The Energy Return on Investment Threshold

Hall and Day (2009) report that the EROI for coal might be as high as 80 and that for hydropower, EROI is 40. Does this mean that coal is twice as ‘good’ as hydro? The answer is no, and in this post I will discuss how this relates to the idea of an EROI Threshold.



The Net Energy Cliff

This post is based on a presentation that I gave at the recent ASPO conference on November 4th, 2011.

We must first realize that EROI is a somewhat theoretical concept; it is a unitless ratio that does not describe actual flows of energy. What society really cares about, and what is really used to grow economies around the world, are actual flows of energy. More precisely, the economy utilizes flows of net energy. What, if anything, can EROI tell us about the flow of net energy?

To understand how EROI influences the flow of net energy, we must first look at the equation for both net energy and EROI, which are:

Net Energy = Eout – Ein
EROI = Eout/Ein

If we solve the EROI equation for Ein and substitute it into the Net Energy equation, we get:

Net Energy = Eout*((EROI-1)/EROI)

From this equation Mearns (2008) created the “Net Energy Cliff” graph. The net energy cliff figure relates the percent of energy delivered as net energy (y-axis, dark grey) and the percent of energy used to procure energy (y-axis, light grey) as a function of EROI (x-axis).



The Net Energy Cliff

The exponential relation between net energy and EROI creates what I am calling an EROI Threshold at roughly 8. Due to the asymptotic nature of the curve at high EROIs, there is little difference in the actual flow of net energy delivered from technologies that have EROIs above 8. The corollary is that extraction/conversion processes with EROIs below 8 result in vastly different flows of net energy.

For example, a drop in the EROI of oil extraction from 50 to 10 would result in a change in net energy flow from 98% (of the gross energy flow) to 90%. Yet, a drop in EROI from 10 to 2 would result in a net energy change from 90% to 50% of the gross energy flow.

This means that the relevance of EROI as a meaningful comparison of extraction/conversion technologies decreases as EROI increases. This is also the reason why I stated in the beginning of the article that coal, with an EROI of 80, is not twice as good as hydro, with an EROI of 40, because the actual difference in the flow of net energy between these two is very small. The truth is that they both deliver well over 90% net energy. What this threshold effect means is that, when substituting renewables for fossil fuels, it is less important to match EROIs (i.e. substituting coal for a renewable that also has an EROI of 80), and more important to focus simply on avoiding very low EROI technologies (EROI < 8).

Major Caveat to the EROI Threshold

There is one major caveat to this discussion. The logic behind the EROI Threshold only applies if the EROIs being compared are actually commensurable: i.e. that the EROI analyses utilize the same set of assumptions. This is often, however, not the case.

One significant difference between the EROIs calculated for fossil fuels and that for renewable technologies results from the intermittent nature of renewable energy. It is commonly thought that scaling renewable energy will require the adoption of some sort of storage system to account for times of over- and under-production. The EROI of wind or solar PV will surely decrease if we allocate the energy costs of those storage systems to the solar PV or wind conversion process. The question is whether this added energy cost will decrease the EROI of these systems below the EROI threshold, but to my knowledge, there are no peer-reviewed papers reporting EROI numbers that included these costs.

Bottom-Line
EROI is a useful metric for comparing across energy extraction/conversion technologies, or for comparing the extraction/conversion process of one resource over time. But as EROI increases, and especially as it increases much beyond 8, its relevance, as it pertains to net energy flows, fades. Furthermore, due to the aggregated nature of the EROI statistic, every analysis involves assumptions. It is important that those who use these EROI statistics understand what those assumptions are and what they indicate about the utility of the EROI statistic produced.

References

1.Hall, C.A.S.; Day, J.W., Revisiting the limits to growth after peak oil. American Scientist 2009, 97, 230-237.

2.Mearns, E. In The global energy crises and its role in the pending collapse of the global economy, Royal Society of Chemists, Aberdeen, Scotland, October 29th, 2008; Aberdeen, Scotland, 2008.

David, your value for hydro seams low. Gagnon got much higher values (110-220). Also, I would like to pint out that all those values are calculated around a system boundary. Value calculated by EOI-LCA tend to be much lower.

The hydro values do seem low when you consider some of the run-of-the-river hydro projects (no dam). There is very little energy input required to build these projects, and the energy output can be enormous.

I saw one in Bhutan that was 1000 megawatts, and they had room for another three like it downstream of the first. Of course, Bhutan is a semi-mythical mountain kingdom high in the Himalayas, a modern version of Shangri-La, so it has enough vertical drop in its rivers to make this work. Most places don't have similar mountains.

The EROI of the Niagara Falls hydro stations must also be enormous. Most people don't know there are hydro stations at Niagara Falls (producing 4400 megawatts of power), nor do they realize that they cut back the flow of the falls when the tourists aren't there. Between 50 and 75% of the flow of the Niagara River goes through the generating stations.

very interesting , thank you , anything below 8 ain't worth doing basically , a though came to mind that Medieval period Feudal system had the peasants giving 7/8 ths of their output to the lord , does this represent a EROI of 7 perhaps? I thinking of agriculture systems compared with our FF one

Forbin

That would be an EROEI of 8. If you consider no population growth, of course.

It is general admited that agrarian society have an EROI between 5 and 10. Note that part of the takes were used to build and maintain energy production/plundering system. Mill were often under the control of the landlord, while army is a way to extract resources with high embodied energy.

It is general admited that agrarian society have an EROI between 5 and 10.

Do you have any sources for that?

If I recall correctly, during pioneer days in North America it required about 25% of the land to feed the horses needed to plow it. That would give an EROI of 4, not 8. So I think the guideline of 8 as a minimum is unrealistic. An EROI of 4 is perfectly acceptable if that is the best choice you have. It did require a lot of cheap land, though.

OTOH, I think the Roman system of using slaves was considerably worse. I doubt the EROI of slave-based farming was higher than 2, which may be one of the reasons for the collapse of the Roman Empire.

The barbarians had a much more efficient system of pastoral agriculture, although it only worked at a much lower population (easily achieved by killing a lot of the Romans).

Tainter cover this aspect very well. From my memory, he has calculated an EROI of 10 from the agricultural process, much more from conquest.

I must also point out that one must distinguish EROI from ERR. The grain used by the horse dont matter as much as the grain used by the farmer. Horse are like tar sand, a lot of tar sand is use to extract tar sand but from the outside this does not matter much.

In primitive agriculture, the only energy inputs and outputs were food energy. So, in primitive subsistence agriculture, the inputs and outputs are equal, and the EROEI was not much more than 1. Being a subsistence farmer in ancient times was not a lot of fun, his main objective was to live until he died, which was not a long time.

Civilization stepped up this rate, until in the late Middle Ages, with the introduction of metal plows and horses the EROEI may have reached 4. They still burned up 25% of their food to feed the horses, but it was a lot better than burning up 100% of their food to grow enough food to work hard enough to feed themselves.

And then the industrial revolution hit, and with the burning of coal for fuel and the use of machines for farming and manufacturing, the EROEI skyrocketed. The introduction of oil for fuel pushed it even higher.

But I don't think an EROEI of 10 for ancient agriculture can be justified. I think it was a lot lower than that.

The way you comes to ten is to calculate how much calorie you get from each calorie spent. It must be higher than one because, you need energy for the elderly and children, who cant feed themselves. You need spare calorie to build you house and make your clothe. Overall, you don't save much however. EROI is not use to calculate the net energy of the economy, just the energy flow in the energy production.

anything below 8 ain't worth doing basically

I don't think this is David's point. His point is that with EROEI above 8 ordinary people don't notice the difference that much. Below 8, energy becomes something everyone starts paying attention to.

Medieval period Feudal system had the peasants giving 7/8 ths of their output to the lord,

Do you mean 7/8ths or 1/8th. If they gave 7/8ths to the lord that would leave them with only 1/8 of their crop to live on for one year. They would have starved.

Anyway you cannot figure EROI from just knowing what they gave away or kept. You must know their original investment. And since that was mostly labor, that would be rather difficult to calculate.

Ron P.

Anyway you cannot figure EROI from just knowing what they gave away or kept.

Actually you can. One can simply stipulate that the food surplus is the net energy, since we know that medieval peasants pretty much worked their butts off and didn't enjoy any luxuries. Thus 7/8 of the crop was what they (and their animals) needed to survive and produce next years' crop, and 1/8 was the net energy. Stuart Staniford has taken this logic further and argued that the EROEI of medieval society was between 1.1 and 1.6.

Now, if large amounts of other energy besides food was used - firewood, for example - that might throw the result off by some amount.

Wait a second! -- You mean tar sands might not be a good idea?!

...But, through the lens of the the coming economic collapse: There's also a complexity component to future US energy sources. i.e. Once the economy simplifies & localizes (via collapse), it's only energy sources that have relatively simple extraction techniques & infrastructure requirements, and that don't require complex supply lines that will be viable at all. (ex: mostly contemporary solar stuff -- firewood, grains & veggies, passive solar, draft animal, solar thermal, etc.)

That these will necessarily supply a significantly lower energy flow rate will not bother the architect(s) of the Thermodynamic Laws one bit. ...And the still-living earth is totally psyched about it.

Roman empire has collapses when its EROI dropped following the conquest of all valuable land. There EROI ended to about 10 following the collapse.

I fear that industrialization may well be a one-time event in human history because of the energy investment needed to harness the residual, more complex fossil fuel sources post-collapse. If we screw up now I think re-industrialization would be very difficult to achieve.

Provided the biosphere remains habitable, after several million years without significant extraction of them, fossil fuels would accumulate again. Geologic processes would slowly create more mineral deposits.

Yes, except that the timescale for that would be more on the scale of billions of years, not millions.

This means that the relevance of EROI as a meaningful comparison of extraction/conversion technologies decreases as EROI increases.

Would this also mean that it is important to take into account Odum's "transformity" (eg a BTU of Coal is not the same as a BTU of electricity) ratios at lower EROI's?

Yes, this is something important too. The EROI presented here are all calculated in primary energy. In the case of electricity to get a 3x factor due to the conversion of BTU to kWh. This improve the EROI of PV compare to it pure energy production. Technology providing EROI lower than 10 are problematic.

I believe Odum has actually put the ratio at 4 to 1 when comparing primary FF energy to electrical energy

Might be, it depend of the technology used.

I believe the extra unit was added to account for the energy used to build the thermal plant which has a conversion efficiency of about one third or so.

It also depends on what the electricity is used for. If it is for heating the transformity may be a less. For process work such as driving a motor, it is probably on the high side... so, it depends.

If it is for heating the transformity may be a less.

Could be a 4, with a heat pump.

Would this also mean that it is important to take into account Odum's "transformity" (eg a BTU of Coal is not the same as a BTU of electricity) ratios at lower EROI's?

Not really. Whether consumers use coal for the heat directly, or consume electricity generated by coal, does not affect the amount of coal (equivalent) that is used to get more coal. That is what principally affects the threshold.

What about with substitutions for coal fired generation such as Solar PV or Wind generated electricity. As we deplete our coal reserves and the EROI's are lower these other technologies might be substituted even more than they are now.

Not quite sure what you're trying to get at.

It's true that the things you're referring to affect EROEI calculations and definitions, in various fashions. But I don't see how it affects David's overall point that above a certain EROEI, comparisons become less significant. His point is that below an EROEI of about 8, everything that affects EROEI becomes more important.

The point is:

input of 1 units of coal and output of 10 units = E-ROI of 10.

input of 1 units of coal and output of 3.3 units of electricity = E-ROI of 10.

How? Because 3.3 units of electricity in an electric motor will dig up as much coal as 10 units of coal in a steam engine.

The point is:

All of these things indeed go into an accurate understanding of the EROEI of given technologies, but David's point in the keypost is a meta-analysis that operates above all that.

We know there is a lot of work going on today to compute the EROI's of various energy supplies such as coal, oil, natural gas, wind, solar PV, etc. David made a good point that the really critical area of this analysis is when the EROI is less than 8 to 1 because the net energy is much much less. Well, this is the same area that some of the renewable projects fall into and renewables often make electricity. So, when we are comparing low EROI renewables against the high EROI fossil fuels (that need to be transformed into electricity to do work) it makes sense to include the transformity in the analysis. For example, one could mistakenly write off some of the low EROI renewable projects as not being economic if the transformity is not considered.

That depends of the actual usage for electricity. If you waste it for direct heating, you'll need the same amount of energy then, lets say, fuel oil.

And coal, no matter how high the eroi is, might not work as well as gasoline/diesel/etc for truck and heavy machinery, which is needed in the extraction of oil and gas.

The total energy input include lots of energy type: oil, gas, electricity and others. In the case of oil or gas extraction, they can be conveniently factorised as oil or gas, since theses energy sources can be of use everywhere, which gives a perfect, i.e. accurate, EROI involving a single energy source.

I would like to see an eroi of the type, lets say:

EROI = coal output / (oil input + gas input + coal input)

And, we should replace the oil inputs and gas by their respective EROI factor (namely (1 + 1/(EROI-1)) * energy input). That is, if you have an oil EROI of 2, then you don't need 10 joules of oil input, but 20, since you will use 10 to extract 20 and then use that oil to extract coal.

That give us an interesting figure where some source a considered to have an high eroi, but since they relies on a generous eroi by other energy sources, their own eroi is good. However, give a worst eroi for the energy extraction of the inputs, and you end up with a bad eroi for theses.

Exemple:

Coal extraction EROI with oil and electricity = 10 TJ of coal output / (0.2 TJ oil + 0.02 TJ electricity)
Oil extraction EROI with oil = 40

Oil factor = 1 + 1/(40-1) =~ 1.025 = fo
Electricity conversion = ( 3 TJ * (1 + 1/(fc-1)) + 0.05 TJ * fo) / 1 TJ = fe

Coal extraction EROI with oil and electricity = 10 TJ / (0.205 TJ + 0.02 TJ * fe)) = fc
fc = 37.3406

So the EROI for coal would be, in energy alone, 45, but since electricity is much more valuable, we get 37.3406 assuming electricity and oil inputs. Everything is factorised as coal or self-sufficient energy sources (using their own EROI).

With changing oil EROI, labeled as x and using the latter assumptions, the Coal EROI would be:

We can see that coal extraction isn't that much dependent on oil EROI until it reaches a value lower then 10 or 20. Now, to make steel and heavy equipment we would need anthracite and others.

In conclusion, I feel the method, given the right values, is much more reliable as EROI because the result is a closed system, like on a drilling platform where you use gas to extract gas. Only here you need a coal mine, an oil well and a coal power plant to make the system closed.

Coal powered steam engines would work quite well in big trucks.

Much coal mining can only be done by electrical equipment, for safety reasons. Much mining, especially underground, has been electric for some time - here's a source of electrical mining equipment. Caterpillar manufactures 200-ton and above mining trucks with both drives. Caterpillar will produce mining trucks for every application—uphill, downhill, flat or extreme conditions — with electric as well as mechanical drive. Here's an electric earth moving truck. Here's an electric mobile strip mining machine, the largest tracked vehicle in the world at 13,500 tons.

to make steel and heavy equipment we would need anthracite and others.

