Energy and the economy

Report on My Activities at SUNY-ESF

As many regular readers know I have been in Syracuse, NY, at the State University of New York, Environmental Sciences and Forestry (SUNY-ESF) working with Dr. Charles Hall on biophysical economics. I've been here since Sept. 15 and it has been an absolutely dream kind of sabbatical leave. I've always thought of myself as a student-for-life and a PhD was my ticket to being able to devote my life to learning everything I could that seemed important to know (see my post: "Subjects of Interest").

I have long had an on-going interest in how energy and the economy related to one another. That is based on a very simple observation: it takes energy to do work, and the economy is nothing if not about doing work to create wealth. Therefore, energy must lie at the heart of how the world's economies function. Realizing that the vast majority of energy used by modern industrial societies, and those wanting to become industrialized, comes from fossil fuels, and fossil fuels are a fixed, finite resource, it long-ago occurred to me that industrial economies could not be sustained. The fuels are going to eventually run out. This was even before greenhouse gasses and global warming had come to the fore.

Over the last several years my interest in energy and the economy has been rekindled by the advent of peak oil. I've been spearheading several efforts to get energy systems curriculum into higher education and started exploring research questions that I might pursue. Several years ago, while reading various articles in The Oil Drum blog, I encountered the concept of energy return on energy invested (EROI or EROEI), which I found comported with my own developing ideas about the possibility that some of the proposed alternative energy methods might not actually be sustainable if it took more energy to build the systems than they would contribute to our energy needs. The concept was the brainchild of Charlie Hall who had developed it after years of studying energy flow in ecosystems. He had learned his trade from Howard T. Odum, the world famous ecologist who pretty much invented systems ecology.

Objective

Originally I had several objectives in coming here to learn about EROI analysis and how I might use it to determine the real sustainability potential for various alternative energy systems. I had built a basic model of what I called a sustainability criterion. What the model was supposed to do was determine if the total energy output from a system, say a photovoltaic (PV) system, would be enough to not only supply the user demand, but also produce enough excess power over its lifetime to feed into replicating itself. Imagine a solar PV manufacturing plant that was run entirely from a PV array. But you would also need to run the parts manufacturing plants from PV arrays, do the shipping using energy derived from PV arrays, and, in fact, run the farms that supplied the food to all the workers using PV arrays. Not so simple.

The motivation for wanting to know if many PV arrays could be generated from a few such arrays is actually essential in knowing whether or not PV is truly sustainable. Of course the above scenario would have to be modified to recognize that PV provides intermittent power, so all of those operations would have to have backup power supplied from the grid. But this would be borrowed power that would be paid back when the sun was shining so that the net effect would be as if everything was powered by PV (e.g., wind and nuclear might also contribute to the grid providing an array of sources to cover each other during down time). Right now, all of the manufacturing of wind turbines, solar collectors, etc. is actually powered by fossil fuels or existing hydroelectric. Fossil fuels are the only way to transport parts and finished products as it stands, and coal or natural gas are burned in most Midwest, southwest, and eastern states to produce electricity. Thus the whole alternative energy industry is currently subsidized by fossil fuels and it is by no means clear — no one has demonstrated with data — that these alternatives could take care of their own power needs while still building out the needed capacity to run the economy.

I came here with the idea that I would learn how to apply EROI methods to the analysis for data to fuel my model. But what I learned is that this is a daunting task. As the above example suggests, the web of interrelations between all the manufacturers, shippers, suppliers, materials extractors, even the food support system for the workers is extremely complex. The strength of the links gets more attenuated the farther from the final manufacturing plant you go. Then there is the installation and maintenance activities to consider. All in all, it is currently a nearly impossible task to collect all of the energy consumption data that would be needed to do a good job of modeling and evaluating the systems. That was the first thing I learned.

But while I was coming to this realization, I was developing another idea that had suggested itself to me in thinking about modeling energy systems in general. There is a concept called the 'boundary limits' which basically tells you something about where you should draw the line around the system of interest to prevent infinite regress. You can see this in the above problem where the tractability of analysis can only be bought by deciding to cut off the analysis effort at some reasonable distance from the center, say we decided to ignore everything beyond the first tier of parts manufacturing and shipping. We could forget about those efforts further out in the web because their total contribution might be negligible in terms of each PV system manufactured. Knowing where to draw the line is as much art as science.

The boundary problem exists for these kinds of complex systems when you are doing a 'bottom-up' kind of analysis. That got me to thinking along the lines of a 'top-down' approach. Sometimes it is easier to model a 'whole' system leaving out a lot of detail, yet capturing the essence of the system's behavior. This is, in fact, the approach I had used in modeling synapses and neurons in my days developing artificial brains for robots. Rather than trying to emulate every little detail of a system, it is sometimes possible to start from first principles and develop a model that captures the overall dynamics of the system and allows you to make predictions about system behavior under different conditions.

