[shortonsense]

I'll reply to the comments in an aggregated form again:

I would venture that your equation for how fast the supply of a resource is coming on line is flawed, and this flaw is correlated with and only specific to a small subset of the actual resource base available for development in the future. It is similar to the "field growth" problem in that if you make the assumption that only what you see is available for development, you massively UNDERestimate how much resource will be converted into reserves and available for use in the future. Laherrere's estimates are very much a reserve estimate rather than a resource estimate, and would lead any model directly into this particular trap. The original modelers in Limits to Growth stumbled into this issue as well, if I recall correctly.

Actually, that isn't an issue in my model or in the Limits to Growth model. One thing that is easy to test is what happens assuming that there are more fossil fuels available. I said before that I tested for this and you get something similar happening, only a bit later. In the original Limits to Growth model the same kind of test was done by duplicating the amount of nonrenewable resources, and exactly the same thing happens.

Actually, I would say there is an issue. For example, I applied your equation to the basic numbers Hubbert provided in his 1956 paper "Nuclear Energy and the Fossil Fuels", in particular, Fig 21 for starters. Your equation, correct me if I'm wrong, is Production Increase = 0.2 *( Fossil Fuel Percent Remaining - 0.5 ) * Current Production.

Applied to Hubberts high estimate in 1956, that solves as approximately

Prod Increase = 0.2 ( (1-(52.4/200) - 0.5 ) * 2.5 = 0.119 , or, 4.7% increase in that particular year. According to the EIA, that particular year was a 2.6% increase, and the AVERAGE from that point in time to peak production in the US was 2.4%.

But lets not forget, Hubbert was using undiscovered resources in his number...the Laherrere numbers you use equate more to reserves rather than resources.

Again using Hubberts Fig 21 numbers and reserves instead of Hubberts resource numbers:

0.2*(( 1 - ( 52.4 / 82.4 ) - 0.5 ) * 2.5BBO/Yr = -0.07BBO/Yr, a decrease in production of approx 3%.

This is exactly what I said, that a low estimate of the resource base will cause your equation to come unglued, when we know from this particular example that no such profile emerged for another 15 years, and your equation has already called peak based solely on low resource estimates.

How does this equation work in a known decline regime? Using EIA data from 1985 to 1990 and your equation, again on reserve numbers rather than an actual resource base ( something Kawata and Fujita did account for ) it solves as:

0.2*(1- (145/171 ) - 0.5 ) * 3.16BBO/Yr = -0.21 BBO/Yr or 6.6% drop. Actual EIA in that year says it dropped 3.2% and the 5 year running average was approx. 3.1%

So...in one of the few environments where Hubberts peak concepts have actually been reasonable, your equation calls peak and decline early on the upside solely because of a pessimistic resource base estimate, at least in this example, and on the downside overestimates the drop by 2X.

This is why I said there was a problem, here is a systemic flaw which in this example fails in a unidirectional manner, particularly when used in conjunction with large underestimates of resources, which Laherrere certainly uses.

Did you test this equation in any systematic way against a full range of production increases, peaks and production decreases to make certain it mimicked something, somewhere, sometime? Copper, whale oil, coal, steel, wood, uranium?

Your equation is nearly guaranteed to force transition of some sort earlier than it should if you stick to reserves instead of resources, at least in oil and gas estimates. And if this equation doesn't work in about the only place where Hubberts prediction has worked out reasonably well, how much confidence can you have in it mimicking correctly something he missed badly, such as natural gas production in the US, oil at the global level?

With different and more reasonable non renewable resource estimates in the model, would you be willing to bet that "final decline being more moderate" could easily turn into "no decline for the foreseeable future"?

Actually, I'd be happy to bet the opposite. Increasing the estimate of nonrenewable energy sources is not actually improving things in the long term, it tends to make things worse because it increases climate change and makes the transition to renewables harder.

So you equate climate change to more difficulty in transitioning to renewables...because of EROEI? The wind blowing harder because of more energy in the atmosphere makes electricity from wind power more difficult? I don't understand the link here I guess. You don't take climate change into account to drop your per capita heating, but the model does take into account a windier planet...blowing over windmills instead of spinning them harder?