Electricity can suffice for iron reduction/smelting.

So, when we are comparing low EROI renewables against the high EROI fossil fuels (that need to be transformed into electricity to do work) it makes sense to include the transformity in the analysis.

In principle yes. But transformity doesn't necessarily lower the EROEI of fossil fuels (see Nick's post).

For example, one could mistakenly write off some of the low EROI renewable projects as not being economic if the transformity is not considered.

To repeat, I think it is the wrong interpretation of David's 'threshold' idea to think that EROEI sources below 8 are not economic. That's not his point.

I agree with your first and second points. However, I also believe that some of the lower EROI renewable projects may be viewed as un-sustainable because of their lower EROI's if transformity is not considered. This may be the right answer in some cases but I wanted to point out that we need to make sure we do consider transformity, where applicable, and especially in the low EROI cases in order to get a truer picture of what is going on.

However, I also believe that some of the lower EROI renewable projects may be viewed as un-sustainable because of their lower EROI's if transformity is not considered.

I think that renewables' economic competitiveness with fossil fuels might be underestimated if transformity is not taken into account. Sustainability is another question, which David is not addressing in the keypost.

In any case, we are not really disagreeing on much of anything here.

When looking at storage and other 'backup' for renewables it is important to consider the also concepts outlined in this report from members of the IEEE Power and Energy Society on "Wind Power Myths Debunked" as there are other ways to at least partially address the storage issue using for example 'energy balancing' between geographic regions. http://www.ohiowind.org/PDFs/Wind%20Power%20Myths%20Debunked.pdf

I assume that the ow EROEI for tar sands is because large amounts of natural gas is currently used in the extraction and refining process. A lot of this NG is stranded from other markets, so I think a more representative measure would be the EROEI of the combined tar sands/NG operation, which would be somewhere between the figure for the two technologies separately .

Also, I wonder if we could estimate the minimum EROEI for industrial society overall that would be needed to sustain a given level of industrial technology. For example, I think we are already beyond the point where manned interplanetary travel is likely to occur, we simple do not have enough spare energy and never will have. At some point technologies will become too capital intensive to develop further - how many more generations of Moore's Law development of semiconductor technology can we sustain before the potential market return falls below the cost of development? Will we ever build a commercial fusion reactor even if we have the technology today to sustain a net energy positive fusion reactor core? If a solar storm took out half the GPS satellites would we have the funds to replace them? In the longer run, could we afford to replace processor fabrication plants?

Large scale renewable energy systems and adaptable power grids are as dependant on electronic control and communications as conventional power stations. In the long run will we need to employ third world level technology because we cannot sustain anything else?

Power for controls is actually very small when compared to the energy it is controlling.

It is important that we remember the non physics and non engineering sides of this issue.

There are always bigger envelopes, at least until you get to Greenish's "alien biologist " pov.

I fully recognize the implications of the fast drop off of useable and useful net energy at an eroi of around 8 or 10 after seeing this graph here for the first time-up until now, I didn't have any good idea just how high eroi must be to achieve stability if not sustainability of an industrial economy.

My hat is off to the author and the originators of this fundamental work.

The non engineering and non physics side-the human side-of this issue may play a far more important part in the short to medium term than expected.

A million tons of concrete and steel can be used to build a mega mall with high rise office and residential towers, or it can be used to build a hydro electric power station.

The first option will consume energy at a prodigous rate for either the life of the infrastructure, maybe fifty to a hundred years, or until the collapse of bau renders it worthless.

The second option will produce prodigious amounts of power times ten or a hundred or a thousand for at least a hundred years or so-and imo, for a lot longer with maintainence, at least at a reduced level of output.

At the personal level, we can either "invest" energy in a beer run in our oversized suvs , or we can invest energy in insulation and a a solar domestic hot water system. At the society level, we can invest in wind, solar, hydro, and maybe some biofuels-or football staduims and jumbo jets to fly road warriors and tourists around..

The thing to keep in mind is that in either case, the energy will be spent.It will not be conserved to any serious extent;we are naked apes, not Vulcans.

Arguments about the fundamental eroi of hydro or solar pv pale into insignificance when this is considered from a practical pov for the short to medium time frame.

Of course over the long term, eroi will rule as firmly and as harshly as the fundamental laws of physics.

In a couple more decades, basic engineering texts will probably have chapters devoted to the concept.

I think Ralph's point was more that the energy required to run an economy capable of producing those controls (the only kown example being the economy of the last n years) is huge. So, if we lose the ability to produce them, for whatever reason, we probably will never be able to produce them again,

Peter.

If we are going to have a power grid at all we will need something to control it. We are still using a lot of old electromechanical protective relays as described in the Applied Protective Relaying hand book by Westinghouse from years ago. http://www.amazon.com/Protective-Relaying-Wetinghouse-Electric-Corporati... However, these controls are clunky, use lots of materials to construct and consume a lot of energy. Some of these relays operate all of the time in the 'fail safe' mode so they can consume a lot of electricity. I think the newer electronic controls will be an improvement over these old electromechanical switches, etc. both from an energy and materials consumption standpoint. Adding intelligence such as sensors, advanced communications and coordinated control systems, and computers to our electrical grid infrastructure can substantially improve efficiency and reliability through enhanced situational awareness, reduced outages, and improved response to disturbances. It also enables flexible electricity pricing that will allow consumers to monitor and control their own energy usage and costs.

Yair...dunno..a new neighbour spent three hundred plus bucks on a you-beaut electronic sensor to shut down a little diesel pump when the tank overflows...and then he saw my jam tin on a piece of fishing line that fills up when my tank overflows and shuts off the Honda...it's even got a little hole in the bottom so the water drips out and a spring resets for the next start...effective, low-tech and cost nix.

Controls are getting just too complicated. Every effort should be made to reduce complexity.

Cheers.

You make a good point ..... However, ecological economist, Elinor Ostrom, recommends that we embrace complexity in order to be able to deal with the challenges of this complex world.

I'm not familiar with Ostrum, but it seems like she doesn't understand that decomplexifying should be the objective.

Antoinetta III

She was the first women to recieve the Nobel Prize in Economics (2009) for her work "Governing the Commons" and she also serves on the board of the Stockholm Resilience Center. http://www.stockholmresilience.org/aboutus.4.aeea46911a3127427980003326.... She is well informed about our our energy and environmental challenges.

Vision
The vision of the Stockholm Resilience Centre is a world where social-ecological systems are understood, governed and managed, to enhance human well-being and the capacity to deal with complexity and change, for the sustainable co-evolution of human civilizations with the biosphere.

Mission
The mission of Stockholm Resilience Centre is to advance research for governance and management of social-ecological systems to secure ecosystem services for human wellbeing and resilience for long-term sustainability. We apply and further develop the scientific advancements of this research within practice, policy, and in academic training.
http://www.stockholmresilience.org/aboutus/visionandmission.4.aeea46911a...

Adding intelligence such as sensors, advanced communications and coordinated control systems, and computers to our electrical grid infrastructure can substantially improve efficiency and reliability through enhanced situational awareness, reduced outages, and improved response to disturbances. It also enables flexible electricity pricing that will allow consumers to monitor and control their own energy usage and costs.

I don;t disagree that these things can improve "situational awareness" , but is this really necessary, or beneficial? The grid has been functioning reliably for decades. I would argue that the overall grid has been far more reliable than any computer software yet written.

All outages where I live (BC) are caused by weather events (or scheduled maintenance), and no amount of software, controls or complexity will change that one iota. Putting cables underground will, but at enormous cost that far exceeds the benefit.

When you add al these sensors and things, they have to be maintained, and upgraded more often than the grid they are controlling. Also, the grid operator can then become dependent on those control systems suppliers to "reprogram" stuff when needed. If those people are on the other side of the country, or world, or have gone out of business, then what?

Real world example - the Sky Train in Vancouver is a driverless transit system, that relies on automated controls, sensors, etc etc. But it still has people sitting in a room controlling that - instead of sitting in the drivers cabin. It costs twice as much per passenger mile to operate and maintain this system as Calgary spends on its train drivers. And when the Sky Train needs to change/update something, they have to get the specialist in from S. Korea - how reliable, really, is that?

As for time of day charging, that can and has been done for decades simply with two meters. For residential customers, adding complex systems to enable many rate options, and real time flexible pricing is pointless - you have all this cost so that homeowners can save 87c a day more than just by having peak and off peak. And that is only if they are constantly gaming system - what is the point?

In keeping with the EROEI theme of this post, I suggest we need more 80/20 solutions, complex things like the "smart grid" end up deliver 90% of the benefits, but for 50% of the cost, not 20. It is taking overall EROEI backwards.

The real savings in electricity are to be had by identifying and eliminating customer wastage, and shifting discretionary loads to off peak. We do not need to get real time electricity prices on our smart phones to do that.

Further real-world example - WMATA trains in Washington, DC. That system is highly automated, although since people sit in the drivers' cabs, it doesn't look it. But every now and then they run short on some exotic 1970s relay or other. Then the drivers actually have to drive the trains - with almost the same disastrous effect on schedule-keeping that's a perpetual everyday debacle with buses...

Yair...I know I always bring things back to the basics but some things realy erk me...For some months we have been enduring protracted road work on the main north/south route through Queensland.

Originaly they started off with the normal reversible stop/slow sign...simple cheap effective...we now have a solar powered traffic light unit mounted on a trailer with a remote and the operators are bored jitless sitting in a reclining camp chair and I suspect they go to sleep.

I understand the solar powered set up rents for two fifty bucks a day...this is progress?

You can be sure there is some concern in there about "safety", for the people having to stand by the road, and out in the sun etc.

That said, the traffic lights are sometimes needed when you still have one lane operation at night, and there is no work going on - then they are worth their cost.

Otherwise, its another example of how many employees want someone to spend money to make their job easier, almost redundant, but still keep their job.

That often contributes to declining EROEI

I can't imagine that they're renting the sign. Not unless they only do road projects for 2 weeks a year.

Yair...Nick the electronic signs are rented...as is the track-hoe and grader and the trucks and equipment on the projects.

Cheers

Good god! That's an expensive way to do things.

It is extremely difficult for a construction company to maintain a steady enough flow of work to justify the enormous capital expenditures involved in buying heavy machinery.

So it is common for a company to own only such machines as it can use very frequently and rent the rest.

On a small highway job a dump truck might be used almost every day, but a backhoe might be used only a couple of days out of the month.

So it's cheaper by far to rent one for that two days.

Rental rates are actually quite reasonable at the "wholesale " rates paid by contractors.

Even a one horse operator such as yours truly can get a nearly new hundred thousand dollar machine, delivered and picked up, for five hundred bucks for a day;I can get it for five to seven days for about nine hundred, and for a month for two grand.No expense except the operator, fuel, and grease.

No property tax, no expensive secure storage yard to park the machine, no mechanic on the payroll to look after it, no truck and driver needed to haul it around.

Otoh, I paid twenty grand for a big four wheel drive backhoe (older and well used) and use it only occasionally-four days rent at five hundred a day is two grand and a first approximation of a ten percent return on my twenty.

I only need to run it ten or twelve days a year to justify owning it, so long as I can afford to have the money tied up in it.

The only way to get reasonably close to full time use out of lots of heavy machines is for somebody to rent them out to lots of different users.

I loan my hoe out to trustworthy neighbors in exchange for the loan of equipment I don't own, such as a dump truck, or in exchange for skilled labor.

Yeh, I've found it is often cheaper to do a week's rental rather than two or three days for several items. Likewise month v weeks.

NAOM

That makes sense.

OTOH, the comment said "..For some months we have been enduring protracted road work on the main north/south route through Queensland."

So, that's a big job, for a long time.

"we now have a solar powered traffic light unit mounted on a trailer with a remote... rents for two fifty bucks a day"

That can't make sense.

Paul - I hear what you are saying and you make good points for the BAU case. However, if we are going to integrate a meaningful amount of intermittant renewable energy (widely distributed generation) into the grid, this will make things a lot more complicated (for VAR, power, frequency and voltage control) which will require more complicated monitoring and control capabilities. If we are just going to run our remaining fossil fuels into the ground using conventional large scale thermal plants then electromechanical protective relays will probably handle this just fine.

Can we include large office developments?
That cement is a bit of a waste.
The primary manufacture of the steel though could supply a useful low-energy mine for catabolism at a later stage I suppose?

I always get a chuckle from the logic that the NG may as well be used to process the tar sands because it stranded. The NG would not be stranded if a pipeline was built to connect to the pipeline system further south in Alberta.

I agree.

If they can get the tar oil to the market, they should be able to get the natural gas to the market.

In point of fact, there are pipelines in Northern Alberta to connect the gas fields to markets as far south as Southern California and Mexico. However, you have to realize that it costs energy to move natural gas down a pipeline - and not a small amount of energy. As things stand, it is more efficient to burn the gas in the oil sands plants, which after all are sitting right on top of the gas fields, rather than send it all the way to Southern California or Mexico.

I have a problem with the energy cliff/threshold idea, and Forbin articulates the thesis I don't agree with: "anything below 8 ain't worth doing basically".

Now, if you have an EROI 7 energy multiplying system that comes for free, then you can connect them serially, and get an EROI 49 system (and again). That is, very low EROI systems may be viable if the energy source is plentiful enough, the labor resource needed to create the system is low enough and the environmental impact is low enough.

Please fill us in on the details of such a system-I'm ready to buy in if it can be done.

It's easy to envision one, at least. For instance, a fully robotic wind power construction and deployment system that is (entirely/almost entirely/arbitrarily) self sufficient (can recycle and repair itself and its deployed wind towers) and gives surplus energy.

Nothing in what you've described is 'serially connected.'

You're right, I take that back, if I may. It's a matter of scale, and the cost of that scale.

Wasn't the "fully robotic wind power construction and deployment system" powered by wind energy?

Yes, but electrons don't smell, so it doesn't really matter. If you set up such a black box, you feed it from the grid with 10 units (you do need to provide it at least at startup), and after a while, it produces, say, 50 units to the grid, for a surplus of 40. If you disconnect it from grid input and just take the energy from the system itself isn't a big issue.

Anyway, if you build another such system (of equal size), you couldn't feed it with the 40 surplus and get 200 units to the grid. That's where my "serial connection" may lead thoughts down the wrong path. You need to scale the systems in size or in numbers to get the 200 surplus. It isn't a magic box that just multiplies whatever energy you feed it with.

Given that the wind power can produce oil, steel, and requirement to make robots then yes it is possible. You'll also have to factor lower EROI caused by transformation of electricity into oil/steel. Oil coming from theses will be much less efficient and with lower EROI.

However, the multiplication thing here is simply a question of time. An EROI of 10 in 1 year is better then an EROI of 30 on 10 years**.

**If your maximum energy output is fixed, then your system output is equal to the inverse of the EROI, no matter how much iteration you make. The better EROI become much more important. The system is always equal to the EROI, but if the maximum energy output isn't the limiting factor, you get get to very very high levels of energy output.

Actually, you don't need to make anything up to show that E-ROI isn't the most important thing - there are good historical examples - see my comment below.