So for the last month+ I have been working on a macro-model of what I am calling an 'abstract economy'. It looks strictly at the energy flow dynamics of an economy based on a fixed, finite fuel source — not much different from our own real economy. What makes an economy is that energy is used to make artifacts that users need and or want (no moral judgments here). These are the assets of an economy. Some are long-lived, like buildings and railroads. Others have intermediate life spans, like washing machines and toasters. Others are consumable, like food, entertainment, and socks. But over time, when there is sufficient available energy, the economy will use that energy to convert natural resources (metals, salts, trees, etc.) into assets. With the realistic assumption that consumers are insatiable if the energy is available, then I posit a growth constant used to compute exponential growth in energy extraction and asset accumulation. In other words, it looks a lot like the classical developed world economy. I've now got a computer model running and some early, somewhat expected results. But I also have a result that would have been hidden from view had I not built the model.

My Model

I started with the idea of a reservoir of fixed, finite size (arbitrary in the model). The reservoir contains unrefined fuels of average energy density. Extracting the fuels from the reservoir takes work. You are working to find places to drill, working to do the drilling, working to do the extraction, and working to alter the fuel for consumption at the end uses (e.g. refining oil to gasoline and kerosene, etc.). All of that work requires energy from the stream of energy flow, and it takes more work to accomplish the same energy production as the resource depletes. This is the EROI aspect. So the model computes for each time step the gross energy extracted, the energy cost of extraction, the net energy available to asset production (gross minus costs — a concept any accountant or financial person can appreciate), and the accumulation of assets. Here is a graph from a typical run of the model.

The model correctly behaves with respect to the peaking of gross energy production (e.g. peak oil; top, blue line) and increasing costs (bottom, red line). Net energy available for asset production can also be seen peaking. But curiously, it peaks before gross energy. The time units are about 5 years, so net peaks about 20 - 25 years before gross peaks. This is significant in that it is the net energy that runs the rest of the economy.

If it is true that we are currently at or past (or about to get to) the peak of oil production, and all of the data are consistent with this hypothesis, and since oil is the 'kingpin' energy source, needed to extract and transport every other fuel, then it is conceivable that we are also very near the peak of gross energy as depicted. If that is the case then this supports evidence that we have already passed the peak of net energy and hence we are in the phase of decelerating asset production.

I have written for some time about this possibility, but I have only had anecdotal evidence in the economic crises we have been faced with for the last ten years or so (seeing globalization and offshoring of manufacturing and jobs as a crisis for the US workers, as well as our general neglect of critical infrastructure). Now this model reinforces what I suspected. It isn't peak oil that will get us, so much as peak net has already done some damage.

Note how even after net energy flow peaks asset accumulation continues. This is because there still is some net energy available to build new assets. But after a while the decline in net energy becomes so intense that asset accumulation itself peaks and its all downhill from there.

It is easy to understand why energy from a fixed, finite resource peaks and then declines after a time. What isn't completely obvious is why assets would decline the way they do. The reason is that the 2^nd Law of Thermodynamics is always at work on our assets. Plus, we are constantly consuming some of those assets. Entropic decay eats at our bridges and buildings. Physical things age and fall apart unless there is a constant supply of energy to fix them up, repair damage (from use), or replace a worn out object. The rate of asset decay is probably less than shown in the graph because for this first pass I have lumped all types of asset into one class and set a somewhat arbitrary decay rate for them all. As I improve the model's resolution and details I will subclass at least three types of assets (long-term, well-built; intermediate-term; and consumable over the time steps used) and have them decay or be consumed at more realistic rates. But that will not change the general trend over the very long term. The decline in net energy is not unrealistic given the assumption that it comes from a finite resource.

A next step in model refinement will be to add back in the less-than 20% gross energy supplied by supposed renewable sources, esp. hydroelectric and nuclear, with provisions for a growth rate in alternatives like wind and solar. Then we can test how various rates of build-out of the alternatives may shape the future of gross and net energies. We fully expect that a large growth rate for alternatives will extend the gross line at a higher level (but below peak for the non-renewables) for a longer time. But what we don't yet know is what the effect will be on the net energy. Everything depends on the wide-sense EROI of these sources. If they are high (meaning above 20:1) and the growth rate of build-out is sufficiently high (which is a research question to ask the model) then we may be able to see net energy maintaining a higher level and at least keep our asset base from shrinking. My own speculations lead me to believe that it will take an EROI of more than 50:1 and a build-out growth rate greater than 15% per annum for the first 20 years followed by steadily declining rates over the next decades as we approach 100% replacement. That is what the model might be able to tell us. And it is important to know.

