What's new
Frozen in Carbonite

Welcome to FiC! Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

EROI and Civilisation

Vorpal

Administrator
Administrator
What kind of energy sources are compatible with our modern way of life?

...

For much of human history, education, arts, and sciences have been an indulgence of a privileged few and associated with wealth. These activities could be thought of as supported by the excess production of the bulk of the population, which was primarily growing food. Some historical schools of economics, initially in China and later imported to France by its Sinologists, recognised this connection and conceptualised agriculture the basis of national wealth—a free gift of nature, which could be transformed by human activity but stays essentially fixed, at least from the point farming labour onward.

The industrial revolution is one reason that changed, by incentivising the labour theory of value, i.e. all labour rather than either a transformable free gift of nature or only agricultural labour. But there could be another direction to generalise: if one thinks of agriculture as a form of exploitation of solar energy (producing food that powers humans), at the base level industry added other forms of energy, chiefly fossil fuels, and the activities of society are still powered by excess energy in some form.

Hence although one doesn't have to develop or take seriously a general energetic theory of value, it's natural to ask the following question: just how much energy returned on investment does it take to support developed society in its modern form?
EROI = (amount of energy obtained by energy-gathering activity)/(amount of energy expended by energy-gathering)​
If it takes more energy to extract and refine oil than one gets from using it, there's not much point in doing it anymore. (Except possibly as a kind of battery due to its convenience, but then this activity should be supported by some other power source anyway.)

Some people have attempted to answer that question, in terms of a Maslow-like societal hierarchy of needs, which shouldn't be taken as notably accurate, but the exercise with oil as the basis is interesting enough by itself:
Lambert, et al. Energy, EROI and quality of life. Energy Policy, Volume 64, pp. 153-167
societal-maslow.png

The undisputed EROI king is hydroelectric, with typical published values in the 80-100 range, but is obviously only situationally useful. Historical oil has approached 50 but is now closer to 20 or so for world average, while poly-Si solar photovoltaics aren't too hot but rapidly improving (meta-study of 2000-2013 published values [pdf]: ∼10 average, but with a very noticeable upward trend throughout the years that probably didn't stop in 2013). Wind seems to be doing better in terms of this metric.

One problem with renewables, however, is that at this rate of adoption, maybe they can serve a quarter of the world energy usage... in half a century or so.

But what can be done about this?
 
Soooo, if I understand it correctly, solar (with current technology, and not mixed with anything else which is purely theoretical exercise of course) cannot support health care?
Also, there's transportation and transportation. Shipping will require seriously lower EROI than aviation, for example.
But what can be done about this?
According to this study (table 2), use hydro wherever possible, gas and wind where they're available locally in sufficient quantity, and nuclear for everything else.
 
Soooo, if I understand it correctly, solar (with current technology, and not mixed with anything else which is purely theoretical exercise of course) cannot support health care?
Well, even keeping in mind that the above are statistical estimates based on performance of oil-based economies, modern poly-Si solar PV can easily have EROI ≳ 15 or even higher in many places, making it at least sometimes competitive with global oil & gas (which is falling and is around 20 today). Solar varies a lot by location for obvious reasons, and in the meantime the technology has been improving a lot.

In other words, maybe it could in a lot of places, though the sheer gap between it and historical fossil fuels is concerning (meanwhile, global oil & gas EROI seems to be steadily falling as the best sites are used up). However, in terms of horizontal irradiance, Germany is about as sunny as Canada, so I can't say I really understand their mad dash for solar specifically.
 
Last edited:
However, in terms of horizontal irradiance, Germany is about as sunny as Canada, so I can't say I really understand their mad dash for solar specifically.
Hm, solar can be good for covering spikes caused by increased use of a/c in hot weather - which is usually also exceptionally sunny.
 
All of this seems to be ignoring the actual point of renewables, which is that fossil fuels incur a large number of negative downstream consequences while renewables do not. This isn't reflected in the EROI. Improving the EROI of renewables is certainly one way of making it more practical to switch to renewables, but it's hardly the only or even most important metric by which people would choose renewables over fossil fuels.