David – A very interesting and logical approach. It parallels the economic aspects of oil/NG drilling. As discussed before the oil patch doesn’t consider EROI when making investment decisions…at least not directly. The actual fuel used to drill a well is not only a small fraction of the total costs but also represents a very small percentage of energy even from a marginally successful well. At the upper end it’s totally irrelevant. The difficult aspect of estimating EROI for a well is estimating the embedded energy in the equipment. And the additional problem of amortizing that energy across the number of wells eventually drilled.

In a similar way to your analysis our economic evaluation makes little difference at the upper end. Two exploratory wells may both cost $6 million but one has a target of $30 million net reserves and the other $60 million. Both are quit economic projects but one is not twice as good as the other. First, the $60 million may find $20 million worth or reserves…or none at all. Dry holes happen. This is why I tease exploration geologists about doing detailed economic analysis of the projects: all exploratory wells are great investments because the geologist is free to pump the reserves up since there is a lack of data…that’s why it’s called exploration.

I’m a career development/reservoir geologist. Economic analysis of these types of projects is more similar to your EROI threshold. The reserve targets tend to be much smaller than exploration goals. Thus energy input is more closely approaching energy output. This is why development projects must have a high probability of success: little room for error. And that has always been the balance: exploration is high risk/high potential yield and development is low risk/low potential.

But now we have major plays that fall between these two extremes: the fractured shale plays. Almost every fractured shale exploratory well is completed because it's nearly impossible to tell what it will produce from indications while drilling. When was the last time anyone reported a dry hole in the Eagle Ford or Marcellus? I’ve seen wells with excellent fracture indicators lose money. And wells with little or no indication of commercial quality produce huge profits. Every shale well completed will produce some hydrocarbons…the question is how much and how fast.

The shale plays are a statistical game. Success isn’t based upon what any one well discovered…despite the fact most of the public oils like to report their big wells (and tend to not put out press releases on the less impressive results). My very rough estimate is that some shale plays will, in general, fall not too far above your threshold. Which isn’t that bad for the public oils. Their primary goal is not high profits (and thus not high EROI). It’s the maintain/grow shareholder equity. As discussed many times Wall Street rewards reserve base growth while typically ignoring ROR. I drill only conventional reservoirs and generate 2 to 3 times the return than even the better Eagle Ford wells. But if my company were public Wall Street would have very little interest in us…we don’t care how much of our reserve base we replace. It’s all about $’s in/$’s out. Of course, we would like to increase our reserve base as much as possible...but not at the cost of ROR.

But in the end there is a factor that has a significant effect on the EROI threshold: oil path economics will kill a project long before the projected EROI gets too low. The total non-energy costs of drilling a well are much greater than the energy consumed. And it’s the total costs that determine what gets drilled. Perhaps not coincidentally I recently estimated the minimum limit for a project to pass economic muster would fall close to an EROI of 10. Though it’s difficult to equate ROR and EROI directly to each other, there is obviously a relationship. But that relationship is made even foggier when the shale plays are analyzed. In theory a public oil could break even (make no profit) on every shale well they drilled but as long as they increaed their reserve base y-o-y Wall Street would bid their stock price up. I suspect such an effort would produce a threshold below your EROI of 8.

From the data I got from the scientific literature money return on investment (MRN) is roughtly equal to the EROI divide by 6 to 12. Hence, anything beloy and EROI of 10, is unlikely to make some profit. As you point out, there is a social cost to a technology that is not factored in the classical LCA analysis. I start a research project on this topic but obviously, there is a lot of job to do.

Yvan - Could you put a ittle more meat on that bone? In the oil patch a return of 15-20% is considered very acceptable. Depending on how it's measured historically 10% ROR is the norm. That would equate to an EROI of 60-120. That seems more than a tad high. And a 1% ROR would be an EROI of 6-12. Now that would seem to fit closer to that threshold of 8.

I am not sure that you can equate ROR to EROI with a simple linear transformation. EOR is net income over revenu. Energy wise it would be ROR=(Money_recovered - f_m *Money_invested)/Money_recovered=1- f_m/EROI (f_m=6-12). Hence,

ROR-1=-f_m/EROI

EROI=f_m/(1-ROR)

Hence for a ROR of 15-20%

EROI=1.2 * f_m

But I think there is a temporal factor that is not taken into account. An easier way to do the calculation is that the economic payback time is longer by a factor f_m, compare to the EPT.

Yvan - I agree fully. I thought ou were supporting that 6/12 relationship. Whatever the relationship certainly not linear. Certainly to complex for a geologist to figure out. Temporal: yes...just the other day I wondered is some one could come up with a NPV equivelent for EROI.

From the data I got from the scientific literature money return on investment (MRN) is roughtly equal to the EROI divide by 6 to 12. Hence, anything beloy and EROI of 10, is unlikely to make some profit.

I disagree. The oilsands, as RMG has frequently pointed out, have an operation EROEI of about 6:1. Add in external embedded energy and this would be lower still.
BUT at current oil prices, these operations are *highly* profitable, the factor EROI to MRN would need to be about 3 to reflect this reality.

This is, quite simply, because different energy sources, in different places (or even different times of day or year) can have *vastly* different monetary values. Oil is currently worth 6x more per BTU than NG, so turning one unit of NG into six of oil will have vastly different returns than turning oil into 6x oil.

Any paper that puts an arbitrary ratio of EROI to MRN is ignoring this fact.

Perhaps the people that wrote that scientific literature ought to have to decide which energy producers they will invest their retirement savings in, and then see if they think that arbitary ratio holds true.

Economics is not a science, but neither is business - the scientists need to recognise this.

I said the 6-12 is an average that I found in one of my research sbout energy efficiency, and is it close that to the average economy EROI. You are right the ratio is lower for the oil industry. Unless you have both energy and money flow, comparison is impossible to do in a strict way. Trust me this is a rare occurrence in the scientific literature. And your are right, you can cheat using NG to leverage.

As for the tar sand, the EROI (~3) is low but the EER (External energy ratio) is around 10 and it is mostly natural gaz, which makes them profitable.

Rockman,

Very intriguing comments. Can you email me this "I recently estimated the minimum limit for a project to pass economic muster would fall close to an EROI of 10."

My email is theoildrumeroi at gmail dot com

David, I have a bit of a quibble with your report, one I've offered before on TOD. I think that the use of the abbreviation "EROI" is seriously confusing. While you have clearly defined what you mean, so that we here on TOD understand the meaning, outside this rather small group, the market oriented world will assume that this means "Energy Return on Investment", as in, dollars invested. I think that the better abbreviation would be "EROEI", as we really need to be passing this message beyond the choir, so to speak. Among the comments already posted this morning, there appears to be a mixing of units, with dollar cost and energy cost being conflated. If the message is garbled at the onset, it will be even harder to deliver it to the world beyond those who already understand...

E. Swanson

Yep, EROI is seriously confusing since the default standard-english meaning of "energy return on investment" is something else entirely, and is a commonly used phrase. Worse, it is a phrase which is often use in the same discussions. It's as though in an alternate world both spaghetti and meatballs are called spaghetti. One comes from ground-up cow parts, one is pasta, but we call them both spaghetti because the Pope does. This causes no end of foulups in restaurants, on shopping lists, and in manufacturing.

Continuing to use EROI because previous academic papers did makes little sense, unless the intent is to prevent the broader public from ever understanding it. EROEI, ER/EI, or even "net energy" is at least not inherently confusing. I beat that drum for a couple years here, but even the academics bold enough to post on TOD will seemingly stick with EROI until it's pried from their cold dead hands.

I've never seen a discussion go very far before the acronym EROI causes confusion. Because it's confusing.

Until the marklar decide to marklar the marklar, marklar will marklar the same marklar.

EROEI, ER/EI, or even "net energy" is at least not inherently confusing.

'Net energy' is not the same as ER/EI, as David made clear in the keypost.

Perhaps the subject is inherently difficult for most people no matter what. Some time ago I removed 'emergy' from the Wikipedia page for EROEI, where it was offered as a synonym.

'Net energy' is not the same as ER/EI, as David made clear in the keypost.

Yes, I was talking about clarity of concepts.... apparently not clearly enough. ER/EI is a dimensionless ratio while "net energy" is expressed in energy units. More importantly, the ratio conceptually establishes the nature of an absolute limit we and all life are up against. It's a deep realization most people never have; that the same limits which apply to the energy expenditure of polar bear foraging also apply to human civilization.

That's an important thing... which makes it a shame to use an inherently-confounding acronym and phrase to refer to it. Thus, one winds up talking about "net energy" instead, which is a similar concept and doesn't suffer from the terminology problem.

Among other things, I've been an activist who has introduced new concepts and terms into contemporary human culture, and it's an easy thing to do wrong... hence my heckling for more clarity.

Until the marklar decide to marklar the marklar, marklar will marklar the same marklar.

I think that's just a THEORY... >;^)

A question. As a novice to The Oil Drum, please forgive me if the question has been answered previously.

Whenever I read about energy returned on energy invested, I think of an ordinary alkaline battery, such as might power a torch or a clock or whatever.

The old received information was that it took 50 times the energy to make that battery than the user gets out of it. Is that an EROEI of 1:50? Or is it properly shown as -50:1?

It seems to me that, to most people, EROEI is irrelevant; what is important is the energy at the point of use, no matter what it takes to get to that point (plus of course the final cost). So it only matters once we pass peak energy, and maybe not then to some!

David

A battery isn't a source of energy, it's a way to store energy provided from some other source. Our basic problem is that humanity discovered ways to recover and use the energy stored in fossil fuels. Fossil fuels are like your battery example, the original energy is provided by the sun and the storage is the result of geological processes. Like your battery, once the fossil fuels have been consumed, there won't be any easy way to replenish that energy source or use the devices which have been optimized for each fossil fuel type. The cost in money is just a comparison between your income and the amount of energy available for you to use. At some point in time, no matter how much green paper the average man has in his pocket, he won't be able to buy enough of those fossil fuels to power a car or heat a big house...

E. Swanson

"A battery isn't a source of energy, it's a way to store energy provided from some other source."

With respect, I understand that; but why does that affect my saying that 50 units of energy are required to produce 1, in other words, EROEI 1:50?

David

It seems to me that, to most people, EROEI is irrelevant; what is important is the energy at the point of use, no matter what it takes to get to that point (plus of course the final cost).

Right on. I have attacked the EROEI concept many, many times since I started commenting on TOD. All to little or no avail.

EROEI has some validity if the form of energy in is the same as the form of energy out. This is true of oil wells mostly.

But when the form of energy in is different from the form of energy out there is often a gain in utility that compensates for very low or even negative EROEI numbers as in the case of batteries that you cite.

This concept seems to be very hard for EROEI believers to grasp.

When different forms of energy are compared as in the chart shown, energy is taken to be concrete and measurable. It is not. All energy is not the same. Each form has its own utility which can in many cases make up for the loss of energy inputs of one form to produce the newer form.

Energy is an abstract noun. It is not concrete. As such, charts showing comparisons of different forms of energy are inherently fallacious. This fallacy is called reification:

From Wikipedia:

Reification (also known as concretism, or the fallacy of misplaced concreteness) is a fallacy of ambiguity, when an abstraction (abstract belief or hypothetical construct) is treated as if it were a concrete, real event, or physical entity.[1] In other words, it is the error of treating as a "real thing" something which is not a real thing, but merely an idea. For example: if the phrase "fighting for justice" is taken literally, justice would be reified.

Another common manifestation is the confusion of a model with reality. Mathematical or simulation models may help understand a system or situation but real life always differs from the model.

Note that reification is generally accepted in literature and other forms of discourse where reified abstractions are understood to be intended metaphorically,[1] but the use of reification in logical arguments is usually regarded as a fallacy. For example, "Justice is blind; the blind cannot read printed laws; therefore, to print laws cannot serve justice." In rhetoric, it may be sometimes difficult to determine if reification was used correctly or incorrectly.

I have attacked the EROEI concept many, many times since I started commenting on TOD. All to little or no avail.

Well, let's try it this way: you may have a valid point in there somewhere - but one "definition" of insanity is repeating the same behavior over and over and expecting different results.

It seems as though energy, on any reasonable definition, is real enough even if different forms do indeed have different utility under different circumstances. It is certainly real enough to be precisely measurable, real enough for the societal "we" to measure precisely the losses when we convert it from one form to another, and even real enough to kill us outright if we get in its way under the right (i.e. wrong) circumstances.

Anyway, I intuitively suspect that buried in there somewhere is a valid point, but I'm at something of a loss to tease it out for sure. Maybe a paraphrase omitting the notion of "reification" might help, I dunno...

you may have a valid point in there somewhere

No, he doesn't really. His argument boils down to saying that because these issues are complex, we can't know anything at all about them. By that kind of logic, no one knows anything about anything. It's pure sophistry, especially the part about reification.

X has a valid point.

I don't have any trouble understanding it;his argument is based on the simple assumption that so long as some forms of enerGy are plentiful and therefore cheap, and it is economically feasible to convert energy from such a form to another that is scarce and therefore more expensive-AND also worth the extra expense-due to increased utility-then EROEI doesn't matter.

And as a matter of fact , it doesn't-from the pov of a businessman.

I must say that it also doesn't matter much in any respect, for now, and for some considerable time into the future, excepting the theoretical respect.This is because all the decisions that are being made, and are going to be made for some time, are going to be made on the basis of cost and utility;the overall situation is analogous to the INVESTMENT DECISION PROCESS in the oil and gas drilling business which Rockman so capably describes.

Now at some future time, it MAY be that enlightened leaders decide that oil for instance must be conserved for use as chemical feedstocks, rather than burned as fuel, because it will require more energy to manufacture these from other raw materials such as organic wastes.

But as a practical matter, if energy is wasted by so burning coal to manufacture lubricants, nobody will give a hoot-excepting by oddballs who hang out on websites such as this one.(My friends sometimes say I am so odd I am weird-way beyond simply eccentric.)

EROEI will become critically important from a practical pov at the same time as energy from oil, coal, and ng becomes scarce and unaffordable-scarce and unaffordable to the extent that we must run our civilization on renewables for the most part.

If I understand this article correctly, we are going to be in very deep doo doo at THAT POINT in time if renewables can't supply copious energy at an EROEI return of about eight or better.(I personally doubt if renewables are going to be up to the job.)

X is a business oriented guy.
He is not concerned with theoretical matters that have no immediate practical applications or implications.
His pov is similar to that of the people who invent or support ponzi schemes such as social security.The less mathematically literate ones may actually believe such systems are sustainable.The smarter ones, privately, admit that they expect to be properly composted before the bill comes due.

I have less than zero use for corn ethanol, except mixed with spring water in moderate quantities.It is a really desperately dumb way to extend gasoline supplies, for many reasons other than a scanty return on energy invested for energy out.

But again, looking at this from x's pov, I can't really be sure what the EROEI of corn ethanol is-or might be, at some future time, when industrial waste heat might be used to carry out the distillation, etc.

