken a position on this issue and dismissed the "accumulated research" when really I'm just referring to this 1 particular study).

It's funny that you bring up the reproducibility crisis in psychology because that crisis is basically the basis for my skepticism. You have all these social scientists cooking up "important" research and studies yet they refuse to let you in the kitchen to see how the food was prepared. I'd be willing to bet problem doesn't just stop with psychology either and I'm sure it extends to most if not all the social sciences. It'd be interesting to see if they begin a reproducibility project in economics which is another field in where the empirical research seems so suspect.

 

[Uchi Mata]

Good post. I think you make some good points especially the one about weighing not only the cost of action, but also the cost of inaction. And again here I want to reiterate the fact that I'm not opposed to fracking legislation and I do not deny that fracking probably is responsible for higher levels of seismic activity in and around the areas where fracking takes place (you're last sentence in your 1st paragraph seems to imply that I've taken a position on this issue and dismissed the "accumulated research" when really I'm just referring to this 1 particular study).

It's funny that you bring up the reproducibility crisis in psychology because that crisis is basically the basis for my skepticism. You have all these social scientists cooking up "important" research and studies yet they refuse to let you in the kitchen to see how the food was prepared. I'd be willing to bet problem doesn't just stop with psychology either and I'm sure it extends to most if not all the social sciences. It'd be interesting to see if they begin a reproducibility project in economics which is another field in where the empirical research seems so suspect.

I'd guess it's probably the case in almost all social sciences. Despite my background in the social sciences (psychology specifically), I don't have hardly any faith in their methods simply because they're too fuzzy in terms of being able to make testable predictions that repeatedly turn out to be accurate. A lot of that has to do with the way that studies can be subtly manipulated to provide statistically significant results. For example, you can easily manipulate the length of a study to try and ensure significance simply by not stopping it until you get the result you want or stopping early once you have your result. It's not easy to police that sort of thing, and because of the imperative to publish significant studies there's a lot of pressure on researchers to get results out there however they can. Sometimes it's just plain hard, since under the typical frequentist framework the same data can be significant or not depending on how you structure your experiment. What the hell? If you have the same data, it should be significant or not but that's not the case (for example, flipping a coin 20 times and getting 13 heads vs. flipping the coin until you get 13 heads and it takes 20 flips have different levels of statistical significance).

Economics is even weirder, since you don't really get to run experiments in many cases and have to rely on extant historical data. Makes it very hard. What data to include or not include can make your study show vastly different results.

 

[ShadowRun]

Man made earthquakes like project destiny? looks like the Core was predicting the future.

 

[AnGrYcRoW]

They are trying to pass a law in Florida, that cities can not ban fracking. Of course we have the most corrupt governor in history of governors, so he must be getting a taste of all the fracking money.
The thing with Florida, is we have low water tables, which leads to our sink hole problem. Fracking would be a disaster in my county. My county has pass a law to ban fracking, but the State will not allow that to be enforced. I love when people crow about less government are usually from the same party that pass bullshit laws that do not allow cities to dictate their own policy.

Doesn't Florida get all its fresh water from aquifers, seems like fracking would be a bad idea in that scenario.

 

[Livermore]

I'd guess it's probably the case in almost all social sciences. Despite my background in the social sciences (psychology specifically), I don't have hardly any faith in their methods simply because they're too fuzzy in terms of being able to make testable predictions that repeatedly turn out to be accurate. A lot of that has to do with the way that studies can be subtly manipulated to provide statistically significant results. For example, you can easily manipulate the length of a study to try and ensure significance simply by not stopping it until you get the result you want or stopping early once you have your result. It's not easy to police that sort of thing, and because of the imperative to publish significant studies there's a lot of pressure on researchers to get results out there however they can. Sometimes it's just plain hard, since under the typical frequentist framework the same data can be significant or not depending on how you structure your experiment. What the hell? If you have the same data, it should be significant or not but that's not the case (for example, flipping a coin 20 times and getting 13 heads vs. flipping the coin until you get 13 heads and it takes 20 flips have different levels of statistical significance).

Economics is even weirder, since you don't really get to run experiments in many cases and have to rely on extant historical data. Makes it very hard. What data to include or not include can make your study show vastly different results.

The bolded part is what I find most interesting about this. An overwhelming majority of researchers and scientists are good intentioned people not trying to commit fraud right. The problem lies in the incentives. Success for a practicing researcher or scientist is contingent on getting as much of your work published as possible in the most prestigious journals as possible. However, the research most likely to get published in these prestigious journals isn't necessarily the research with the "best" results but research whose results are considered most beautiful, positive, and clean. The rub (as you well know) is that most research isn't beautiful, positive, or clean and so researchers and scientists are incentivized to do sketchy work. Add to this the fact that some researchers do go into the lab with ideological priors and what we end up getting is bunch of bad science. It's unfortunate.

 

[ben236]

What is the average strength of these earthquakes? Which I would think would be more important than the frequency.

 

[mjmj]

Considering fracking has been going on for DECADES, I think we have a good idea of the affects. Little to none.

 

[2DUM2TAP]

So I am a statistician, or more precisely a data scientist, and I can tell you that the trope that 'correlation doesn't equal causation' is highly abused. It is certainly true that correlation is not sufficient on its own to prove causation, but it's a pretty damn good indicator and working statisticians use it all the time when discussing causation. In fact, the most common statistic cited for the strength of regression relationships (models which seek to establish the nature and strength of a causal link between variables), R^2, is essentially a measure of correlation since it's based upon the size of residuals which are essentially no more than measures of correlation.

The other thing to keep in mind is that when you're evaluating a process and trying to determine deviations from that process and if something has changed in a global way, one important thing to determine is how stable that process has been and for how long. Given the extensive seismic record the USGS keeps, the normal range of seismic activity is pretty well understood. When you see a major deviation from that distribution, it's completely sensible to assume that it's probably not natural since you have a really good notion of what the natural variation in activity looks like. And it's also perfectly reasonable to attribute the increase in seismic activity to a process of resource extraction that you know impacts the structure of the Earth's crust, explicitly weakening it (that is after all the whole point of fracking). You can certainly say that there's some doubt as to whether or not fracking is the true cause, but that doesn't make it shoddy science to say it's the most likely cause. You literally cannot remove all doubt from any scientific result, and pointing out that doubt still exists is the scientific example of the Cartesian cogito. Yeah, it does. So what? Doesn't mean we don't take drugs when we're sick even though there's some (theoretical) doubt as to whether or not they make any difference. The standard in science is/should be preponderance of evidence, combined from a practical decision making standpoint with the cost/benefit calculus of being right or wrong.

NERD

 

[Uchi Mata]

NERD

Yeah. But it pays well.

 

[44nutman]

Doesn't Florida get all its fresh water from aquifers, seems like fracking would be a bad idea in that scenario.

That is what everyone is arguing, but the bought and paid for State Congress are trying to make it illegal for cities or counties to pass laws, banning fracking. The grift is so obvious.