Also, once you open up the non renewable resource base a little bit, you'll allow more CO2 output which might provide another 2 or 3C temperature increase, +6C has got to cause less heating needs somewhere.

According to the book "Six degrees", if we reach that point, the lowering of our heating needs won't be the most pressing of our concerns.

Stopping another Azolla event from crashing our planet back into another Ice Age would definitely take precedence on the list of concerns to humanity, I certainly understand that one!

One of the reasons I used World3 was because it's well known, so I hope that there will be many more people that are familiar with at least parts of the model and can detect any major flaws.

You've got me interested in this World3 thing now, particularly since Shanny mentioned its open source, although I find debugging a programmers logic or code distasteful.

[WebHubbleTelescope]

The concept of dispersive discovery explains the shape of the classical Hubbert curve. No use bringing in heuristics to the discussion. Commenters like shortonsense jump on it because a plain heuristic doesn't derive from any fundamental understanding. So unless you can back it up with some physical and/or statistical model, good luck trying to defend your empiricism.

[shortonsense]

The concept of dispersive discovery explains the shape of the classical Hubbert curve. No use bringing in heuristics to the discussion.

I think her model requires something which does more than just mimic the classic Hubbert curve however. She is using this thing for all non renewables, which means she needs something with quite a bit of flexibility in it.

Commenters like shortonsense jump on it because a plain heuristic doesn't derive from any fundamental understanding. So unless you can back it up with some physical and/or statistical model, good luck trying to defend your empiricism.

Maybe she's got one? When I questioned it before she seemed to think it wasn't an issue.

[yesplease]

Assuming we had tremendous FF based EROEIs and are transitioning to much lower renewable EROEIs would probably induce a relatively inaccurate model compared to using oil at an EROEI of ~5:1 or lower, and then transitioning to renewables with EROEIs in the ~10-30:1 range.

All the studies I've seen comparing EROEIs of different sources agree that fossil fuels generally have higher EROEI than renewables, or at least used to have.

Only if we selectively look at the EROEI of certain steps and don't mention that we're doing so, ala some earlier posts I've seen on TOD, as opposed to the final EROEI of the delivered products. As you can see from my post here, the references of 100+:1, 30:1, and so on, EROEI, refer to the EROEI of discovery and production respectively. Conveniently, whenever I see that info posted on a PO site, the OP leaves that info out. If we do a bit more digging regarding the energetic refined products we get from a bbl of oil as well as the energy input needed for refining we can see that even if oil's EROEI in the early 1900s was a billion to one, the energy needed for refining compared to the energy we get out caps this at ~6:1.

The problem clearly isn't the work in the papers I cited in my linked posts, but the use of that work as a all in one EROEI figure for oil, even though it's likely been around ~5:1 for as long as we've used it, at least in the context of refining it into fuels. Of course that still doesn't include the exergy of that energy, how much of it is available for useful work given the application. For instance electricity from solar panels has much higher exergy than gasoline from a refinery due to the limits of the Carnot cycle versus electric motor efficiency. Realistically speaking, a kWh of electricity will take an EV about three times farther than a kWh of gasoline will take a conventional vehicle, so that would also throw things out of wack, although that depends on the application and is a pain to quantify.

Coal too had a great best case via Cutler at the mine mouth of 80:1 in the past. The problem is when we toss in the ~1 billion kWh of diesel fuel needed to transport every ~36 billion kWh of coal, as well as the processing costs of it, we probably were and are at best at ~15:1 for coal before we even burn it. Unfortunately, doing this reduces the amount of energy to about a third of what it was, so we're looking at a ~5:1 EROEI for electricity from coal.

Overall, we seem to have two problems, using an inaccurate figure for the EROEI of oil's energetic products, and ignoring the difference in efficiency of use of chemical energy from FFs compared to electricity from renewables, although this is more of a pain to quantify because it depends on application. I doubt a model will be accurate given these two, as well as the others I mentioned below.