Assigning an energy value to the high protein livestock feed left over is a very tricky business in and of itself-the answers found will depend on numerous assumptions involving the embedded energy in other fungible feedstuffs, not to mention the basic question of whether we should even be eating animals raised in confinement

The weight of the actual PRACTICAL evidence is on his side-what we are actually DOING follows his reasoning precisely.

The theoretical evidence is on everybody elses side.

We will eventually have to face up to that theoretical evidence-but not for a while yet.

Bau will rule-until bau collapses.

X may be as wrong as wrong can get, but hos type are in control of the levers.

OFM,

I made a previous reply to your post, which for some reason disappeared, perhaps because I was not nice enough to X. This reply is not the same; the other one was better. This is the best I can do instead.

And as a matter of fact , it doesn't-from the pov of a businessman.

Sure it does. First of all, if the business man is a venture capitalist looking at the energy industry, he ought to be really interested in EROEI. It will help tell him which technologies are not worth investing in, if the net energy is negative and there is no other advantage. Secondly, studying EROEI is just another piece of information that may tell you things about the direction of costs and prices in your energy related industry. Is it the only thing to consider? No one ever said it was. Is it stupid or a waste of time to consider it? I don't think so. Changes in EROEI will affect industry in all sorts of ways, and its useful for any businessman to have a clue about what those might be.

But again, looking at this from x's pov, I can't really be sure what the EROEI of corn ethanol is-or might be, at some future time, when industrial waste heat might be used to carry out the distillation, etc.

No, you can't be sure, but you can make an educated guess using science. Or at least you can listen to those who make those estimates. And that might lead to better decisions if you happen to be considering whether to stay in the the ethanol business or not. If X does not realize that, I think that's unfortunate for him.

Well stated.

If EROEI was the only concern, then no one would bother with electricity (EROEI no better than 0.4), and yet we do. The concept is a boundary case, and is of academic interest. As a practical matter, it doesn't matter. Something else will shut you down first if you are Rockman, or you will do it anyway if you can convert something readily available (coal, wind, sunlight) to a more useful form (diesel, electricity).

These EROI work because original EROI is high enough from start. If the original EROI was low, the conversion factor will bring the price of the final product to uneconomic model.

As a practical matter, it doesn't matter.

As a practical matter, it is a factor (albiet not the only one) that determines your business success or failure. It certainly does matter.

Something else will shut you down first if you are Rockman, or you will do it anyway if you can convert something readily available (coal, wind, sunlight) to a more useful form (diesel, electricity)

If you understand why EROEI matters, you might realize you have a better chance of taking your capital and moving to a new business before you get shut down. If you do not understand why EROEI matters, it will just happen to you and you will be unprepared mentally and probably financially.

I have some difficulty understanding why anyone who posts on this site dedicated to 'energy and our future' does not think that EROEI has practical implications.

Eroei simply DOES NOT HAVE any practical implications from a businessman's pov when considered in a time frame used to make business decisions on an ongoing basis.

Businessmen compute profits and losses and returns on investments and calculate or estimate their risks in MONETARY terms.

Now of course they do consider whatever estimates are available respecting the future costs of various sources of energy-IN MONEY, not in ergs or calories or kilowatt hours-and the MONEY cost is adjusted up or down according to what they estimate the energy will be worth depending on its FORM.

So a businessman may estimate that ten years from now that a gallon of gasoline will be worth eight bucks before taxes and that coal containing an equal amount of energy is worth only a buck and a half.

He will not give a hoot whether energy is wasted in converting coal to gasoline-so long as it looks like there is money to be made doing it.

OF COURSE he will consider the possibility that the PRICE of coal and the OTHER INPUTS NEEDED manufacture synthetic gasolinemight increase to the point that the investment in a coal to liquids plant might not pay off.

But all these considerations will be made on the basis of prices and expenses in money rather than energy.

I have an older Ford F150 4x4 that I have set aside hoping to convert it to a woodburner someday when time permits.My research indicates that 10 kilos of good dry oak chips will take me about as far as a gallon of gasoline if I ever get her running on wood.

Anybody who wants some good dry oak can buy it from me for a hundred dollars a ton cash underground economy price picked up at my place cut to stove length.Chipping will be another forty bucks.

That 240 dollars worth of wood chips(which would probably cost twice as much at retail in the above ground economy) will take me about as far as a hundred gallons of gasoline, which will cost me about 350 bucks.

But considering how much trouble it is going to be to drive that Ford burning wood, I would much rather pay TWICE or even three times as much for gasoline, so long as I can get it and have the money to pay for it.

I could care less about the amount of energy embedded in the oak as compared to the gasoline;the gasoline has far far greater more utility as ice fuel-enough greater utility that I would gladly pay two or three or even more times as much for the MOTOR FUEL ENERGY EQUIVALENT amount of gasoline.

(Just starting up a wood burner takes fifteen minutes, and it must be serviced every five hundred miles or so; and the engine , which should last another hundred thousand miles on gasoline will probably not last more than twenty thousand burning wood.Refueling will probably take at least fifteen minutes.Half of the cargo box will be taken up by the wood gasifier and a chip supply adequate for a hundred mile trip.)

Eroei will become important -critically important-someday.It is a very useful metric, an invaluable metric, in computing the outer bounds of what is sustainable.

But businessmen are interested in profit and loss, not sustainability.

A coal to liquid plant might be built with a planned life span of thirty five years.

A business man will not likely look any farther ahead than that.Liquid fuel on an equivalent energy basis will probably be worth a LOT MORE thirty five years from now than coal.The decision to build or not to build will be based on relative prices, not energy content.

Businessmen compute profits and losses and returns on investments and calculate or estimate their risks in MONETARY terms.

And how do they assess their likely monetary returns on investments that will take many years to pay themselves back? By considering the many real world factors that might affect their cash flow. EROEI is a real world factor that many would do well to put in their stack of considerations.

But businessmen are interested in profit and loss, not sustainability.

That depends if they are the type of businessmen who want to make a profit only this quarter, or for the rest of their lives. There is nothing inherently anti-sustainability about business. It is a moral choice for a businessperson to completely ignore sustainability. If businessmen had no obligations to consider sustainability, then we could say that there is nothing wrong with Ponzi schemes. But of course, there is something wrong with them.

Did you miss my response to your post upthread? I don't get the impression that you really thought about anything I said in that post, and if this last post was meant as a response, I don't think you made any new points in it.

Sorry JB

I see that we are talking past each other.

My contention is that business men DO look as far as possible into the future, in estimating costs and returns-certainly any company building an electric power plant must come to the conclusion that they will be able to operate it for quite a long time in order to make a profit overall-maybe for twenty or thirty years or more.It will probably take that long to pay off the loans!

Business men ARE interested in sustainability in a narrow sense of the word-meaning that the business model will last at least for the time frame under consideration.That would be forty years for a power plant designed to last for forty years.

I tend to use the word sustainable in a broader sense-meaning a practice is sustainable if and only if it can be practiced for an indefinitely long period of time, such as centuries or millenniums.

Slash and burn agriculture in a vast tropical jungle or rain forest would be sustainable if only five thousand humans were practicing it in a territory of a thousand square miles.The farmers would not need to return to any given spot for a century or longer-giving that spot plenty of time to recover.This could continue for thousands of years if nobody barged in and the locals didn't develop any new technologies

The kind of farming I do is sustainable in business terms;my business planning time frame is about two decades, which is as long as I expect to live, if I'm lucky, and as far ahead as I can see the future-thru a dark glass, dimly.Our little farm is well taken care of in just about every respect.But we are utterly dependent on petroleum products;we started mechanizing in the twenties, two decades before I was born, and the process was substantially complete by 1950 or so, even though my maternal grandfather owned one last pet mule until sometime in the nineties.

So we are not sustainably farming, in the true sense of the word.Whether we can continue our own little bau business for even another half century is extremely doubtful.

And I do recognize that EROEI has enormous , critical, grave implications for our society as a whole.It is not hard to understand that a barrel of oil produced in East Texas in the 1930's required only a trivial investment in drilling machinery and concrete and steel pipe and so forth, whereas a deep water well requires very large per barrel investment of equipment and materials.This necessarily means that we have less capital and for machinery and materials to supply ourselves with other goods and services-a thousand tons of concrete and steel buried thousands of feet deep in an oil well is a thousand tons less available to builds a bride or hospital of course.

I guess what I am really disputing is the meaning of the words we are using-sustainable, theoretical, and practical.

To me sustainably means indefinitely; practical means "bau" and theoretical -in this context- means "of academic interest only" because "bau" is the controlling paradigm.

Best, OFM

Fair enough OFM. If your point is that we are in a paradigm (BAU) where the vast majority of (business) people completely ignore EROEI, I can't dispute that. I just don't think it's a good paradigm for us all to be operating in, and I think that at least to a certain degree we can choose to think about things differently. I'd guess that probably you can agree with that.

In order to avoid fruitless arguments, I would simply take the EROEI for that battery to be approxmately 0.02. That is, in discharging it one gets back 2 units of energy for every 100 used in producing it.

This implies that batteries of that sort wouldn't be practical as primary energy sources, since any system relying on them as such would quickly grind to a halt with each recirculation being 2% of its predecessor. It follows that EROEI itself isn't a reason to manufacture and use them, so we might choose to think of them mainly in some other way - such as a convenient and expensive way to get a little useful energy to a spot where otherwise there would be none at all. (Some folks might take said choice to be a matter of dogma, but that's unnecessary - the alternatives are merely points of view.)

In reality, the lengthiest arguments will be over in-between cases - such as corn ethanol used as fuel, which might be seen as an energy source with EROEI a touch over 1.0, or instead as a Rube Goldberg way to convert natural gas or coal to liquid fuel. Arguing over whether corn ethanol is an energy source or an energy carrier is somewhat fruitless, often a matter of "spin" where the outcome depends on where one's bread is buttered. The real question is whether it is worth doing. In the right circumstances, 70 joules of liquid fuel may well be worth a great deal more than 100 joules of coal - but even then, using corn as an intermediate step could be far from optimal.

A mountain lion could hunt wild boar and get an EROI of 50 and meet all of its energy needs. The lion could also hunt antelope at an EROI of 25 and catch two antelope instead of one boar and still meet all of its energy needs, although it had to spend more time hunting. The lion could catch 20 rabbits at EROI of 10 and still meet all of its energy needs even though it would be hunting nearly all of the time. However it could not hunt mice with an EROI of 4 because it would not have the time to do all of the other things a lion must do to stay alive.

Declining overall EROI and declining oil imports will expand the energy procurement part of the economy and shrink other sectors. To maintain the existing level of complexity will require greater and greater draw down of lesser and lesser EROI resources. After it's determined that future delivery of energy and resources will not be forthcoming with interest - bills, notes and bonds will be defaulted on. We can try to meet energy contraction with conservation and expanded government risk guarantees for energy producers. Along this path many more things will become too expensive to produce and there won't be credit or cheaper Asian labor to make up for increasing costs of energy. The next likely target for economizing will be labor costs and associated expenses in the West.

Where does it stabilize? When ninety percent of the population are involved in producing food and energy and the automobile, colleges, and most medical care have withered on the vine?

However it could not hunt mice with an EROI of 4 because it would not have the time to do all of the other things a lion must do to stay alive. [...] When ninety percent of the population are involved in producing food and energy

Precisely, it's about time! And for human civilization, an EROI system of 4 might suffice, since it might not take that much of our precious time away, if the system is automated enough. (The lion is the system - humans create systems, and that's a big difference.) EROI and time may not be that linked, and for renewable resources, depletion is not an issue but environmental impact may be. I simply don't think EROI is that important.

But the "system" is already automated. Just think of "civilization" as an automated system for extracting energy from the environment to make more "civilization". This process has been going on for a long time. Early on, the EROI was low (hunting/gathering/subsistence agriculture), then with improved technology this was boosted to something like 100 (early coal and oil). This high EROI allowed a much bigger civilization, which in turn tried really hard, using the best available technology to extract more energy. The process is ongoing but EROI is falling because the process is intrinsically self-limiting. There is just no way around it because any EROI greater than unity implies exponential growth, which is not sustainable in a finite world.

But the "system" is already automated.

And gets more so all the time. Which is why EROI matters less.

any EROI greater than unity implies exponential growth

What, why? This makes no sense. Look at the lion/boar EROI example of Dopamine. The lion has a great EROI from hunting, and then use the surplus energy when lying around yawning. He won't grow exponentially.

The EROI of the lion hunting needs to be greater than one if the is allowed to do things like sleep have sex and rest. You could design a "better", more efficient lion that didn't need to sleep. What would be the result. The lion would be able to hunt more. The supply of antelope would fall as result. The lion would then seek to improve the EROI by making more automated hunting systems (lion cubs). After a few iterations, the lion population would crash.

You are basically arguing that humans are different and are not limited by resource constraints, that energy usage can increase to infinity. This is unlikely to be true.

Yes, the lion population would crash, but that would be the start of a convergence toward a new equilibrium in relation to its prey. And at equilibrium the EROI of the lion hunt would still be high, and the lion would still not grow exponentially.

You are basically arguing that humans are different and are not limited by resource constraints, that energy usage can increase to infinity.

No, I'm arguing humans, since they are building systems that have degrees of separation from themselves and their use of time and effort, are not (very much) limited by EROI. However, the systems are subject to resource constraints and environmental constraints.

A cheap low-EROI oil sands operation, for example, can fuel our civilisation only as long as the resource base isn't depleted or the climate doesn't kill us. A cheap low-EROI breeder operation, OTOH, may be infinitely sustainable since the fuel resource base is extremely large. (Ok, nuclear breeder EROI is likely to be high, but anyway, let's assume it isn't.)

Your premise in your last paragraph is that nuclear fission breeder operations (I will assume that means as the majority source for Humanity's energy consumption) will be cheap.

That premise is unproven.

I suppose we should first define 'cheap', including defining all costs involved.

A second point: You may wish to claim that your preferred process is 'sustainable for a long time' vice 'infinitely sustainable'.

Your premise in your last paragraph is that nuclear fission breeder operations (I will assume that means as the majority source for Humanity's energy consumption) will be cheap.

That premise is unproven.

Yes, it is, but nevertheless it will be. I won't try to convince anyone in this thread though - too off topic.

I suppose we should first define 'cheap', including defining all costs involved.

No, we don't, since external costs are insignificant. Or perhaps you mean that we should be careful to include external costs of alternatives when we compare to today's energy costs? In that case, it's a good point.

A second point: You may wish to claim that your preferred process is 'sustainable for a long time' vice 'infinitely sustainable'.

No, not really. Or perhaps I do, to avoid pointless arguments. I could be extremely conservative and say "millions of years" instead.

Do you recommend any books which outline the vision you advocate of large-scale breeder fission electric power reactors?

Also, do you recommend any web sites which outline the vision you advocate of large-scale breeder fission electric power reactors?

To your knowledge, have their been any key posts on TOD which outline the vision you advocate of large-scale breeder fission electric power reactors?

This are honest questions...you obviously are convinced of the potential success of this approach, and I like to read, and wish to read the same sources you have read.

I am looking for sources that provide adequate details.