If we determine that our assets can be stabilized at something like the above numbers, then it is crucial to know which alternatives will provide us with adequate sustainability (high, sustained EROI) and direct our remaining net energy investments into building those top performers at a rate that will achieve that stability. Right now we are guessing and hoping, hardly a predicament to be in for a supposedly scientific age and mind set.

But therein lies the real problem. We have scientists like Hall and many others who are earnestly trying to ascertain the numbers so we can make informed decisions. The problem is we are in the minority and often just ignored because we bring potentially bad economic and political news to the only people who can do something about all of this. And our less than sapient leaders, indeed the majority of the citizens, just don't want to hear bad news. Achieving what I have in mind will take extreme sacrifice on the part of everyone to allow diversion of our remaining ability to do useful work toward the build-out of these alternative energy sources (and, BTW, this also includes increased efficiency and conservation efforts). No politician is ready to spread the message of sacrifice. I'm sorely disappointed in Obama in not following through on his inaugural address when he said that we would need to sacrifice. In retrospect I guess this was code for 'all you working class folks are going to have to sacrifice (lose your income) while the bankers prosper'. Too bad.

There is one more thing I need to point out that will seem a little untoward, but is essential to grasp. The assets I mentioned include all human biomass! That may sound perverse, but the fact is that our population is supported by the energy flow through the economy. A lot of that energy flow is actually directed to managing the agricultural and food distribution system. More net energy means more and bigger (obese) people are able to exist. Consider, if you will, what the meaning of a crash, as depicted in the graph, means for human life. I'll leave you with that thought. And with a question: What do you think we should do about it?

Comments

[Greg Studen]

I travelled by car from Ohio to Evanston, Illinois, a couple of weeks ago, including a beautiful drive on a sunny Sunday along the length of I-94 as it traversed the Çhicago megalopolis from south to north. I was struck by a thought that resonates with your analysis: how are we going to find the energy to replace all this massive built environment as it wears out? It is hard to conceive of how we are going to run enough wind and solar farms to mine, manufacture, and erect the replacements for the countless buildings, bridges, roads, etc., of our thousands of large cities worldwide--in addition to maintaining and replacing the energy sources themselves. How many joules over the years did it take to build present-day Chicago? How many will it take to maintain it indefinitely, along with the 2.9 million people who live there--food, waste, cars, gadgets? I thought as a drove by the massive skyscrapers and soaring bridges that I was watching a civilization as it hurtled toward a cliff.

[FiniteResource]

Over here in the UK, today's papers are full of the story that Spain generated 50% of its electricity from wind yesterday. You had to read to the end to find out that Spain's medium term target for sustained wind generation is 13%.

Even at 13%, Spain will be light years ahead of the UK on renewables. Like the US, we appear to be in a kind of net energy trap. We're broke. The only thing preventing us sinking deeper into the quicksand is a sky hook of humongous debt, borrowed against growth that will never occur.

The more the UK struggles to get off fossil fuel dependency, the more of its meagre ration of future fossil fuel it burns up.

Everything Robert Hirsch said about needing decades of advance preparation for Peak Energy is coming true.

[Kevin Bressette]

George,
Nothing will be accomplished until the current "leadership" is replaced with, as you say, sapient, technically minded people that understand the problem and it's very dire nature.
We need to make an immediate change in the education system and the media needs to use it's power to shape the ethos toward this essential awareness.
Good Luck.
No one out here is going to change anything until they are forced to so a collapse of the current archaic monetary system is inevitable.
Like you and many others I am convinced that this line of thinking is without a doubt the correct direction and will ultimately result in humans seeing themselves as part of the larger ecosystem and not the "boss" life form.
I just hope it is not too late.
I wish the field was sufficiently advanced to allow me to get involved on a professional level.

Porge from TOD

[George Mobus]

Greg,

I came out east via 94 through Chicago going the other way. Good observation.

Finite,
Good points. Any preparation we do now, assuming TPTB ever recognize a need to prepare, will come at great sacrifices to the general public in order to marshal resources for the work of building sustainable social systems.

Kevin,
(Hi Porge!) Have you read The Upside of Down, by Homer-Dixon? Sometimes the best thing that can happen to an overly complex and energy intensive civilization is to crash so that something new and (generally) better can be built in its place. Like you, I hope we're not to late to even get ready to rebuild.