Certainly we wouldn't have started producing electricity from coal, rather than wind, if coal wasn't easier. My model assumes that the transition to renewables happens when the EROEI of renewables is comparable to the EROEI of fossil fuels.

Coal was easier at one time, although that isn't due to EROEI per say, as opposed to availability, cost, and so on. In terms of a transition to renewables, that's happening as costs, both the usual internalized stuff we're used to and the externalized stuff we're starting to include, drop below the cost of FFs.

In terms of the World3 model, a huge problem is no cost substitution and another is no recycling. No cost substitution for instance is horrendously inaccurate at modeling demand. We aren't going to use electricity from wind turbines for smelting iron ore, just for replacing electricity from coal. EVs aren't going to use as much energy as conventional vehicles do because a kWh of electricity can do more work than a kWh of gasoline in that context, just like we won't use as much because the energy storage limitations of batteries require smaller vehicles, which means even less energy consumption. No recycling is obviously a huge problem given how much we do it today depending on the app.

[kiwichick]

re increasing effiency of renewables

www.bluglass.com.au

claims of possible efficency up to 50% conversion sunlight to power

huge!!!??????

[Doly]

My replies:

Your equation, correct me if I'm wrong, is Production Increase = 0.2 *( Fossil Fuel Percent Remaining - 0.5 ) * Current Production.

That's where you got confused. That equation is the maximum production increase, not the actual production increase. The actual production increase is determined by demand, and only if demand exceeds the possible maximum, production is calculated using that equation. The peak happens when the maximum production increase goes negative.

So you equate climate change to more difficulty in transitioning to renewables...because of EROEI? The wind blowing harder because of more energy in the atmosphere makes electricity from wind power more difficult?

No, it's because climate change causes problems in food production, and humans have this tendency to think that food is a more pressing concern than anything else, including development of renewable energy.

Coal was easier at one time, although that isn't due to EROEI per say, as opposed to availability, cost, and so on.

And availability and cost have nothing to do with EROEI? If something is less available, presumably it takes more energy to get it? And cost is related to many things, but mostly to the cost of the equipment needed to extract that energy, and the cost of the equipment often is correlated to the energy needed to build it.

[shortonsense]

My replies:

Your equation, correct me if I'm wrong, is Production Increase = 0.2 *( Fossil Fuel Percent Remaining - 0.5 ) * Current Production.

That's where you got confused. That equation is the maximum production increase, not the actual production increase. The actual production increase is determined by demand, and only if demand exceeds the possible maximum, production is calculated using that equation. The peak happens when the maximum production increase goes negative.

Unfortunately, that just makes it all worse. Whereas your equation says the maximum increase in production is only 4% in 1956, and it was increasing 2%/year for 15 years. Using even Hubberts numbers in this scenario ( much, much higher than an equivalent Laherrere estimate ), it would have called peak approximately 13 years early. Using a Laherrere equivalent it would have called for peak on the spot.

The important question however is, how well did testing of this equation do against reality in some other commodity? If you are basing maximum production rates on it, or calls for peak on it, perhaps it works for coal, natural gas, uranium, etc etc?

And this doesn't address the insufficiency of Laherrere's numbers in this scheme either,

So you equate climate change to more difficulty in transitioning to renewables...because of EROEI? The wind blowing harder because of more energy in the atmosphere makes electricity from wind power more difficult?

No, it's because climate change causes problems in food production, and humans have this tendency to think that food is a more pressing concern than anything else, including development of renewable energy.

Well, its certainly an assumption that climate change causes all sorts of problems, but considering how well mankind thrived during the last large scale climate warming episode, it doesn't strike me as a quid pro quo that possible climate change = definite food production problems = the conclusion starving people can't build windmills and solar panels. This is where some basic correlation would come in real handy...otherwise the entire model is dependent upon what can easily be a faulty assumption with no allowance for what strikes me as the obvious...no such causal links between climate change and inability to build out renewable infrastructure exists. For example, can you point to a single example where the past 100 years of climate change ( assuming that it even exists of course, not an assumption I would naturally make myself ) has caused a single solar panel factory to close because of the mass starvation of its workers, specifically related to climate change? Or a windmill factory?