I think many designs would do, including the traditional fast breeder. However, the somewhat hyped LFTR breeder seems to show most promise for economy and safety. I'm sure you've heard about it, and if you want to know more, you can always start at the wikipedia entry and then google.

I'm not very inclined to give you more than that, since I know that whatever I give you, you'll follow up with claims that the information is inadequate and provide some miscellaneous accusations to go along with that. You're on your own.

I have already read that Wikipedia article, as well as every other Wikipedia article for all the past, present, and future fission reactor designs listed there that I could find. I also have read numerous sources on Gen III, GEN IV, Gen V fission reactor designs in the past found using Google search.

Thank you for your time to provide your reply...but I can do fine without you!

I would think that to in all cases, EROI decreases to the minimum necessary to maintain equilibrium between the species and its resources. If a lion doubles it's efficiency in hunting, the population of prey will decrease until a the EROI decreases to its original value. In principle, the population of lions could fall in order to maintain the higher EROI (few lions, lots of prey), but that would require a conscious choice on the part of the lions to limit their population.

In any case, the EROI of lions is 1.0. The EROI of hunting is higher only because very artificial boundaries are set for calculating energy. The energy lions consume while resting/other activities is also necessary for their survival.

The the human civilisation also has an EROI of 1.

This article is about the extraction of energy from the environment to run machines, the lion hunting was someones analogy. But clearly EROI as defined in this article also has to converge to one in the long term. This is also clear from the first graph above.

You have to be cautious here, because the supply of rabbit might go down and probably not be sufficient. With higher consumption because of the lower EROI, the rabbit population may simply go down, increasing the EROI because of population changes, which will drive to further ressource depletion and so on until the EROI reach 1.

That means, the population of lions will be limited the the population of rabbits much more because of its lower EROI. A better EROI, such as the antelope, would put less strain its population since the lions would need less hunting.

In conclusion, the EROI is not necessarily a static value, it may depend of usage.
Secondly, lower EROI drive up the ressource depletion (or total utilisation in the case of a replenishable supply).

Does this analysis take into account 'entropic dissipation' of energy generation ? In other words do the EROI figures take into account the amount of energy required to remove the bi-products of energy production ? This could well prove critical if carbon capture technologies are taken into account. Also the energy cliff graph is my opinion doesn't seem all that alarmist.. all it is really saying is that if EROI drops below 8, then more GDP and resources are spent on acquiring energy resources rather than consuming or using the energy resources themselves.. not bad if unemployment stands at 9% - putting idle resources into play. Seems to suggest that current high unemployment is transitional from a consumer based economy to an energy based economy ? Important thing to consider on a Black Friday shopping day.

This EROEI issue seems to be plagued with difficulties, as Jeppen's comment and the rude reply to it illustrate.

For example, it's not obvious to me that 8 is a magic number that's OK, while 7 guarantees utter collapse. Nor is it obvious that as EROEI decreases there's any particular value at which the general flow diagram must suddenly be reconfigured to a new topology. Nor is it obvious to me that low-ish EROEI is intrinsically a problem; it would seem to depend on the labor intensity and the size of the primary supply.

It must surely be a tacit assumption - and one that probably needs to be made more explicit - that the input energy to an EROEI calculation must be, in some sense yet to be defined rigorously, economically (oh, the horror of that hated word) relevant. After all, in the real universe, the Earth only intercepts around 0.00000004% of the output of the sun, and we don't factor the 99.99999996% of wasted sunlight into EROEI since it's economically utterly irrelevant. No labor or other human input is required to produce it, it's simply there no matter what we do.

So, certainly, in principle, one could imagine a set of black boxes with an EROEI of 1.1, supplying 10TW for human use, but with a total energy flow of 110TW, 100TW of which circulates back internally to feed the harvesting/production process. Of course that probably couldn't function on tacitly-assumed horse-and-buggy technology since there wouldn't be any time for human beings to do the other things they need to do in order to live. OTOH, we don't live in Roman times any more; if our black boxes were highly automated, then we would indeed have time to do other things, so why not?

Now, one might certainly argue that 110TW couldn't conceivably be made available for that black box to run on, without creating other undesirable impacts, but of course that's a different argument. After all, many already argue that we can't go on producing 10TW by present-day, fairly-high EROEI methods, without creating excessive impacts: their argument is nothing much to do with EROEI.

It's true that this EROEI concept has a lot of problems, which is the main reason why it is not one of the metrics used in the energy industry to determine if a project is viable or not. However, as Rockman has pointed out various times, if the EROEI of a project is too low, then it will fail several of the other metrics which are actually used by the project analysts, and it will not be done regardless.

This is not absolutely true under all conditions: A project which converts one form to another form might proceed if the output energy has a higher utility than the input energy. An example would be a gas-to-liquid (GTL) or coal-to-liquid (CTL) plant with an EROEI of 0.6 - although even in that case it might take wartime conditions to justify the economics.

It must surely be a tacit assumption - and one that probably needs to be made more explicit - that the input energy to an EROEI calculation must be, in some sense yet to be defined rigorously, economically (oh, the horror of that hated word) relevant.

Absolutely. The tacit assumption is that EROEI stands for "energy returned to human beings over energy invested by human beings. The italicized part makes it inherently economic.

What most of us are trying to do here on TOD (whether we know it or not) is reform/recreate/revolutionize economics so that it would account for energy flows in the (human) economy. We may belittle 'economists' because that profession has become dominated by a caste of people who ideologically deny the obvious connection between the physical world and the human one, but there is no question that 'we' are competing with 'them' for authority over the same subject area. We can decide to argue them out of their word ('economics'), or we can decide to come up with a new word to call our own, but that's just a matter of rhetorical tactics.

This EROEI issue seems to be plagued with difficulties, as Jeppen's comment and the rude reply to it illustrate.

Understatement of the year......

The problem is more about the iterations necessary to reach your goal and the amount of energy to start with.

With an EROI of 1.1 and 1 MW as a starting point, you'll need 169 iterations and waste 91% of the ressource.
With better EROI, such as 40, you'll need 5 iteration and only waste only 2.5% of the ressource.

Here I assume the EROI stay constant over the iterations, which is historically very wrong. It should go up a little and then reach a point where the inputs become more and more rare which drive up the ressource depletion and lower the EROI to 1. This is so because there all sorts of inputs and usually one come to depletion much faster then the others (be it time, labor, materials inputs, space, etc).

So yes, EROI dosen't answer everything but unless you can find a lower EROI source with much bigger ressource (with all inputs included) and no pratical limit on iterations it is likely to prevail.

It is commonly thought that scaling renewable energy will require the adoption of some sort of storage system to account for times of over- and under-production. ... The question is whether this added energy cost [for storage] will decrease the EROI of these systems below the EROI threshold, but to my knowledge, there are no peer-reviewed papers reporting EROI numbers that included these costs.

The real question is whether we really need that much storage. Is there any reason to believe that intermittent renewable energy won't still be economical, at least for a significant portion of use, simply when it is available? The belief that the 24/7/365 electrical grid must continue ought not to be casually assumed. What is 'commonly thought' ought to be thought a bit less commonly.

We'll see if some scalable, economical use for intermittent energy surfaces that provides some appreciable floor for electricity prices. I've read about the occasional negative spot price in high-wind-penetration regions with feed-in-tariffs. Wind operators won't turn their turbines off-wind until the spot price is the negative spot price, and thus there should be a market for resistor-bank operators charging money to burn off the surplus electricity.

We ought to be able to come up with better dump loads than resistor banks. The question is what loads can use intermittent sources, and could be left off when there isn't excess intermittent energy.

There should be an ample market for the energy in cold climates during the winter season-resistance heaters are dirt cheap, and the heat could be "dumped " into homes and offices , thereby saving a lot of ng, heating oil, and coal too-the coal savings being indirect through using less coal generated electricity for heat..

It might even be possible to use it to heat an underground reservoir, so that a community sized ground water heat pump drawing heat from the reservoir would operate more efficiently with a higher output.

But in the end, I believe thye economics will work out in such a way that the windfarms will not need to sell every kwh they can generate-the price of coal and gas will probably rise to the point that it will be profitable to shut down every possible kW of conventional capacity.

Most businesses operate profitably without deploying assets with maximum efficiency-even passenger jets sit around unused part of the time.

Most factories shut down on weekends and holidays.Farmers and loggers work when the weather is favorable.

An inability on the part of wind farms to sell 100 percent of their possible output is not likely to break the industry.

It also seems likely that new coal and gas generating plants can be designed to be more flexible in terms of fast shutdowns and startups-even if this must be accomplished by going to modular construction.

All the steam from a plant need not come from one or two giant boilers-and all the electricity need not be generated by one or two giant dynamos.

The actual price of fossil fuel is eventually going to trump capital costs of conventional plants to the extent that the owners will be glad to shut them down and "burn" wind whenever they can;and anyway, the utilities are likely to own most of the windfarms in the end, which will take most of the infighting out of the equation.

But in the end, I believe thye economics will work out in such a way that the windfarms will not need to sell every kwh they can generate

I just did a simple simulation with a hourly data set from Sweden's first eight months of this year. For a simple spot price model, I just count hours that produce less than half of consumption from wind as fully paid, and other hours I set to zero payment. Today, 4% is produced from wind and naturally, 100% of wind generation was paid under the model. Then I scaled wind production by multiples and looked at the numbers then:

1x today: 4% of consumption, 100% paid.
5x today: 20% of consumption, 95% paid.
10x today: 39% of consumption, 47% paid.
15x today: 55% of consumption, 23% paid.
20x today: 67% of consumption, 12% paid.
25x today: 74% of consumption, 7% paid.
30x today: 80% of consumption, 5% paid.

Actually, I think this calculation may have converted me from nuclear to wind. I didn't think this much of consumption would be covered. 25x current 4% is 100%, and my calculation shows only 26% of that is wasted, which isn't that much. So, you could cover a reasonably big country's needs to 74% by very moderate overbuilding, if you have the wind resources. Also, you could probably get a few percentage points higher by smart grids and long distance transmission. It needs support, since the electricity produced won't "sell", as my numbers show, but that's a bit of socialism I could live with. It also needs almost full backup, but I guess that could be handled too.

J,

Do you happen to have a source for that data from which I could download?

Swedish source data is here. The middle series has the hourly data for the last 10 years (I used just the 2011 figures). If you'd like to see my mangling too, I could try and find a way to make it available.

Thanks!

"Most factories shut down on weekends and holidays. Farmers and loggers work when the weather is favorable."

That's the kicker, isn't it? The first time I read the article a couple months ago about a concentrating solar factory (skipping the electrical transmission part of solar), I thought it was laughable. Then I drove past the gigantic cement plant on the way out of Tucson, and it sunk in that most industrial processes mostly require heat, not electricity and it clicked - storing electricity is both hard and expensive. Storing heat is doable, but still not ideal. Storing concrete is dirt simple. Storing smelted steel is easy. Storing milled grain has been done since grain was milled by hand. Plenty of people work only when the work is available now (contractors, etc.). If you want to talk about economic substitution for oil, I think one of the necessary substitutions is going to be in the temporal aspect of industrial processes. When it's cloudy, most of the employees at the cement plant get unpaid vacation, when it's sunny, overtime pay. Working 24/7 without regard to the environment is a very new development in most human endeavours. Food production still largely follows the seasons- let's be happy that the intermittency of renewables fluctuates weekly, not yearly.

Best hopes for good meteorologists & flexible work schedules,

Steve

How do you fix the system boundary?

Some of the figures being quoted seem ludicrously large, but if you can draw the boundary where you want, you can get pretty much any figure you want by cancelling out flows within the boundary.

For example, I draw the boundary far out, and I see 1 W going in and 100 W coming out. Looks good. I peer a little more closely into what is going on inside the boundary and I see a 1 kW flow. If I tighten the boundary, I now have 1001 W going in and 1100 W going out. Not so impressive.

Is there some way of deciding what is an input, what is an output and what are interior flows that it is valid to cancel out, that isn't down to however the person doing the calculation decides to define it?

Good points and questions.

As a general rule, if the point of looking at EROEI is to compare the net energy of different technologies (which is what the point should be), then the larger the boundary area, the more meaningful the calculation. We want to know how technology choices affect the net energy available to a society. Looking at small pieces of the process doesn't show us that. A lot of the confusion about EROEI that goes on in threads here involves people not understanding the difference.

Of course, producing accurate calculations, which can be compared as apples to apples, is harder than taking a walk in the park.

This is a classical problem with the life cycle analysis. There is a standard ISO 140040 procedure to handle that. Off course, nobody is using it in the biophysical economy community. Nevertheless, they have been some effort to do this formally reticently. Part of my research project next year.

Another way to go round the boundaries problem is to use the EIO-LCA approach they calculate all the energy flow trough the society. Carnegie Mellon Green Design Institute are the pioneer in that field. Another research project for the next year.

LCA is rather different and this problem doesn't happen. With LCA the problem is how far do you need to evaluate the consequences of the consequences of the consequences.

I agree that you can use an LCA methodology, but I don't see why its any better than some other methodology. LCA methodology is an agreed way of terminating a calculation sequence that is expected to converge. As you include more recycles in an EROI, it diverges.

Does gas split off to generate power for running a gas field count both as an input and an output or as neither? Does fuel burnt in situ count both to the input and the output, or to neither? In practice you will calculate different numbers doing the LCA one way rather than another, but its a different calculation method for what is in principle the same thing. The EROI is different in principle as well as in practice.

So, is there a way of deciding which inputs can be cancelled with outputs and which can't, that isn't based on arbitrary choice?

You use an LCA package and it calculates 10 in, 100 out, EROI=10. I identify an interior flow of 1000 and calculate 1010 in, 1100 out, EROI = 1.1. Someone else sees a way to use 9 of the output for inputs and calculates 1 in, 91 out, EROI=91.

We have numbers from 1.1 to 91 and its all down to an arbitrary choice of what is counted as a recycle and what is counted as an input plus an output.

Indeed, we could probably all use the same LCA package and by slightly different definitions of what we were doing the analysis on, come up with widely divergent numbers. Those different definitions would have trivial consequences for LCA, but arbritrarily large ones for EROI.

This is why there is an effort in the biophysical economy community to formalize this. Everyone is aware of these problems. Already, there is some effort to address these issues.

So basically EROI is rubbish, but some people are trying to find a way to make it not be rubbish and so far they don't have a clue? I thought you would have a method and I wanted to know what it was. There are plenty of ways it could be done, and economists having been using them for ages because cash is cash, just like energy is energy.

e.g. the energy profile over time is calculated. The deepest minimum is taken as the input and the net surplus at the end is taken as the output.

It won't allow things like "society" to have a robust EROI associated with them, but it would allow a robust calculation for something like a nuclear power plant you don't get a result that wasn't an arbitrary number dependent on just how the nuclear power plant was described. Comparing a nuclear power plant against a THAI project would be rather less robust, but that has more to do with not all energy being equally useful, and the costs and benefits spread over different time periods rather than not having a method to decide whether something is a recycle or an input and an output.

Where the different qualities are important you might use exergy rather than energy, and a rate of return where time periods are different.