George

[George Mobus]

qe-0002 sent this comment in via e-mail:

EROI: I have read (don't recall where, sorry) a comparison of the EROI of solar panels, counting only "raw" energy costs, as 6mo-1y in Phoenix and 1y-2y in Seattle. So there's hope, but you are right total energy is the key. I look forward to more analysis!

Collecting all the component sources: You have to be one of just thousands/millions who want that data. Lots of folks have some of the data and would be happy to share. It suggests a separate and useful project is a "library" of component costs. This is beyond your project,
but I hope as you talk with economists, etc., you will suggest it.

Energy cost: I understand the graph is just a model, in reality does the historical energy cost keep increasing like that? I note the price of solar and wind has been falling due in part to less embedded energy. I'd guess that now the embedded energy per kWh of generating capacity is at or below fossil fuels.

Energy costs 2: I've seen some off-topic remarks in RMI's hypercar paper that suggest nuclear's embedded energy cost and global warming footprint are higher per kWh than everything but coal, and other sources saying it is barely competitive with just burning natural gas.

Thanks for posting, great stuff!

[Richard Leckinger]

In New Zealand we are blessed with about 65% renewable electricity on average, and enough fossil resources to keep our stationary energy systems going for some time. What do we do with this largess? We build more roads, to expand the one area of our economy that is hopelessly dependent on imported energy!

We are expanding assets that are essentially redundant and can serve no purpose in the future that is hurtling towards us.

Keep up the good work George, and do give us updates as your model progresses. There are more people listening than you may suspect - some of us are even politicians...

[George Mobus]

Hi Richard.

Thanks for the comment and vote of confidence. I guess I should be careful to not paint politicians with too broad a brush!

Regards

George

[Dan Olner]

This is great stuff. The lack of model-work on these subjects is pretty shocking. I'm in the process of trying to build one myself - concentrating on energy cost changes in an abstract spatial economy. I'm wondering how much space can substitute for energy; you can get "positive localisation externalities" that effect costs, but no-one seems very sure the extent to which these rely on a specific amount of energy input. In fact, Krugman's work suggests agglomeration relies on low transport costs - if this is true, increasing them may well be a de-localising force. Interesting when one compares the push for localisation as a solution to energy crisis. (Though the importance of e.g. Transition initiatives may lie in their getting people active, perhaps, rather than actual local economic outcomes.) Again, it's something we know woefully little about - perhaps there's some vital studies I've yet to come across, but I doubt it.

I'm particularly interested in your 'abstraction' choices; all models have to make them. You're allowed to build what Krugman calls 'silly models' in certain parts of economics, but mostly elsewhere, everyone seems to think more data = better model. This often leads to the most ridulously bloated, GIS-based, data-crammed works of fiction that can say nothing about key dynamics. It's difficult, though, and it's a problem I have yet to crack convincingly.

Do you have any more technical details on your model? A working paper or anything? In my own, I'm imposing energy cost changes exogenously on agents who then have to make economic and spatial decisions. (I've linked to my blog; in case you're worried I might steal anything from your model, you can go over there and tell people! Oh - I'm from the University of Leeds in the UK.)

Bye for now.

[George Mobus]

Dan,

Thanks. I will check your web site.

The paper that Charlie Hall and I are working on will take a bit more work. We decided to modify the model to break out different asset types (long-term ((fixed)), intermediate-term, and consumables, along with human biomass!) and try to apply some realistic depreciation/decay/consumption rate constants. Then with those results we think we'll have less of a 'silly' model (love that term).

George

[Dan Olner]

Having just checked, Krugman actually talks about 'silly assumptions' -

"To get this system or aggregate level description required, of course, accepting the basically silly assumptions of symmetry that underlay the Dixit-Stiglitz and related models. Yet these silly assumptions seemed to let me tell stories that were persuasive, and that could not be told using the hallowed assumptions of the standard competitive model. What I began to realize was that in economics we are always making silly assumptions; it's just that some of them have been made so often that they come to seem natural. And so one should not reject a model as silly until one sees where its assumptions lead."

Which leads to one of his recommendations: dare to be silly! Well worth reading -

http://web.mit.edu/krugman/www/howiwork.html

[eman]

What type of vehicle are you driving Prof. Mobus?

[George Mobus]

Dan,

Thanks for the link.

Eman,

Except in extreme weather (like snow around here) I ride a Honda 250cc (Rebel) motorcycle that gets between 75 and 78 MPG. This is my commuting vehicle or for trips to the store. At other times I drive a Toyota Corolla which gets about 35 MPG city.

George

[Term papers]

I would like to say that it is very interesting to read your blog.Alternative energy system is vital for any country.