This just seems like an elaborate way to say, quantifiably, "Renewables aren't as economical as fossil fuels". Yeah, we already knew that, that's not why we're doing it.
 
This just seems like an elaborate way to say, quantifiably, "Renewables aren't as economical as fossil fuels". Yeah, we already knew that, that's not why we're doing it.
No, you're misreading things. If you take the statistical argument of the study, which is based specifically on EROI of oil, it is actually that long-term, reliance on petroleum will have us unable to support civilisation because the more energetically accessible petroleum can't last and its EROI is getting worse. Stated generally, this is not at all a deep conclusion (one might call it is very obvious long-term because the source is finite)—it was simply somewhat interesting to me that someone actually tried to statistically estimate how the energetic difficulty of accessing petroleum corresponds to performance of social institutions in a civilisation primarily based on oil*.

You can't get 'renewables are worse much than historical petroleum' from it without also 'present petroleum is much worse than historical petroleum, and gradually getting worse' (and both are generally correct, except for some minor wrinkles such as one the most consistently EROI-performing source in these kinds of studies is a renewable, though not a widely deployable one. Also, I've already emphasised that EROI of renewables, poly-Si solar PV as a particular example, is improving as technology develops, while global petroleum is falling. So if this can be hoped to at least roughly translate to other energy sources, that might be a rough guide for at which point they are good enough. However, areal problem with taking it too seriously is lack of standard methodology, especially across different energy types.

*If what you're getting from this is that we non-petroleum sources will never have high enough EROI, then the conclusion wouldn't even be against renewables per se, but simply a nihilistic position that that everyone is doomed regardless of any additional externalities or downstream consequences.
 
But what can be done about this?
You really can't. The best way that renewables can be used is as part of a supplement for nuclear reactors. Basically, renewables are very niche when it comes to placement and has periods where they're not even viable in the first place (solar and the night for example or no wind for windmills). Hydroelectric is the only one that is viable but all the viable dam locations have been used up (or so close that the difference is minimal) already. So that leaves us with one alternative: nuclear.

... and guess who killed nuclear in the western world for the most part? Yep, coal and oil companies.
 
Well, Russia's is betting more on a closed nuclear cycle and less on renewables. Hopefully France doesn't start going against nuclear either.
 
Well, Russia's is betting more on a closed nuclear cycle and less on renewables. Hopefully France doesn't start going against nuclear either.
Given that France's nuclear reactors are an anomaly in terms of the West in how much power is generated by them (well over three-quarters), with Japan looking to decommission theirs?

I wouldn't bet on it lasting. Period. If we are going to fight anti-nuclear, we'll probably have to treat it in a similar way as to Anti-Vaxx: a memetic hazard induced anti-society/civilization movement and treat it like any info/memetic hazard.
 
Well, Russia's is betting more on a closed nuclear cycle and less on renewables. Hopefully France doesn't start going against nuclear either.
I've read a lot about solar and wind projects in Russia in the last couple of years. I don't get why solar (subsidies?), but wind certainly has potential in some areas. In some isolated areas the main source of electricity are diesel generators, which get resupplied once or twice a year, depending on what weather allows. There wind could seriously decrease consumption of diesel fuel and prevent situations when the community runs out of fuel. I've read something about a similar situation in the Far East some years back, when in the end the US Coast Guard had to come to the rescue...
I think that for Russia the combination of gas + closed nuclear cycle + renewables as opportunistic sources of electricity is optimal. Also, not deploying renewables would lead to the same problem as with the large gas turbines in the past - lack of technological competence and serious lag in an important market.
 
No, you're misreading things. If you take the statistical argument of the study, which is based specifically on EROI of oil, it is actually that long-term, reliance on petroleum will have us unable to support civilisation because the more energetically accessible petroleum can't last and its EROI is getting worse. Stated generally, this is not at all a deep conclusion (one might call it is very obvious long-term because the source is finite)—it was simply somewhat interesting to me that someone actually tried to statistically estimate how the energetic difficulty of accessing petroleum corresponds to performance of social institutions in a civilisation primarily based on oil*.