Actually, there is something like this. The key issue is that it has not yet been formalized in the same way as the LCA in the ISO 14000 norm. That just that. Most people working with EROI understand these issue. There is effort to standardize the analysis. This does not means that the previous analysis were wrong, they lack of a very standard definition of boundaries. These papers have been published this year:

http://www.mdpi.com/2071-1050/3/10/1972/
http://www.mdpi.com/2071-1050/3/10/1908/
http://www.mdpi.com/2071-1050/3/10/1888/
http://www.mdpi.com/1996-1073/4/8/1211/

They cover all the concern that has been expressed here about the EROI.

The EROI as defined in the first paper suffers from exactly the deficiencies I outlined earlier. Far from dealing with them, it glosses over them and assumes that you won't notice that the sort of thing they are talking about as an EROI in one part of the paper isn't the same thing as what they are talking about as an EROI in another.

Far from appreciating and acknowledging this they state "EROI is a physical property of an energy
source" and use this as a basis for their discussion. As they use it, it is an arbitrary construct and they manifestly fail to appreciate this as they pull EROI values from different sources and use them comparatively without any discussion of the manner in which the authors of the other papers defined and calculated their EROI. Whether or not other things are equal, you've got to take some pains to make sure that the thing you are dealing with is.

They show an appreciation that there is a difference between capital and income, but they build a flimsy construct on shaky foundations with it rather than trying to sort out exactly what is capital and what is income.

...

The second paper is a lot better, but it still doesn't address the problem of why one boundary is the right one to choose. They give a method which is capable of general use, show they can use it to analyses a system which it is easy for their method to work with, but it still boils down to, "Our method is best because if everybody used it their results would be consistent with one another when used on the same system". I think the analysis itself has interesting things to say about wind turbines, and if you did something similar on a shale frac you would come up with some interesting results on shale fracking, but I don't think comparing 6 (the EROI for a windturbine from their analysis) with 4 or 8 or whatever you came up with for shale fracking would tell you anything interesting.

Again they lack an appreciation that the EROI changes in an arbitrary manner depending on whether something that you can identify is classed as an internal recycle rather than an output plus an input. They use an example system where they can finesse it because forms in which input and outputs are embodied are different. Use it on a thermal heavy oil extraction process where the heat can be obtained by buying in fuel, or burning some of the produced oil as fuel and it will fail.

Its a nice analysis, and an interesting paper but a shame they used EROI as the way in which to present it.

...

The third one.

Lol.

Another paper which tries to hide the basic shortcoming behind pretentious overcomplication.

We can't come up with a single defensible way to to define the boundary so we'll define it in a score of different ways and hope no one notices that we now have 15 numbers each of which is subject to arbitrary choices of where boundaries lie rather than 1.

Not so much "ceteribus paribus" as "reductio ad absurdam".

...

The fourth one.

Rather like the third, but with some potentially useful methods in it. I get the feeling that these authors reckon EROI is pretty rubbish, but don't dare say so because they wouldn't get published. Its a pity they don't actually use their methods to analyse a few systems, because the detail of the analysis could be interesting, though an EROI they quoted at the end wouldn't be.

...

Thanks for giving me the references. I think they show some interesting things being done in energy analysis, and that EROI is a good for a laugh and not much else.

Yvan Dutil
Thanks for the reference to the Carnegie Mellon Green Design Institute http://gdi.ce.cmu.edu/ Some of their publications look interesting.

Please keep us posted on your research, findings & publications.

Should have said this above...

The overall point of David's piece here is new to me, very well presented by him, and obviously of deep importance.

I think what is lacking is also a time frame. What is the EROEI over say a 20 year period, 50 year period etc? The thing which is lacking here IMO is the understanding of recycling existing infrastructure or in the case of hydro/nuclear long lived structures. When you already have wind turbine shafts, transmission equipment etc and the ability to recycle the existing generators and blades on an existing wind power installation or similar solar installation what is the second/third generation EROEI? If we expect fossil fuels to start winding down in 20 years time then we should be taking into consideration what the second generation power equipment on site performance would be like. I think the renewable energy won't assure us of cheap energy in the near future but in the far future it definitely will.

This is the great tragedy of not embracing renewable energy sooner. Had people been moving towards renewable energy in a strong way 20 years ago then we would be refurbishing the existing sites with higher capacity equipment with excellent ROI and EROEI and at the lower prices fossil fuels like coal and natural gas would flat out be uncompetitive. Since that can is continually being kicked down the road we're left trying to push through first generation sites with all the associated costs.

What is the EROEI over say a 20 year period, 50 year period etc? The thing which is lacking here IMO is the understanding of recycling existing infrastructure or in the case of hydro/nuclear long lived structures.

All of those things can be considered in an EROEI calculation, and the more seriously and comprehensively they are considered, the more credible the calculation can be said to be. But these things don't really affect David's central point.

A proper calculation of EROEI includes the cradle to grave energy consumption and output, including recycling the equipment at end-of-life. Nuclear power plants are designed and licensed for a lifetime of 40 years and it's not clear that extending their use beyond 40 years is a good idea. Dams last an even longer time, but there is usually point at which the lake behind the dam fills with sediment and the dam no longer provides useful energy storage, only run of the river power. Then too, hydro turbines wear out from the constant abrasion of the sediment carried by the water and corrosion is a constant problem...

E. Swanson

The AP-1000 has a 60 year design life.

Nuclear reactor itself do not affect much the EROI. Full processing is the limiting factor. Lenzen did a very good meta-analysis about that a few years ago.

How much energy is used in the decommissioning process for each reactor?

There is a link in one of the recent DBs to an article stating that Switzerland will spend ~ $22.5B (expressed in USD) to decommission 5 nuclear reactors. However, as you said above, I think the article said that the majority of that cost was for disposal of the fuel assemblies.

E-ROI analysis is a useful tool, but scale is more important.

The analogy here is Soviet toilet paper: if you subsidize it too much, the toilet paper makers start buying it from the stores, and using it to make new toilet paper, because it's cheaper than the normal wood feedstock!

Obviously, you don't want that kind of thing, but....

Look at the transition from hunter-gathering to farming: hunter-gathering was far higher E-ROI, but it didn't scale.

Similarly, local wood is higher E-ROI than distant, deep coal (the figures above are mine-mouth), but local wood didn't scale.

Scalability (maximum volume that can be produced) is far more important than E-ROI.

Scalability (maximum volume that can be produced) is far more important than E-ROI.

I don't agree at all. A society that invests 1 unit of energy and gets back 50 units to consume is going to look very different from a society that invests 40 units and gets back 50 to consume. They both produce energy on the same scale but the first society gets to spend almost 5 times as much energy on luxuries.

This is like the debate about population and energy-per-capita. The best answer is that they both make a big difference and that it is not meaningful to say that one is more important than the other.

It doesn't matter what the E-ROI is: if it doesn't scale, it can't be used.

Hunting and gathering has been pretty thoroughly abandoned (except for a few places in the US South...). And, yes, an agricultural society looks very different - not nearly as much time to play - but it can support at least 10x as many people on the same land.

So, scalability is more important - it's the gatekeeper, the sine qua non.

-------------------------------------------------------

I do agree that there are many things that must be considered along with E-ROI: cost (mostly labor) per unit of energy, ratio of capex to opex, energy quality, energy timing, waste disposal and other externalities come to mind. But scalability is different.

It doesn't matter what the E-ROI is: if it doesn't scale, it can't be used.

And if it scales but doesn't have a (sufficiently?) positive net energy, then it (eventually) can't be used either.

So, scalability is more important - it's the gatekeeper, the sine qua non.

The sine qua non for what? Total energy use? Population? Are those the only things that are 'important'? Were the lifestyle changes between hunter/gathering and agrarian society 'not important'? I don't think so. Perhaps you should use more precise language.

David's point in the keypost is basically that we could agree with you when EROEI is above 8 or so. That's fine. But I don't think you can extend your point much farther than that though.

They both produce energy on the same scale but the first society gets to spend almost 5 times as much energy on luxuries.

Maybe, maybe not. The 40/50-unit energy production might conceivably be scaled up to 196/245, yielding the same 49 net units. If we assume that this must be done exclusively with horse-and-buggy technology, it probably won't work out due to the necessary labor being unavailable. But if the energy production is highly automated, that might not be a big issue - and after all, even now there are enormous quantities of energy flowing outside of the accounted-for parts of the system.

Now, as I said elsewhere, it's fair to ask whether it's practical to scale up the 40/50 system that far, in terms of resource size etc., but that question is separate from EROEI, and the answer will depend strongly on what system we're trying to scale up (simply not enough land to scale up corn ethanol enough to really matter, for example.) And note that as Rockman tirelessly points out, systems with very low (whatever that means) EROEI typically won't work out financially anyhow, since the energy inputs do cost money. So the practical folks can essentially always safely ignore EROEI as an independent consideration.

Maybe, maybe not. The 40/50-unit energy production might conceivably be scaled up to 196/245, yielding the same 49 net units

Both of them can be postulated as larger or smaller; the magnitude of the example numbers is irrelevant. The point is that the two societies are going to have very different social and economic structures for the same scale of energy production. And that is 'important'.

So the practical folks can essentially always safely ignore EROEI as an independent consideration.

Just because the 'practical folks' - of which I happen to be one - can usually translate EROEI into dollars does not mean EROEI is not an 'independent consideration'. When I try to install a solar array to maximize production, I understand that the goal is to increase both the EROEI and the financial ROI. If I didn't understand that one leads to the other, I wouldn't know why it was important to do so. Some of my colleagues who don't understand it so well do our customers a disservice because they don't work as hard to do it. The financial return to them doesn't change. And as for the customers, if they don't understand it, they are more likely to get cheated.

Both of them can be postulated as larger or smaller; the magnitude of the example numbers is irrelevant. The point is that the two societies are going to have very different social and economic structures for the same scale of energy production. And that is 'important'.

The magnitude of the numbers is very important - that is the essence of scalability. I can reduce the water use of a bathroom faucet by 75% by changing a 2gpm aerator to to 0.5gpm, for the cost of about $1. This will save about 1.5 gal/capita per day. But that's it, it cant scale any more as you only use the bathroom faucet so many times (about 2 minutes for an average person) But by changing the toilet from a 5gpf to a 0.8/1,25 gpf dual flush, I can save 20gal/person/day, at a cost of $300 for the changeout. The EROEI of the faucet aerator is way ahead, but it only scales to 1.5 gpm/day, or about 2% of per capita water use. The toilet is far lower EROEI, but scales to 20% of per capita water use.

If a town is trying to avoid an expansion of water and sewage capacity - which one should they go for? Changing all the aerators seems like great value for money, but the savings are almost noise, and won;t put off a capacity expansion, so they don;t actually *achieve* much. The toilets actually achieve a significant decrease, and can/do delay/avoid expansions.

It all depends whether the objective is to get the best return on investment, or to make the biggest total change. An investor looks for % return, but how good is a 200% return on investment if you are only allowed to invest $1, compared to a 20% return on investment, if you can invest $300? The aerator creates $2value per capita/per year, and the toilet creates $60 per capita per year.

The scalable solutions are the only ones that are able to make a significant change. This is the basic theme of people like Tom Murphy (at Do the Math) and David Mackay (Sustainable Energy Without the Hot Air). All of these seemingly great "solutions" are of very limited use unless they can be scaled up large enough to actually make a significant difference.

Wind and solar can be scaled to very large amounts. Hydro has rather strict limits, but has been scaled enough to be useful. Collecting energy from people travelling down in elevators, for example, has a great EROEI, but can;t be scaled to make even 0.001% of total energy use, so it might be worth doing for whoever can do it, but it is not going to solve any of societies problems.

Which are the solutions that seem to get the most "PR" these days - not the scalable ones!

The magnitude of the numbers is very important - that is the essence of scalability.

The magnitude of the numbers in the example is not important to the point I was making.. They were unitless numbers that could signify any scale. Just as 50/1 is going to be very different from 50/40, it will also very different from 245/196, and also very different from 250/5. The experience of living in any of those societies is going to be very different from the experience of living in any of the others, which shows that EROEI is important, too.

Try to read other people's comments more carefully. I was not saying that scale is not important, in fact I explicitly said that I wasn't saying that. What I said in exactly so many words was "it is not meaningful to say that one is more important than the other."

Actually, if energy production in the lower-EROEI society is highly automated, it need not force a dramatically different social or even economic structure. All that matters in these respects is the amount of people's time the sector consumes and the amount of output it produces; with automation it might conceivably have very large recirculating flows without requiring a noticeably burdensome number of employees. (Naturally, and as I implied, if one insists on implementing laborious horse-and-buggy technology, then one forces a different social structure, but that owes to withdrawing modern technology rather than to low EROEI itself.)

Of course, and lest we forget, depending on the nature of the recirculating flows, their size may certainly matter in other respects, such as there isn't enough arable land on Earth for corn ethanol to be a truly major source. But it's not useful or necessary to conflate separate factors - it just creates confusion, and the factors will have their effects just fine without any need to conflate them.

A society powered largely by wind turbines would be a close approximation to your concept - once they are all built.

This is not a new idea, of course - it was explored by HG Wells in "The Sleeper Awakes".

His transport system was actually moving ways - large conveyor belts. For in city peak transport they would probably be pretty energy efficient, once you resolve the problem of getting on/off them.

And, unlike corn, there is plenty of scope for expanding wind (and solar) - at a cost, of course, but once done, the man hours are minimal.

Sounds like Heinlein stole the idea for "The Roads Must Roll". Asimov also used it in his Robot novels...

Actually, if energy production in the lower-EROEI society is highly automated, it need not force a dramatically different social or even economic structure.

I still don't agree. A high degree of automation, or the lack thereof, is an important social and economic difference. I realize that to a certain degree this is semantic, but my point is that changing from one type of society to the other will involve changes that ordinary people experience as revolutionary. A high degree of automation means that either more people will be monitoring the automation, or else, if the automation can take care of itself, we'll be handling our lives over to the robots. In the former case, we need massive changes to our educational system, constituting a big social change. As for the latter case, just look at all the scary science fiction that scenario has inspired, and you can easily see that it has social implications as well.

We're going to have to agree on how to calculate EROEI for hybrid electrical systems, particularly
- solar thermal with inbuilt gas burners to keep molten salt hot
- wind systems with gas backup elsewhere on the grid.
However that calculation is done it should be equivalent to a weighted average along the lines
x [EROEI(wind)] + (1 - x)[EROEI(gas)]
The EROEI of the backup includes not only fuel but financing costs, administration and transmission. The purpose of this exercise is to highlight the fact that fuel burn backup will show diminishing EROEI over time so the intermittent energy source is less attractive. Some have suggested when gas backup is prohibitive we burn biomass like hay or woodchips to keep steam boilers running. In that case the weighted average approach should give a quick answer.

Illustrative calculation for gas backed wind, with EROEI for wind 20 and for gas 15, weights 30% and 70%
.3 (20 ) + .7 (15) = 6 + 10.5 = 16.5
Suppose some time from now gas is in steep decline and woodchips look better with EROEI 5
.3 (20) + .7 (5) = 6 + 3.5 = 9.5 not so good.