Coal was easier at one time, although that isn't due to EROEI per say, as opposed to availability, cost, and so on.

And availability and cost have nothing to do with EROEI?

Certainly cost has nothing to do with an EROEI argument at its most basic level. Which is why the EROEI calculations used in any debate which comingles with GDP or economics or peak something or resource depletion generally doesn't work and should be avoided. Its not the primary consideration when deciding to develop a resource, it isn't a measure for why you continue to use a given resource after development begins, and its change through time to a lower value doesn't dictate when the use of a given resource should stop.

EROEI by itself is sort of an interesting number when comparing two different resources, but by itself it is pretty irrelevant to anything related to the use, development or substitution between those two resources.

[yesplease]

Coal was easier at one time, although that isn't due to EROEI per say, as opposed to availability, cost, and so on.

And availability and cost have nothing to do with EROEI?

In the case of coal versus renewables, it's mostly available tech, production level, and it's cost. Not that EROEI doesn't factor in, just that it tends to take the back seat when contributing to cost compared to infrastructure and labor costs. For instance natural gas, by far the largest energy source used in oil extraction/refining in the states, only contributes a percent or two to the cost of gasoline and other refined products. As for the market basing it's use off of EROEI, that's incorrect AFAIK. If it was, we would've been using renewables in bulk a long time ago. As it stands, a shortfall between production and consumption can make renewables favorable in terms of energy costs, w/ no change in EROEI from any resource.

If something is less available, presumably it takes more energy to get it?

That depends... If demand surges above supply, price will rise a lot, but the resource doesn't take any more or less energy to get, so right there we see a change in availability w/ no change in energy required to get something.

And cost is related to many things, but mostly to the cost of the equipment needed to extract that energy, and the cost of the equipment often is correlated to the energy needed to build it.

I wish! Cost in terms of manufacturing is mostly related to volume/availability. If we cranked out oil platforms like cars, cost per ton given the material inputs would probably be comparable, but since we don't, specialized oil equipment is expensive and commonly used stuff is relatively cheap. Compare the cost of a custom build car to a mass produced car, and you'll see how much more people pay for specialized equipment.

[Doly]

My replies:

Unfortunately, that just makes it all worse. Whereas your equation says the maximum increase in production is only 4% in 1956, and it was increasing 2%/year for 15 years.

And where is the problem? 4% is less than 2%

The important question however is, how well did testing of this equation do against reality in some other commodity? If you are basing maximum production rates on it, or calls for peak on it, perhaps it works for coal, natural gas, uranium, etc etc?

I have checked that the numbers for production of all fossil fuels are approximately correct. I'm less sure about uranium because I couldn't find very reliable data for that.

And this doesn't address the insufficiency of Laherrere's numbers in this scheme either

Like I said before, I have checked with higher reserves numbers for fossil fuels and the same kind of thing still happens, only a bit later.

Well, its certainly an assumption that climate change causes all sorts of problems, but considering how well mankind thrived during the last large scale climate warming episode, it doesn't strike me as a quid pro quo that possible climate change = definite food production problems = the conclusion starving people can't build windmills and solar panels

It's based on available data on what climate change has already done to crops. Climate change has already reduced crop yield, that was part of the food crisis we had last year. The study can be found here:

http://www.iop.org/EJ/article/1748-9326 ... 14002.html

This is where some basic correlation would come in real handy...otherwise the entire model is dependent upon what can easily be a faulty assumption with no allowance for what strikes me as the obvious...no such causal links between climate change and inability to build out renewable infrastructure exists.

I never said that the link is direct. It's a link with several intermediate steps, and you can take issue with any of them, if you like. The chain of links is: carbon dioxide>land yield>food>food per capita>fraction of industrial output allocated to agriculture>fraction of industrial output allocated to investment

For example, can you point to a single example where the past 100 years of climate change ( assuming that it even exists of course, not an assumption I would naturally make myself ) has caused a single solar panel factory to close because of the mass starvation of its workers, specifically related to climate change? Or a windmill factory?