You can't get 'renewables are worse much than historical petroleum' from it without also 'present petroleum is much worse than historical petroleum, and gradually getting worse' (and both are generally correct, except for some minor wrinkles such as one the most consistently EROI-performing source in these kinds of studies is a renewable, though not a widely deployable one. Also, I've already emphasised that EROI of renewables, poly-Si solar PV as a particular example, is improving as technology develops, while global petroleum is falling. So if this can be hoped to at least roughly translate to other energy sources, that might be a rough guide for at which point they are good enough. However, areal problem with taking it too seriously is lack of standard methodology, especially across different energy types.

*If what you're getting from this is that we non-petroleum sources will never have high enough EROI, then the conclusion wouldn't even be against renewables per se, but simply a nihilistic position that that everyone is doomed regardless of any additional externalities or downstream consequences.
Well, your post emphasized the disparity between renewables and oil while dismissing the pyramid as not "notably accurate", so I presumed you wanted to talk about the former rather than the latter.

I'm not sure that the EROI is actually all that useful a metric for discussing this either, because it seems to me that they are simply using it as a proxy for the price of energy (their entire calculation is based simply on joule per dollar, with a modifier based on "economic intensity"). And "cheaper energy results in better quality of life" is not a novel idea. The paper discusses the relationship between energy prices and QOL, but oddly they don't perform any analysis to demonstrate that their EROI has a better correlation with QOL than simply using the local cost of energy.

The notion of a social hierarchy of needs with arts at the top, I think, horribly misunderstands human psychology and trends across time. I'm not terribly worried about arts and healthcare being the first to disappear as we run out of fossil fuels. Part of the problem with the assumption that there is some minimum EROI necessary to maintain modern civilisation is the fact that improvements in energy efficiency are driven by high costs of energy, so as EROI decreases, so will that minimum EROI.
 
Last edited:
@Heliostorm No, it's the accuracy of the specific EROI numbers themselves that I have suspicious about (both in quoted the pyramid illustration or when studies find a particular EROI for a specific energy type), but the conclusion believable at least in an approximate sense.

I'm not sure what's wrong with performance arts being on top—presumably, this doesn't mean that some form of entertainment industry exists (it did in some form probably ever since agriculture did), but rather things orchestras, operas, museums, etc., which are made for a relatively well-to-do educated population, so it makes sense for these thing to have ultimately higher requirements than education itself to provide the general population with. Relative position to health care may be more uncertain and vary a bit across what individual societies value. US is more... exceptional in that sense than many others.

Price of energy itself is completely meaningless to quality of life, although one could take a ratio with some other quantity, (energy price)/(X). For example, X could be (population)/(money spent on energy) to get per capita energy usage, which is indeed known to correlate very well to quality of life, and the paper mentioned that. X could be some income statistic, such as average or median, but that seems worse than the previous choice.

The paper probably only mentions per capita energy usage because it's not the point to get the best possible predictor of QoL, but to try to see how well indicators that are proxies for physical EROI do. There could be very well be multiple statistics that correlate with it but not each other; for something as ultimately ambiguously multi-dimensional as QoL that is assured.

A lot of not-insignificant improvements in societal efficiency are possible but wouldn't really do much but work into that approximate disclaimer above at best. ... Also, I would almost always be personally better off driving my own car rather than using mass transit or riding a bicycle to get where I need or want to go. That's why people go for what is essentially a moral argument about downstream consequences, chiefly involving people that haven't yet been born. And compared to changes on that scale, things like driving smaller cars (which I do) are basically roundoffs. ... Ultimately, there are also efficiency limits; e.g. a significant chunk of electricity consumption is lighting, which with modern LEDs is already as close to the maximum physically possible limit of lumens per watt, assuming daylight-like output.

If EROI < 1, it makes absolutely no difference what your economic policies are or what your equipment efficiencies are. You simply won't have an energy source, but rather a glorified battery at best. If it's, say, 2, then it would take too much effort to build tractors and combines or transport food across vast distances, and you're better off trying to subsistence farm with horses or something. In an ancient society where is the chief sources of energy are grown food and grass, with the chief 'machines' powered by that energy are humans and beasts of burden, it takes the exploitation of a huge fraction of the population to maybe provide a nice life for some small elite class on top. And as you start find better and better energy sources, such as building windmills and waterwheels, you need less and less underclass for that (beasts of burden fall under that category too, ultimately); if a waterwheel didn't provide substantially more energy over its lifetime than it took to build, there wouldn't be much point in spending the effort in building it, especially considering the up-front costs of doing so.