Illustrative calculation for gas backed wind, with EROEI for wind 20 and for gas 15, weights 30% and 70%
.3 (20 ) + .7 (15) = 6 + 10.5 = 16.5
Suppose some time from now gas is in steep decline and woodchips look better with EROEI 5
.3 (20) + .7 (5) = 6 + 3.5 = 9.5 not so good.

The calculation is not quite as simple as that. Keep in mind, with electricity, that we have very distinct peak and off peak periods.
For simplicity, lets say the peak/off peak are each =50% of the time, peak demand is 2x off peak, and that backup mostly only operates during the peak period.

So for off peak, we have one unit of wind only at 20, and for on peak we have two units at your combination at 16.5. We end up with (20+2x16.5)/3= 17.67

For the biomass solution, we have (20+2x9.5)/3=13 Still pretty good.

Note that in both examples wind is supplying only its capacity factor of 33%, but we could get more by overbuilding
The backup is available 24/7, but is *very* rarely used in the off peak hours.

The fact that many NG peaker plants are simple cycle turbines, at about 3/5 the efficiency of combined cycle, is an example of accepting a lower EROEI for those vital peak periods, where the energy has a higher utility.

This argument doesn't apply to oil, of course - which is just another reason why we have to be careful in making broad brush statements about "energy" use - there are significant differences between type, time and place.

A concern that I have is not just the falling EROI and its impact on the modern global economy, but the $trillions of dollars of so called wealth-captial that has been sunk into increasingly worthless derivatives. Matt Simmons stated that he believed the USO ETF was created to short oil. If we look at this 3 year chart we can see how the USO has fallen in value compared to the WTIC price.

http://67.19.64.18/news/GoldenJackass/2011/10-26gj/2.jpg

Request: Any tips on posting pictures here. I used the img tag and still no luck.

Wall Street's ability to invest into reserves of Shale Oil Wells can only take place in this continued Fiat Currency Fractional Reserve Banking System we are currently utilizing.

With the collapse of MF Global and probably others (Jefferies), investors are starting to pull money out of futures and options. Trust is eroding away in the futures markets that used to be taken for grant it just a few years ago.

I believe the EROI of oil and gas is presently considerably lower than what has been estimated by Hall, Gagnon, Lysle, Murphy and Cleveland to name a few. The huge siphoning of captial into derivatives, especially in the past 10-12 years has created a severe mispricing of commodities... ETF's are not physical assets.

There is a real threat that the current Fiat Currency Economic System will disintegrate. If this is so, then where will the venture capital come from to allow Shale Oil-Tar Sands and other low EROI energy production to continue? This, I believe is something very few are considering.

Lastly, if we have a global depression, Peak metals will more than likely occur within in the next year or two. I wrote about this in my article Peak Silver Revisited: Impacts of a Global Depression, Declining ore grades, and a Falling EROI:

http://www.marketoracle.co.uk/Article30905.html

but the $trillions of dollars of so called wealth-captial that has been sunk into increasingly worthless derivatives.

Sunk how? Would the problem be that it gobbles up money supply, making money too tight?

Request: Any tips on posting pictures here. I used the img tag and still no luck.

The img-tag is the way to go. I typically use something like this: <img src="url" width="400">

The huge siphoning of captial into derivatives, especially in the past 10-12 years has created a severe mispricing of commodities...

What does the underlying mechanism look like? To my mind, derivatives can't affect commodity prices very much.

Fiat Currency Economic System will disintegrate. If this is so, then where will the venture capital come from to allow Shale Oil-Tar Sands and other low EROI energy production to continue?

Of course, if the credit system freezes up, we have a problem, but the problem is general and not connected to tar sands. Why wouldn't the tar sands extraction be viable in another (functioning) system? I don't see why fiat would have anything to do with it.

jeppen...it is my contention that the derivatives markets (not all) have been designed to pull money away from physical assets and into financial paper. The Falling EROI is the disease, the money printing & explosion of derivatives are the symptoms.

If $10 trillion is invested into derivatives (i.e. Annunties, Treasuries, CD's, IRA's, Futures, Options and etc) than it is not flowing into the ACTUAL PHYSICAL ASSET. I realize the modern economic system we live in allows us to create every financial product under the sun, but it does not mean they are REAL ASSETS.

Investing into a silver derivative is not the same as parking your hard earned money into physical bullion kept in a safe place in ones home or private secure institution. For example "Silver Certificates" are not backed by actual physical bullion at a bank, but rather by the "Assets of the Bank"... whatever they may be. These assets may include MBS or CMBS that may turn out to be liabilities rather than assets.

The amount of paper silver derivatives compared to physical metal maybe estimated at over 100/1. The collapse of MF Global has put a huge damper in the trust of the futures markets. When investors finally wake up and demand physical silver from their certificates, pool accounts, ETF's, and futures-options, there will not the metal to back up these paper instruments. This goes for many other commodities as well.

The implosion of the Fiat Currency System will destroy the derivatives markets which will destroy Annuties, US Treasuries, Money Market Accounts, IRA's, Bonds and etc. These instruments are the very means at where the investment comes to fund future projects in all aspects of the economy.

You say that you think derivatives don't affect commodity prices. How can they not?? 80% of the total derivative market, estimated to be in the hundreds of trillions or even more, is in Interest Rate Swaps. Interest Rate Swaps are backing the US Treasury market. Without them, the US Treasuries would be acting more like many of the European Bond markets. There is no free market and there is no real price discovery mechanisim in commodities.

Again, when the Fiat Currency Fractional Reserve system implodes, so will a great deal of the presumed wealth-capital along with it. There will not be the available captial to invest in Low EROI energy future projects. This is also true in the mining industry. Almost 10% of the world's energy is consumed in the mining industry.

I hope I am wrong about my assumption, but I believe 2012 will be the year the cow excrement finally hits the wind turbine. No one knows for sure the ramifications... but I believe they have been greatly underestimated... as well as the EROI ratios.

In one of my ASPO presentations, I offered up four caveats to EROEI that I think should always be kept in mind.

1. A process could have a poor EROEI and be economical (but not sustainable)
2. There is no time component in EROEI
a. If one process is 2/1 and one is 10/1 – what is the turnover time?
b. Should be standardized to annual basis
3. Biomass inputs are frequently omitted
a. Sugarcane ethanol not really 8/1; that is output/fossil fuel input
4. Inputs and consumed energy often conflated
a. This mistake is the source of EROEI of 0.8 for gasoline and 1.5 for ethanol

I don't believe one can say that even 2 to 1 isn't good enough. Imagine that I can turn that process over every day, but my 10/1 has a turnover time of 1 year. Which one nets me the most energy per year? The 2 to 1 process.

So I think EROEI has utility, but it needs to be used with some understanding of the caveats.

I might repeat something that already been say EROI is calculated from cradle to grave and indeed include the time. Bob Henrenden has worked on this question since the end of the 70's.

I am not sure what you are saying here, but if it is that EROEI -- as discussed in this post -- does include a time component, then that is inaccurate. When someone reports an EROEI of 2 or 10 they have not calculated those on the basis of an equivalent time period.

If on the other hand you are saying that some have worked on this problem, I would agree. It is just that those caveats are never applied in practice when people discuss EROEI.

So I think EROEI has utility, but it needs to be used with some understanding of the caveats.

Could you please explain its utility, without the caveats negating that utility.

An energy metric in an economic system just don't appear to make sense. Certainly Rockman has verified that industry does nothing with these types of metrics, and I have yet to see anyone, in any industry, ever, claim that a project gets a go/no go decision based on the energy input and output ratio. I might be wrong, but I have been searching for quite some time and haven't found it, heard of it, seen it published or blogged about yet.

There is a direct relation between energy and money. People do calculation in money, but those calculation could be done in energy and equivalent decision factor can be generated.

There is a direct relation between energy and money.

Sure there is. And it is non linear in nature, varies somewhere between (and possibly including) zero and infinity, and changes from day to day, and place to place. All BTUs are not created equal, haven't stayed equal throughout human history, and certainly won't be the same tomorrow as they are today.

Can academics manufacture some cool assumptions, generate a number, and pontificate at length about alleged conclusions (while guys like Rockman can testify to its uselessness in their industry)? Sure. The result? Hall and Cleveland using net energy in 1981 to show that all drilling in the US should cease sometime before 2005.

Find me an industry which makes it decisions in the real world based on this metric, awards employees who have improved their units EROEI, promotes the EROEI leaders from within, demands proven EROEI success from new hires, shows the bank why they need all this new capital to increase their EROEI further, and trumpets the results of EROEI improvements to Wall Street, who rewards their stock price handsomely. It is a rhetorical, and satirical demand, of course. The relationship you claim, if direct, would allow all of this to happen. Yet none of it does.

They do it implicitly. When you by 1$ of good or services, you by the X BTU that as been necessary to do produce the good or services. When you do a calculation in energy instead of dollar, you usually don't make any money if you are producing energy with an EROI is below 10.

Also, there is an inverse relationship between EROI and the price of oil. Those thing are all factored out in the economic system. Actually, the relationship would be even closer, if calculation were done in term of exergy instead of energy.

Could you please explain its utility, without the caveats negating that utility. ...

1> Just because industry doesn't use it doesn't mean it has no utility. That is like arguing that steam engines have no utility because no one was using them in the 15th century.
2> Even if it has no utility concerning short term for-profit investments, that doesn't mean it has no utility in long term predictions and planning. If there is one area where EROEI probably has the most utility, it would be in considering whether energy technologies are worthy of venture capital, or research funds. If they have an EROEI significantly below 1, they probably aren't worthy, especially if there is no other benefit.
3> Even if it does have no utility to industry, that doesn't mean it has no utility to policy makers and ordinary people in planning for the future.

"Could you please explain its utility, without the caveats negating that utility."

Sure. Let's say we are talking about fungible or relatively fungible energy inputs. It does not make sense to encourage energy systems with a low EROEI. As the simplest example, one would not convert 1 BTU of gasoline into 1 BTU of ethanol. However, subsidies may be in place that encourage those types of conversion. EROEI tells you that they make no sense.

The classic example is Soviet toilet paper: if you subsidize it too much, the toilet paper makers start buying it from the stores, and using it to make new toilet paper, because it's cheaper than the normal wood feedstock!

This clarifies that subsidies are the distortion of $-ROI that E-ROI helps catch.

2. There is no time component in EROEI

I think a better way to say this would be "a final EROEI figure does not tell you anything about the timescale of the energy flow". There are plenty of EROEI figures that include a 'time component' in the calculation.

b. Should be standardized to annual basis

It seems to me that proper EROEI calculations already do this, so I don't understand what you're suggesting. If a solar array requires 5 units of energy invested, and produces 30 units of energy over it's 30 year lifetime, then the EROEI is 6. The 'annualized EROEI' is also 6. What's the difference?

Payback time is important. If payback is to long and your are growing at exponential pace (ex: PV), you might end up without any energy produced until a very long period of time. People do take account of those factor. What you see in TOD is never the complete picture of the calculation, which is much more complex.

I understand that payback time can be important. What I don't understand is how Robert is proposing to account for this by 'standardizing to an annual basis'. It seems to me that the only way to account for it is to develop models that incorporate the time component as a variable that is separate from EROEI.

(Note also that with renewable energies the time component is related to the EROEI. If a solar array lasts longer and produces more energy, its EROEI will be greater. And the opposite is also true.)

Yvan – Exactly. A real life example: the Eagle Ford Shale play. Let’s assume the average well has an EROEI of 10. The average well pays out in 8 to 14 months yielding a return of 25% over 4 years. Given the high decline rate there's not much cash flow after that. But what’s very critical is the early high cash flow allows that revenue to be recycled to drilling the next well…gotta keep drilling to replace those produced reserves. Now a conventional well in a tite low flow rate reservoir. This well also has an EROEI of 10 and costs the same to drill. Except due to the low flow rate it takes 7 years to pay out and 14 years to produce the same volume of oil as the EFS well.

So both wells have an EROEI of 10…so what’s the difference? The EFS well will get drilled…the conventional well won’t. The second well, even though it eventually produces the same amount of oil it generates an unacceptable low return on the investment. The Net Present Value of the cash flow is insufficient even though it returns an acceptable amount of energy. Or does it really? Would you spend 10 million BTU’s today if it took you 7 years to recover that energy back and another 7 years to recover the gained energy?

So there’s the $64,000 question: would you burn 1 million bo today in order to recover 10 million bo over 14 years? Could you even afford to do it given a high current demand for energy? A simple example: you can invest this year’s paycheck in a deal that will absolutely double your money in 4 years. But if you put all your money into the deal you don’t have enough left to pay your mortgage or kid’s college tuition. OTOH if you could make 50% gain in 6 months you would figure out a way to make it work.

So can someone cook up a Net Present Energy analysis?

"It seems to me that proper EROEI calculations already do this, so I don't understand what you're suggesting."

The classic example for me is oil and ethanol. When we say that oil has an EROEI of 10 or 20, that is on a much different time frame than ethanol's 1.4. The latter takes place over an entire growing season, while the turnover for the former is in days. Those are not equivalent (i.e., in this case the annual energy benefit of oil over ethanol is far greater than 10 to 1).

Robert,

That appears to exaggerate the difference.

Oil also involves time delays between planning, exploration, testing, drilling, production, refining, and delivery.

Furthermore, even if ethanol takes, say, 3 additional months to deliver that's not a big difference: in $-ROI/time value terms, that might reduce the value of the end product by 2%.

I'm not clear why turnover would be meaningful here.

"I'm not clear why turnover would be meaningful here."

Here is an example to satisfy yourself. Process A has an EROEI of 1.2, but delivers that EROEI every day. Process B has an EROEI of 10 delivered on an annual basis. Which process delivers more energy over the course of a year?

The reason time is important is that without it, you can't make broad claims about one process being better than another on the basis of EROEI.

Here is an example to satisfy yourself. Process A has an EROEI of 1.2, but delivers that EROEI every day. Process B has an EROEI of 10 delivered on an annual basis. Which process delivers more energy over the course of a year?

The assumption of your question seems to be that Process A would be done once a day, and Process B once a year. That isn't really an equal comparison. Process B could be done 9 times over the course of each year and deliver more net energy than Process A being done 365 times. Depending on the details, that doesn't necessarily make Process B 'better'. But suppose the energy input for each process could be done in a day. Would you rather work 9 days a year, or 365, for the same pay?

The reason time is important is that without it, you can't make broad claims about one process being better than another on the basis of EROEI.

To be clear, I'm not disagreeing with this statement. I just don't understand how you propose to 'standardize' EROEI on a time basis, or how that would better help us assess the utility of an EROEI number.

The assumption of your question seems to be that Process A would be done once a day, and Process B once a year. That isn't really an equal comparison.

That was what I was specifying. In once case, we have a system that can be turned over quickly, and in another a system that could take an entire growing season -- or longer -- to deliver it's energy return. It's like a financial return. If I say an investment has a return of 50%, it makes a big difference if that's 50% a week or 50% over a year. Yet if you are using EROEI, both cases would have an EROEI of 1.5.

