Doly wrote:I suppose this is the right place to mention the model I just got published in The Oil Drum, which takes the classic World3 "Limits to growth" model to the world of climate change, peak oil and EROEI. Hope you enjoy it:
http://europe.theoildrum.com/node/5145
Iaato wrote:From the graphs, it looks like you used Stella?
I don't understand why you feel that non-renewable resources such as minerals can be deleted as not important?
What happens if you leave demand out of the model? Is demand essential to include?
Doesn't demand follow supply, and won't we essentially use supply as fast as we can?
Since there is no way to have more than matching supply and demand, why bother?
Might the inclusion of demand create errors by artificially improving supply?
What causes the blip in descent of industrial output, and why is the decline so prolonged?
Doly wrote:Of course you may ask questions! That was the point of posting here. I'm not a regular poster as I used to be (not working in a callcentre any more with plenty of five-minute gaps to fill in the day), but I still believe in discussion forums.
shortonsense wrote:What efforts were made to test the sensitivity of your final answer on ANY of the uncertainties involved in your core assumptions?
shortonsense wrote:I would also ask about Graph 3.1.2.1 representing the flow diagram to model energy demand. I did not see an explanation on correlation between the various components or how they were handled?
shortonsense wrote:For example, the correlation between the reserve numbers you used and the political stability necessary to utilize those reserves?
shortonsense wrote:The correlation between climate change which is implicit in the model and used to negatively effect agriculture, and the heating demand figure you assumed at 400 kg/capita? I would think that there would be a negative correlation there of some size ( less agriculture because of CO2, more heat, less heating demand, correlation of -0.25 or some such? )
Doly wrote:
I'm very aware that as soon as there are uncertainties in parameters, the whole model can go into question. There are even uncertainties on whether the equations used are correct. I have tested the most uncertain ones by trying a range of values on them, and got an idea of the variability on the results. The whole concept of considering different scenarios is exactly to test what happens under different assumptions and see if there are common themes emerging.
A few things that I have tested for:
1) Different ultimate reserves for nonrenewable energy sources. As you would expect, that delays the decline, but not for long, and the final result is similar.
Doly wrote:2) Different assumptions about EROEI decline in nonrenewable energy sources. The results don't change a lot, unless you put an assumption where EROEI declines rather sharply. In that case, industrial output can collapse as the result of peak oil alone, because the world can't switch to other energy sources on time.
Doly wrote:3) Higher EROEI for renewables. If this is done, renewables to start being used in significant amounts earlier, the decline is slower, but the general pattern is similar.
Doly wrote:4) A policy to increase the use of renewables early, regardless of EROEI. This results in renewables increasing early on, and the final decline being more moderate.
Doly wrote:I'm interested in any other ideas of assumptions that are worth testing.
Doly wrote:shortonsense wrote:For example, the correlation between the reserve numbers you used and the political stability necessary to utilize those reserves?
There isn't anything in my model to represent "political stability", and it's one of the things that have been mentioned before (no wars on the model!) But I don't see any way such a complicated concept could be represented in equations.
Doly wrote:shortonsense wrote:The correlation between climate change which is implicit in the model and used to negatively effect agriculture, and the heating demand figure you assumed at 400 kg/capita? I would think that there would be a negative correlation there of some size ( less agriculture because of CO2, more heat, less heating demand, correlation of -0.25 or some such? )
That's a good point, and it depends a lot on what the expected temperatures are. My model doesn't calculate the average increase in temperature, but it may prove necessary to include it. At the levels of CO2 that my model is predicting, the average increase in temperature wouldn't be more than 3C, from what I have read. I'm not sure if this would affect significantly enough the demand for heating.
Professor Membrane wrote: Not now son, I'm making ... TOAST!
yesplease wrote:What range of EROEI values did ya use for oil Doly?
Professor Membrane wrote: Not now son, I'm making ... TOAST!
yesplease wrote:I asked there too, but I didn't receive a reply and the comments are closed. Based on what I did see, the OP assumed oil's EROEI was at ~100:1 in the 1930s, and fell from there on, using the work of Cleveland. However, unlike Cleveland, the OP didn't preface this with the qualifier that the 100:1 figure was for discovery only, not including production costs over the life of the well, refining costs, and transportation costs. The other figures only include production costs, and also do not include refining or transportation costs. The reason this is important is that refining puts a cap on oil's EROEI of ~5:1, so at best this is what we could have been at during oil's heydays. 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.
Professor Membrane wrote: Not now son, I'm making ... TOAST!
yesplease wrote:
Anyway, like I said before, a model that selectively uses extremely high EROEI figures that are only for discovery and/or extraction, and compares them to EROEI numbers for renewables over their entire life probably isn't going to be accurate. Not to mention that it ignores the difference in exergy, the amount of energy available for useful work, which is another reason why renewables are more cost effective. From 1kWh of electricity I can get ~.7kWh of useful work done in an EV, but in a conventional car 1kWh of oil will only get me ~.1-.2kWh of useful work.
Professor Membrane wrote: Not now son, I'm making ... TOAST!
yesplease wrote:Based on what I've read she hasn't tested for low EROEI in terms of oil and possibly other FFs, just changed the rate of decline from the erroneous 100:1 figure and bumped up renewables a bit in order to see how those outcomes change. Assuming oil and other FFs have spectacular EROEIs when they're worse than most renewables and provide less useful energy to boot hasn't been tested so far, unless of course I've missed something.
Shannymara wrote:She said it's based on the World3 model, which IIRC are open source and readily available. So it's not a black box, generally speaking.
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.
In this case, it would then be reasonable to make the EROEI decline rate a range in a Monte Carlo simulation and quantify "sharply" from whatever else you are normally using.
Interesting. Can you tell from running these scenario's what level of EROEI is necessary to completely NEGATE the use of non renewables?
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"?
Imagine a "Sim City" scenario, make some assumptions about per capita GDP, put a randomizer into the model which triggers a random event of random size, base its frequency on that per capita GDP, allow it to remain stable MOST of the time, but allow in the occasional meltdown through war, and some of those would be in resource rich areas and would suddenly drive the economics in such a way as to boost the need for non renewable development in a short time frame.
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.
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.
After reading the Dollys entire article I wasn't clear on how well value, as in her GDP measures and calculations, mixes with an EROEI basis for transitioning around energy supplies which of course impact GDP.
While its a situation modelers are undoubtedly are forced into working in collaborative efforts, its pretty scary to invest bunches of time and effort, and then when someone asks you the question..."and so what happens then?" and you are forced to punt..."well, someone else built that part, its the XYZ model" and unless the asker knows what the XYZ model is, and does, its just turned into some magical black box answer which you can't even be certain is just randomly multiplying numbers together and pretending its meaningful.
Return to Peak oil studies, reports & models
Users browsing this forum: No registered users and 61 guests