In general terms, these are extremely obvious physical considerations, so the real issue is how much confidence we can place in specific estimates and/or how fuzzy an uncertain they are.
 
Okay before i read the rest of that, it should agriculture is still a form of labor, the nature of the peasantry just leads to different forms of alienation and different forms of class consciousness. Also, the ruling class views workers as a fixed asset I think, so until we cross a certain threshold the answer to your question will always be manpower.
 
@Heliostorm No, it's the accuracy of the specific EROI numbers themselves that I have suspicious about (both in quoted the pyramid illustration or when studies find a particular EROI for a specific energy type), but the conclusion believable at least in an approximate sense.

I'm not sure what's wrong with performance arts being on top—presumably, this doesn't mean that some form of entertainment industry exists (it did in some form probably ever since agriculture did), but rather things orchestras, operas, museums, etc., which are made for a relatively well-to-do educated population, so it makes sense for these thing to have ultimately higher requirements than education itself to provide the general population with. Relative position to health care may be more uncertain and vary a bit across what individual societies value. US is more... exceptional in that sense than many others.
"Arts and Other" meaning "orchestras, operas, and museums, etc." sounds like an extremely forced interpretation to me. Surely the relevant metrics here are not how many orchestras, operas, and museums we have. If one is referring to "the arts" in a general sense and excluding the most popular forms of it, i.e. movies, pop songs, television, etc., I'm not sure what relevance it actually has to discussing the function of modern civilization as a whole. Surely there is no reason to distinguish between performance arts and pop culture as forms of discretionary social amenities.

My criticism, which is one that can also be leveled at Maslow's hierarchy, is that many social and psychological needs are not truly higher up on any hierarchy, but rather are basic. In Maslow's hierarchy, for instance, "love and belonging" should really be in the same rung as "safety" and arguably even mixed in with physiological needs. Given how many examples there are of individuals choosing to prioritize psychosocial needs over even physiological needs under some circumstances, it's apparent that there is no clear hierarchy. Similarly, I would argue that what these authors term "discretionary" is actually essential. Societies cannot choose to eschew all social amenities in a time of crisis (some amenities, certainly, but then that's true with all economic functions).

Price of energy itself is completely meaningless to quality of life, although one could take a ratio with some other quantity, (energy price)/(X). For example, X could be (population)/(money spent on energy) to get per capita energy usage, which is indeed known to correlate very well to quality of life, and the paper mentioned that. X could be some income statistic, such as average or median, but that seems worse than the previous choice.

The paper probably only mentions per capita energy usage because it's not the point to get the best possible predictor of QoL, but to try to see how well indicators that are proxies for physical EROI do. There could be very well be multiple statistics that correlate with it but not each other; for something as ultimately ambiguously multi-dimensional as QoL that is assured.
It might not be the point of this particular article, but I think there is a need for someone who is making an argument about quality of life based on EROI demonstrate that their EROI is more meaningful than alternative measures.

The thing is that price is inherently a form of measure of ROI. In a free market, you pay X value for a unit of energy only if you determine that you can derive X value of utility from that unit of energy. So instead we could look at the cost of the selling price of a unit of fuel/the cost to extract it. Now, the values being measured are more nebulous, but given that this paper is basing its calculation of energy input on price and GDP anyways, I'm not convinced that their measure is actually a better measure of the ROI of energy.

If EROI < 1, it makes absolutely no difference what your economic policies are or what your equipment efficiencies are. You simply won't have an energy source, but rather a glorified battery at best. If it's, say, 2, then it would take too much effort to build tractors and combines or transport food across vast distances, and you're better off trying to subsistence farm with horses or something. In an ancient society where is the chief sources of energy are grown food and grass, with the chief 'machines' powered by that energy are humans and beasts of burden, it takes the exploitation of a huge fraction of the population to maybe provide a nice life for some small elite class on top. And as you start find better and better energy sources, such as building windmills and waterwheels, you need less and less underclass for that (beasts of burden fall under that category too, ultimately); if a waterwheel didn't provide substantially more energy over its lifetime than it took to build, there wouldn't be much point in spending the effort in building it, especially considering the up-front costs of doing so.