However, if those returns are annualized, you have an apples to apples comparison.

But suppose the energy input for each process could be done in a day.

You are specifying a different problem. EROEI doesn't tell you that, which is why it needs to be standardized to a time frame. Look at the graph in the post and ask yourself over what time frame each of those EROEIs is achieved.

However, if those returns are annualized, you have an apples to apples comparison.

Arguably you don't. Arguably the apples to apples comparison would involve the same amount of energy investment over the same amount of time.

You are specifying a different problem. EROEI doesn't tell you that, which is why it needs to be standardized to a time frame.

I get what the problem is. I don't get how your proposal is a solution. If you 'standardize' to a time frame, the time frame is arbitrary. For example, solar looks horrible compared to ethanol if the time frame is one year, but ethanol looks bad compared to solar if the time frame is 30 years. And not only that, but after one month, solar looks better than ethanol (even if both look horrible).

Also, the piece of the time frame you choose to look at is arbitrary. For solar, EROEI in the first year might be 1/6. In the second year it would be 1/0, or infinite. The only way to reasonably 'standardize' this to a year is to average all years, in which case you just end up back at the EROEI.

If you 'standardize' to a time frame, the time frame is arbitrary.

It doesn't matter, as long as the time frame is the same for the sources you are comparing. I think an annualized EROEI would work just fine though.

It doesn't matter,

It does matter, because given certain time aspects of the sources, the energy return comparison may come out different depending on the time frame you pick. (Did my last post not give a good enough example of that?)

I think what Robert is saying would make more sense if he said 'net-energy-over-time is a more meaningful figure than EROEI'. I have been struggling to understand what he means by 'standardizing EROEI to a year'. I think he really means to say 'annualized net energy', not 'annualized EROEI'.

To more or less repeat what I've said elsewhere, scalability is only 'more important' than EROEI if you take no interest in anything besides total net energy. With respect to total energy flows (which is important, ecologically, socially, economically) it is not meaningful to say that one is more important than the other.

I think it's fair to say that sufficiency of volume has always been the first priority for customers of energy systems. It has come well before ecological and social concerns.

That was the case for hunter-gathers; wood burners; and it will be the case for oil, gas & coal users, if they have to make a choice.

Fortunately, they don't really have to make that choice: renewables are affordable, scalable, high-E-ROI, etc, etc.

I'd say the energy delivered depends on the scale of the resource (in the ground, or whatever), not the input or the E-ROI.

Now, if you can increase oil production because the E-ROI is higher, that's different - but I don't see how that would work.

I'd say the energy delivered depends on the scale of the resource (in the ground, or whatever), not the E-ROI.

Would you still say that's true if the EROEI is less than 1? I think you need to guard against taking your point too far.

I agree.

Still, Robert is willing to go down to an E-ROI of 1.1 in his example of a theoretically viable process...

I'd say the energy delivered depends on the scale of the resource (in the ground, or whatever), not the input or the E-ROI.

For oil, but not for everything. Assume I have a photosynthetic microbe process that can turn over in a couple of days. I want to compare that to corn ethanol. If I am going to use and EROEI approach, I must have those on equivalent time frames to calculate actual returns. In this case, if my microbe process has an EROEI of 1.1 and ethanol is 1.4, the microbe process is going to be much better from an energy return point of view. If we annualized the returns, that would be clear.

If the process can be turned over every two days (versus once per year )it increases the effective output/energy return . That reduces your amortized capital (tanks, buildings, etc) and labor costs per unit of energy output. So, you're really saying what I've said elsewhere: scalability is more important than E-ROI, as long as E-ROI is positive.

You're also saying that with sufficient scale other costs may be more important than energy costs/E-ROI (again, with positive E-ROI).

Of course, greater turnover also increases the input proportionately, so the E-ROI stays the same. You could just count the input at the beginning of the first cycle, if you wanted to, but I don't think you should - each cycle starts new. Heck, almost any process could be considered to have arbitrarily high E-ROI if you only count the first cycle. I'd say counting the input once over a period of a year would be mighty arbitrary - if you want to do it that way, you'd probable need to use the life of the equipment, or installation (and E-ROI would be even higher).

So: faster turnover will dramatically improve net energy return, $-ROI and overall value of the process, but it won't affect E-ROI.

Robert - there are some further factors worth considering with regard to your oil to corn ethanol observations.

First, the ability to store the organic feedstock appears to nullify the otherwise critical turnover-time factor, in favour of the relevance of $-RIO & E-ROI.

Next, using the example of wood as a potentially far more sustainable feedstock than corn (grown by the absurd re-dedication of farmland), the dollar and energy returns will decline with the scale of the processing facility due to transport and process costs and inputs, with only limited gain from the process's economy of scale.

For example, the wood I cut on the farm for my own use and burn in an 80% efficient stove takes less than 1kg of petrol (~18kwh) to cut and haul the 1.6 tonnes (1 tonne after barn drying, with ~4900kwh potential) that yields about 3920kwh in hot water and space heat. An E-ROI of about 218 is thus achieved.

At the other end of the scale spectrum there is both a heavy outlay of diesel on long-distance haulage of a fuel with low enegy density, and on chipping to homogenize the wood for processing, and on handling, conveying to hoppers, etc. I lack data on quite how much poorer is the E-ROI of large industrial biomass plants, but FWIW I know of scientific research by the UK Forestry Comn stating that coppice wood needed to be grown within a few miles of the facility to be economically viable for power generation under 1980's energy prices.

This implies that processing wood and perhaps other sustainable energy harvests may actually lack scaleability in the conventional sense, but be entirely viable at say village scale with a local feedstock supply, with plants' profitability and E-ROI resting on the under-appreciated economies of sustainable replication, and not those of monopolistic scale.

Regards,

Backstop

To understand how EROI influences the flow of net energy, we must first look at the equation for both net energy and EROI, which are:

Net Energy = Eout – Ein
EROI = Eout/Ein

If we solve the EROI equation for Ein and substitute it into the Net Energy equation, we get:
Net Energy = Eout*((EROI-1)/EROI)

An easier equation to use for EN is: EN = (1-1/ERoEI)*EG; where EG equals the above Eout or the potential energy of one unit (140,000 BTU/gal for oil). ERoEI is expressed as a real number; that is, 20 for 20:1. ERoEI can then be expressed as ERoEI = EG/EG-EN or ERoEI = 1/(1-EN/EG). Since the set of equations that define the ERoEI curve are non-deterministic, this form of mathematical structure greatly facilities the numerical analysis needed to solve the function.

I'm not so sure about the method they used to come up with 8:1, but they are very close. Our study has determined that the world' oil reserve will be completely depleted when its average ERoEI hits 7.46:1, at $407 per barrel; (margin of error +- 5%). That will be the point where the higher ERoEI fields can no longer produce enough energy to subsidize the lower ERoEI fields. It is also be the point where the end consumer must invest 140,000 BTU of energy to produce the work needed to exchange for a gallon of oil product.

Have you taken account of the societal energy cost of the energy production? This can be calculated using the EOI-LCA technique. If you do it, the EROI will drop and will end up with a critical price below 407$. My educated guess: 200$/bbl.

Actually, the critical price is about $80/bbl. At that point substitutes (EVs, etc) start to be cheaper. At that price oil will become obsolete in perhaps 20 years, as R&D is expanded, capex barriers to entry are surmounted, and economies of scale start to really take effect.

Of course, there are substantial delays/lags in implementing the substitutes, so I'd say the effective cap is around $125 - that's the price where low-cost but inferior substitutes like carpooling, online shopping, and e-bikes start to become attractive to more and more people.

FOR ALL

Seems like there's still a disagreement as to the importance of time in EROEI analysis. To be honest I haven’t analyzed the math presented here…seems too convoluted IMHO.

But again two projects both with an EROEI of 15. One recovers the energy in 3 years and the second in 12 years. Do both projects have the same value to society? If you say yes then time isn’t a factor.

Back to the point Robert made earlier. Net Present Value is a valid method to compare investments that generate revenue over different time frames. Two projects invest $1 million. One has a NPV of $2 million and the second $4 million. In the oil patch NPV isn’t always the determining factor...but usually weighs heavily. Risk profile might favor the lower NPV.

Assume 2 projects: “A” has an EROEI of 15 and “B” an EROEI of 10. But A has a NPV less than B. Which well gets drilled? What if B pays out much quicker than A? That appears to be a driving factor in some fractured shale plays: the EROEI is not very impressive nor the NPV as high as a conventional well (like Rockman drills). But the public oils make their profit by increasing stock equity. And for every conventional well left to drill there are dozens, maybe many dozens, of unconventioal wells to drill. And that profit motivation shows up in no EROEI evaluation I’ve seen yet. Add the quick recovery of capex for these wells doesn’t show up either. But that cash flow allows more wells to be drilled. But as I’ve detailed before these factors are driving many of the shale plays…and not EROEI. But that shouldn’t be a surprise: the oil patch has never used to make drilling decisions. And isn’t used today.

I understand you point. I think in that case power output would be a better metric. In Quebec, we have very large hydroelectric powerplants. They are awfully expensive to build but they have an extremely long lifetime. Over a long period EROI is very good, but not much impressive on short term. Same for cash flow. But almost all hydro-power in Canada is government owned.

Off course, any economist will tell you that we discard the future. That's probably the most important factor.

Yvan - And that's exactly THE critical point I've been trying to express. Your example is perfect. Any company or investment group evaluating a hydroproject might reject it because of its low ROR and/or pay out. Essentially why we have utility commissions: to set profitable return rates for investors in such projects so they'll pay for the project. Or have it funded entirely by the govt which does't require any profit. The govt may not profit but hopefully the public will. OTOH how many such project didn't return a net benefit to the govt or public?

And there's THE problem: govt investment will be necessary to ramp up the alts quickly IMHO. But that also requires the govt to pick potential winners. And that's where my confidence falls flat. The politics involved in such a process would likely kill much of the potential.

OTOH how many such project didn't return a net benefit to the govt or public?

Old hydroelectric project in Quebec made a lot of money because they were very cheap. This is no longer the case now, since new project are less profitable. This is where to politics comes into play. Somehow, the government must keep the dam building industry alive.

Government are useful toll of societal reorganization. Unfortunately, they are a burden if they try to protect the business as usual. This is why I doing some politics within the Quebec Green Party. This is mostly a psychological Band Aid but I will be able to stare my sons in the eyes in the future.

But again two projects both with an EROEI of 15. One recovers the energy in 3 years and the second in 12 years. Do both projects have the same value to society? If you say yes then time isn’t a factor.

The more interesting case is when the longer process has a higher EROEI. At certain numbers, they deliver equal net energy over time. The question is whether this amount of time is within a scale that humans require to benefit from investments. That gets pretty fuzzy.

I certainly don't know a lot about NPV analysis, but isn't the 'discount rate' a key variable? And isn't the choice of a discount rate somewhat arbitrary? I think NPV analysis would be helpful here, but there would still be a range of results depending on the discount rates chosen. At least we might be able to say what the maximum time scale difference is for two different EROEIs, within which they would possibly be economically competitive.

jag – A very valid point. There’s a whole range of permutations. That’s what NPV analysis is for: to equalize the value. The discount factor is required to put time factor into play. At a DR of 10% $1 produced next year is worth $0.90. A $1 the next year is worth $0.80. Etc, etc. Add the values of all the years of discounted cash flow and you have NPV. Not really arbitrary but it can be. But changing the DR will change the NPV as a function of time: the higher the DR the less value future quantities will display. A 1% DR will change future vales very little. A 50% DR will make values beyond just a few years meaningless. One way to look at DR’s is a form of reverse interest. If a project has a NPV of $100 at DR of 10% then it would represent investing $100 at 10% interest for the same time period.

So that adds a very interesting dimension to Net Present Energy. FF produced 10 years from now could be of greater benefit/value then today. That’s a key factor in calculating NPV of a producing well. One has to assume a future price platform for the production stream. If I inflated the FF at 10% per annum and used a 10% DR they would essentially offset each other and NPV would be the cumulative cash flow. So how do we inflate the energy content? We can do it monetarily: a BTU today is worth $X and worth $3X in 10 years. But how do you do it in energy units?

Again: why the oil patch doesn’t use EROEI.

Discount is not normally taken into account for EROI analysis. But, I have seen in a few instance that climate change must be taken into account in building retrofit. Over 50 years, the effect is strong enough to increase or reduce the HVAC load by a lartge factor, which mess up the EROI. Actually, this is a problem with any LCA analysis involving infrastructure because over their lifetime energy basket is changing as many other industrial factor.

Yvan - "Discount is not normally taken into account for EROI analysis". But won't that kill the utility of EROEI in some cases? Without a discount rate a revenue $(or a BTU) produced in Year 10 has the same value as produced in Year 1. In the analysis of a drilling prospect using the typical 10% DR revenue produced in Year 10 adds almost nothing to the rate of return/NPV.

The "lifetime energy basket" point may explain why EROEI of drilling prospects is of little value but useful in other projects. The value of a well, under constant pricing, declines every day it produces. OTOH, a dam produces a more static volume of energy over time. That doesn't change a diminished NPV of its income over time. But if it's generating $X of energy in Year 10 as it did in Year 1 you have a very different dynamic from a oil well that is not only depleted in Year 10 but also represents a plugging liability...IOW a negative value.

There is one oil patch situation similar to the dam: low flow rate/high reserve. Typically stripper wells. I've analyzed such fields that had almost no decline rate: currently making X bopd...in 20 years it would be making about the same. But NPV says it was worth $Y and would pay out in 5 years. But if you did the NPV of the field in Year 6 it still had a NPV of $Y. I dubbed these fields as constant NPV. Imagine trying to buy such a field from a small operator: you can only offer $Y for the property but the operator could keep it, make $Y over the next 5 years and then still sell it to you for $Y (or more if oil prices have risen). Almost impossible to negotiate. Been there and banged my head against that wall. LOL

I would imagine buying an existing hydro or solar project solar would face the same dilemma. The new owner has to make a ROR X. But unless the current owner has to liquidate why would they sell thir "cash cow"?

Yvan Dutil said:

Have you taken account of the societal energy cost of the energy production? This can be calculated using the EOI-LCA technique. If you do it, the EROI will drop and will end up with a critical price below 407$. My educated guess: 200$/bbl.

Thanks for the input. My concern with that approach is two fold: 1) the input-output tables of the EOI-LCA technique are based on a US data set, and oil is a globally used commodity with a global set of inter-dependencies. 2) The time frame between $200 and $400 oil will be less than a decade. It is possible that during that period the world will be pushed into a state of chaotic disassociation. To analyze that kind of situation, there is no historical precedent to rely upon.

My WAG, however, is that $200/bbl oil is probably pretty close to the mark. $200/bbl oil will put transportation fuels at about $8.50/gal; enough to begin seriously affecting global trade and the world economy in general. $407 is merely a point where the curves from two energy functions cross; it does not reflect any societal impact of the event. It is a point that the laws of physics tells us we can not exceed.