That reminds me of the argument that people had two years ago about how impossible it was to have another Great Depression, because we didn't have any in the last 50 years.

There are plenty of historical examples of manufacturing being partially abandoned in times of food scarcity, not by mass starvation of the workers, but because workers and resources are very quickly redirected towards production of food when food is scarce.

That depends... If demand surges above supply, price will rise a lot, but the resource doesn't take any more or less energy to get, so right there we see a change in availability w/ no change in energy required to get something.

That is correct, but it doesn't explain which energy source will be chosen to substitute when one of them starts being a problem because supply can't meet demand. My model uses EROEI to explain allocation.

If we cranked out oil platforms like cars, cost per ton given the material inputs would probably be comparable, but since we don't, specialized oil equipment is expensive and commonly used stuff is relatively cheap. Compare the cost of a custom build car to a mass produced car, and you'll see how much more people pay for specialized equipment.

As long as specialized equipment remains specialized, I don't see that is a problem. I think there is still a correlation between energy needed to produce it and cost.

[yesplease]

That is correct, but it doesn't explain which energy source will be chosen to substitute when one of them starts being a problem because supply can't meet demand. My model uses EROEI to explain allocation.

We can't ignore price in terms of resource use because that's what drives it, not just EROEI. EROEI drives price to an extent, depending on the energy source, but it's just part of the costs that determine what energy source we use.

If we cranked out oil platforms like cars, cost per ton given the material inputs would probably be comparable, but since we don't, specialized oil equipment is expensive and commonly used stuff is relatively cheap. Compare the cost of a custom build car to a mass produced car, and you'll see how much more people pay for specialized equipment.

As long as specialized equipment remains specialized, I don't see that is a problem. I think there is still a correlation between energy needed to produce it and cost.

I don't think a correlation between energy and cost would be easy or accurate for low volume industrial equipment. The additional cost tends to be associated with more specialized man hours, not more energy per say. I have a cousin who is a mud engineer, and he doesn't make six figures a year because he uses an order of magnitude or two more energy than other workers do, but because his specialized experience commands a premium. If anything, adding the embodied energy in the materials to the industry average for energy costs during construction would be the way to go.

[Quinny]

and how would you measure price - in dollars?

That is correct, but it doesn't explain which energy source will be chosen to substitute when one of them starts being a problem because supply can't meet demand. My model uses EROEI to explain allocation.

We can't ignore price in terms of resource use because that's what drives it, not just EROEI. EROEI drives price to an extent, depending on the energy source, but it's just part of the costs that determine what energy source we use.

If we cranked out oil platforms like cars, cost per ton given the material inputs would probably be comparable, but since we don't, specialized oil equipment is expensive and commonly used stuff is relatively cheap. Compare the cost of a custom build car to a mass produced car, and you'll see how much more people pay for specialized equipment.

As long as specialized equipment remains specialized, I don't see that is a problem. I think there is still a correlation between energy needed to produce it and cost.

I don't think a correlation between energy and cost would be easy or accurate for low volume industrial equipment. The additional cost tends to be associated with more specialized man hours, not more energy per say. I have a cousin who is a mud engineer, and he doesn't make six figures a year because he uses an order of magnitude or two more energy than other workers do, but because his specialized experience commands a premium. If anything, adding the embodied energy in the materials to the industry average for energy costs during construction would be the way to go.

[yesplease]

Dollars, Euros, Yen, whatever. We have currency indexes so we can see how a currency is doing against a weighted basket of other currencies as well as direct exchange rates.

[Quinny]

It seems to me that SOS an YP jsut don't accept the concept of EROEI and keep harping back to what is IMO a discredited financial system.

I first called for Energy based accounting back in the 70's when efficient pit's were being closed because of a fall of the Polish currency against the pound. (This was despite much worse productivity and environmental degradation in the Polish pits.)

When someone puts forward a model which then gets criticised because of the lack of a link (or clarity) between EROEI and fiat money I can only assume that those doing the crticism just don't accept any of the fundamentals of PO.