In general terms, these are extremely obvious physical considerations, so the real issue is how much confidence we can place in specific estimates and/or how fuzzy an uncertain they are.
I don't think that's true, at least for this specific measure of EROI.

For example, if you take their equation and apply to say, a loaf of bread, (Calories in a loaf of bread / (total calories consumed in US yearly/US GDP * price of bread)), the EROI of Walmart-brand white bread ($.88 per loaf) is apparently on the order of 100.

Now, it's late at night so I'm not going to do it now since data for pre-modern periods is much harder to find, but I would imagine if you did this same calculation for a loaf of bread in 1300s France, you would get a significantly different value. Their measure of EROI is inextricably linked to economic conditions (which is itself linked to technological efficiency), by incorporating GDP and price into the calculation. Their "energy intensity" modifier, Et/GDP, is essentially an inverse measure of the energy efficiency of the economy. So a simple way to improve the EROI of any energy source, for instance, would be for the government to subsidize it, driving down the price and therefore increasing the EROI. Another would be by improving societal and technological efficiency, increasing the domestic product per unit of energy consumption, driving down the energy intensity. This EROI is not any inherent physical property of an energy source; it is an economic property of that energy source under a given system.
 
Last edited:
Apologies; by equipment I meant things greater economy that the energy source powers. If your energy is obtained with EROI < 1, it no longer makes any difference how efficient your light-bulbs are, whether your society prefers individual cars or mass transit, or which things you try to tax and which you try to subsidise. ... Yes, technology in general affects EROI. That's e.g. why EROI of solar PVs has been going up: they tend to be grow provide a somewhat higher fraction of incident power, and more importantly, take less energy to manufacture.

"Arts and Other" meaning "orchestras, operas, and museums, etc." sounds like an extremely forced interpretation to me. Surely the relevant metrics here are not how many orchestras, operas, and museums we have. If one is referring to "the arts" in a general sense and excluding the most popular forms of it, i.e. movies, pop songs, television, etc., I'm not sure what relevance it actually has to discussing the function of modern civilization as a whole. Surely there is no reason to distinguish between performance arts and pop culture as forms of discretionary social amenities.
The paper specified 'perfomance art', which is usually interpreted as a subclass of 'fine art', so this particular gloss of 'arts & other' is probably appropriate. This is usually conceived as being primary created for aesthetic purposes, so no, most popular forms don't apply. The beautiful and the entertaining can of course overlap, but they're not equivalent.

As for a general criticism of a hierarchy of needs, well, in context one possible litmus test could be: if the going gets tough, especially in an energy crisis, are some categories of pubic funding more likely to be cut off than others? Is health care more likely to get cuts than education? Another might be: do things on 'top' have things on the 'bottom' as prerequisites in order to function as they do in modern society? And so forth. Some of these are not obvious (e.g. I don't know whether fine arts would usually be cut before health care, although it seems plausible in a typical democratic society), but I think interpreting things in those kinds of terms would be a more charitable interpretation, since there's no particular reason to take the label as anything other than an analogy.

It might not be the point of this particular article, but I think there is a need for someone who is making an argument about quality of life based on EROI demonstrate that their EROI is more meaningful than alternative measures.
To be clear, EROISOC is a relatively meaningful indicator of QoL, and it being so compared in the paper, but the big fudge is actually how well this EROISOC corresponds to physical EROI, which in depends how tightly EIR and EROI couple. In other words, the chain depends on the assumption that energy intensity ratio and EROI are proxies for each other, and for oil (and coal) there is at least some independent research to somewhat support that, unlike for entirely bread-powered societies. It is, however, very hampered by the fact that EROI studies do not themselves have standardised methodology and sometimes use economic scaling to estimate unknowns.

The thing is that price is inherently a form of measure of ROI. In a free market, you pay X value for a unit of energy only if you determine that you can derive X value of utility from that unit of energy. So instead we could look at the cost of the selling price of a unit of fuel/the cost to extract it. Now, the values being measured are more nebulous, but given that this paper is basing its calculation of energy input on price and GDP anyways, I'm not convinced that their measure is actually a better measure of the ROI of energy.
Good point, but I think again the authors want to make a link to a physical indicator rather than an economic one and so assume the above as a proxy. It's notable that they don't do anything like 'renewable EROI is below such-and-such number, so it will never work', or anything even vaguely of the sort.

So a simple way to improve the EROI of any energy source, for instance, would be for the government to subsidize it, driving down the price and therefore increasing the EROI.
Another assumption is a global free market for energy, which is not quite true but decently close for oil, which is incompatible with this scheme. Like I said to Wakko above, these numbers are specifically for oil-based economies (and I wouldn't trust it as anything remotely near exact even there).

Another would be by improving societal and technological efficiency, increasing the domestic product per unit of energy consumption, driving down the energy intensity. This EROI is not any inherent physical property of an energy source; it is an economic property of that energy source under a given system.
That's correct in general. How much one can jump to inherent EROI for oil and coal depend on assumptions somewhat supported by things like this.
 
The paper specified 'perfomance art', which is usually interpreted as a subclass of 'fine art', so this particular gloss of 'arts & other' is probably appropriate. This is usually conceived as being primary created for aesthetic purposes, so no, most popular forms don't apply. The beautiful and the entertaining can of course overlap, but they're not equivalent.

As for a general criticism of a hierarchy of needs, well, in context one possible litmus test could be: if the going gets tough, especially in an energy crisis, are some categories of pubic funding more likely to be cut off than others? Is health care more likely to get cuts than education? Another might be: do things on 'top' have things on the 'bottom' as prerequisites in order to function as they do in modern society? And so forth. Some of these are not obvious (e.g. I don't know whether fine arts would usually be cut before health care, although it seems plausible in a typical democratic society), but I think interpreting things in those kinds of terms would be a more charitable interpretation, since there's no particular reason to take the label as anything other than an analogy.
But why? What distinguishes performance art or fine art from popular media such that it merits being separated out in terms of discretionary energy spending? How is one more or less discretionary than the other?

I'm pretty sure that's exactly what they're intending, since that's how Maslow's Hierarchy is supposed to work. But again, what makes "fine arts" different from popular media in this regard? One could argue that fine arts are more insulated from when the "going gets tough" than popular media, because fine arts tend to be perused by the wealthy, who are less likely to be affected by economic downturns. The fact that society is stratified like this is of course, an additional layer of complexity relative to Maslow's that further challenges the meaningfulness of such a hierarchy; different strata of society have different levels of resources to devote to discretionary spending and different preferences, and are not equally impacted by economic adversity.


To be clear, EROISOC is a relatively meaningful indicator of QoL, and it being so compared in the paper, but the big fudge is actually how well this EROISOC corresponds to physical EROI, which in depends how tightly EIR and EROI couple. In other words, the chain depends on the assumption that energy intensity ratio and EROI are proxies for each other, and for oil (and coal) there is at least some independent research to somewhat support that, unlike for entirely bread-powered societies. It is, however, very hampered by the fact that EROI studies do not themselves have standardised methodology and sometimes use economic scaling to estimate unknowns.

Good point, but I think again the authors want to make a link to a physical indicator rather than an economic one and so assume the above as a proxy. It's notable that they don't do anything like 'renewable EROI is below such-and-such number, so it will never work', or anything even vaguely of the sort.
Right, but I'm not comfortable with trying to present a physical indicator that is actually a proxy for an economic indicator. It seems like it adds extraneous complexity and additional assumptions compared to simply using the economic indicator, and I don't see what additional insight is granted by using it. Feel free to enlighten me on this if you do.
 
Only when people have the time and money to do so.

They always do. That's what makes it popular.

The number of societies in which there has been no popular culture is zero. The amount of time and money needed to create Clementine was negligible. Not so for The Ring Cycle.
 
Back
Top Bottom