The thing is I'd love it if they could disprove PO theory because I'd feel much happier for my kids future. Unfortunately they just try to 'nit-pick' holes in other peoples contributions. I suppose they should be given credit for the volume they write, unfortunately I find much of it based on worship of a failed system. After a while it starts to get on your nerves.

[yesplease]

I think you should reread my posts Quinny. I've talked about EROEI in a fair bit of detail on the previous page. That said, EROEI is not the only factor in terms of cost. We have energy inputs and their costs, human inputs and their costs, material inputs and their costs, as well as taxes, subsidies, and so on, all of which work together to determine cost. I don't think we should ignore EROEI, but it's clearly not the only thing considered when we choose what energy sources we use, and a model IMO should reflect this reality. Whether or not you consider the structure of our costs fair, for instance should there be more taxation to account for externalized costs and so on, is valid, but if we're trying to accurately model something we need to look at cost, of which EROEI is just one part, not just EROEI, in terms of energy usage.

[Quinny]

I don't agree. IMO the problem with using any currency in a global model is that the basic flaws in a debt based money creation system are carried into the model. Doly is correct in trying to eliminate cost factors and focus on EROEI.

[yesplease]

The world uses money as a basis economic decisions, including energy use. Regardless of whether or not you agree personally with our monetary system, ignoring it when trying to accurately model a system won't result in an accurate model. Should EROEI be included in an accurate model? Sure, it's part of costs. Should it be the only thing we use? No way. We need material costs, labor costs, and energy costs, which includes EROEI, but are not completely dependent on EROEI.

[Quinny]

You obviously see Energy as just another resource. We have to agree to disagree, I think it's the trump card that underpins all other inputs.

I'd accept that there could be other extremely scarce resources who's cost is not reflected by the energy used to get them, and hence they could be used as constraining factors. They are however IMO not on the critical path as it were.

There is no point in using 'material or labour costs' when they can fluctuate with no recognition of the effort required to extraxt, or the amount of labour required. Like the highly efficient English pits being closed (and to all intents and purposes sealed in perpetuity) in favour of inefficient Polish pits because of 'costs' that had nothing to do with effectiveness or best use of resources.

[yesplease]

So where's that leave matter?

Anyway, regardless of what you believe, people don't behave based solely on what nets the most energy, and as such any attempt to model behavior on that probably won't work well. The point of using market costs of labor, materials, and energy used to get other energy is that it's accurate because that's what we base energy extraction on. EROEI is a part of it, but it's not the end all be all. People and materials are still involved and the amount of time/scarcity of ability also determines cost and what energy sources we use.

[Quinny]

Although you've tried hard to pick holes in Doly's model, I am confident that any model that tries to introduce more financial costs into it would be much easier to criticise.

Real wealth is represented by goods that people can utilise in some way and food they can eat etc. produced by the application of energy to raw materials. You don't seem to accept this and assume that wealth can be measured by abstract concepts such as GDP. Following the collapse in value of major industrial concerns and financial institutions can you really not see the danger of introducing 'funny money' into a model?

[yesplease]

The only problem with doly's addition to the world3 model was using EROEI solely for energy use when that isn't case in practice. As for whether or not a model based on what energy used is based on, cost, with it's components including EROEI, is easier to criticize, that depends on the changes in the model. It would certainly be more complex, but that doesn't mean it would be more or less accurate than the previous iteration. It depends. The only way to find out would be to try then wait.

Real wealth is represented by goods that people can utilise in some way and food they can eat etc. produced by the application of energy to raw materials. You don't seem to accept this and assume that wealth can be measured by abstract concepts such as GDP.

GDP includes goods that people can utilize as well as food they can eat and raw materials used in their production. It also includes services, what people do and how much they get paid to do them, because without those, we can't utilize raw materials to make goods and foods.

Following the collapse in value of major industrial concerns and financial institutions can you really not see the danger of introducing 'funny money' into a model?

The collapse of the different financial institutions has nothing to do with funny money and everything to do with fraud. Lenders made incredibly risky loans, which were labeled as totally safe investments, and the value of these institutions went up until it became obvious that they were sitting on a pile of lead rather than a pile of gold, at which point they fell like dominoes.