Report from SUNY-ESF

Every serious academic dreams of the opportunity for a sabbatical leave. Sabbaticals are not extended vacations! They are supposed to be periods of intense research and, generally, expanding one's realm of knowledge. The idea is to get away to somewhere new and learn something you didn't know before (or at least extend your repertoire of expertise). Some academics use the time to go work with colleagues they already know and collaborate more deeply in some project they are already working on. Some use the time to write books or develop new academic programs. There is no one 'right' kind of sabbatical other than it should be some form of professional development. When I started thinking about my sabbatical several years ago, I thought I would use it to write the two books that I've been outlining in the blog series', "Sapience" and "The Science of Systems" (see Series Index for links to both of these.

That was before I went to the First Biophysical Economics Meeting here in Syracuse at State University of New York - Environmental Science and Forestry (SUNY-ESF) and met Charlie Hall and his students last fall. I'd known about Charlie's work on energy return on energy investment from a number of papers he had written for The Oil Drum. I went to the conference even though it had no direct relation with my academic career as a computer scientist or roboticist because I had become very interested in the issues surrounding energy after learning something about the peak oil phenomenon. Regular readers already know something of the evolution of my interests. I had worked in the solar energy space heating business back in the 80s when the oil shocks of the 70s and 80s got everyone thinking about alternative energy to escape the threats of being a net oil importer and OPEC having something of potential strangle hold on the rest of the world. Those were heady days for us naive energy enthusiasts. We were convinced that solar energy would save the day. But I got some deep experience engineering these systems and an idea struck me. I got to thinking that while a solar system on the roof might provide the home owner with lower heating bills (especially with tax incentives) that the amount of energy that was needed to manufacture the flat plate collectors (mine and smelt the aluminum and copper, glass, and the oil content of urethane foam, etc.) was about as much, and could be more, than the energy benefit gained. In other words, the country as a whole could be using up precious energy to build systems that would not produce a net gain in energy! Well that is exactly the question Charlie Hall was asking about all energy systems. I wanted to know more.

So I went last fall and found the most intriguing world of academic pursuit I had yet been exposed to. This question of energy returned for energy invested (EROI or, in other words, how much energy do you have to spend to get a unit of energy?) could very well be the most important question we can be asking right now. I was extremely fortunate that Charlie and ESF agreed to host my sabbatical visit. I am learning in high gear. And gratified to see that some of the ideas that I had developed independently actually have relevance here. This is just a quick update on how things have been going and some notes on what I have been gleaning.

One of the biggest questions I had when I came here was to find out how large the boundaries around the energy system had been chosen. In the above example I just looked at the energy requirement from mining raw materials to final manufacture and delivery of the solar panels. Most of the data had to be derived or guessed at based on dollar costs of direct energy inputs. An extremely tight boundary would have been just the factory where the panels were manufactured, adding up just the costs of the materials and the energy utilities used at the factory. That does give a preliminary estimate but only if the cost (in dollars) data has some basic relation to energy measured in energy units. This is not always the case unfortunately. But prices can act as rough surrogates for energy, so that is what has to be used most of the time.

A wider boundary would include all indirect energy inputs. Everything from the calories consumed by workers (both in the factory and in the offices) and their families (if they were the sole bread winner) to the energy used in lighting — in other words all overhead and wages at the factory. To get a really accurate measure of the energy consumed in making, say a solar photovoltaic panel, would require extraordinary efforts. I have developed an idea for how it might be done (later) but it would not be easy or cheap. Yet, here is the problem. If we don't have an accurate grasp on the total inputs to a particular technology we could be making some huge policy and market mistakes in what alternative energy technologies we decide to invest in. As in all network of processes relationships, a local optimum does not imply a global optimum. What looks good at one node in the network, the choice to go to that other node instead of any other choices just because it looks like the best outcome can have disastrous impacts if all of the subsequent choices lead to extremely high costs.

Building lots of wind turbines looks like a pretty good choice right now from the perspective of financial analysis of where we are and where we need to be. But I think we are working with a very small amount of information when making this decision. Until we know the full energy impact of turning massively to wind (such as the energy that will be required to create a distribution grid that can handle the intermittency problem) should we be racing ahead with plans to build huge wind farms? What would happen if down the road we discovered that the total energy required to replace and repair turbines, delivery and installation, distribution and temporary storage somewhere in the grid, and who knows what other infrastructure requirements, all turn out to take more energy than we expected (in EROI language, a ratio of, say, less than 5 to 1 — five units out for every unit we spend). If you know that for most of the 20^th century the EROI of oil was > 40:1, you will be able to see that we could not begin to sustain a civilization based on that latter value with a technology of 5:1. Right now the EROI estimates that have been given for wind have been much higher, in the teens on average. But these are largely based on fairly narrow boundary choices so they do not look at all of the energy costs that fold into the final product.

This diagram might help to visualize the problem. In the diagram is a very partial map of some of the nodes (work processes) where energy is used and some useful work is done (note that labor means all of the people who perform work at all the other nodes; Marx would approve I think!). A few of the energy flow relations are shown. If I tried to show every energy flow relation you wouldn't be able to see much of anything. This is what we call a 'dense' graph (network) in math speak. Most of the lines are bi-directional, meaning that energies of one kind or another flow in both directions.

 

Take special note that every single node requires fossil fuel inputs, some directly, like delivery (transportation), and some indirectly, like coal-fired generation of electricity. Solar energy only enters for the most part from the left side at agriculture and from the right side, operations of the wind turbines. Of course solar energy warms the whole system in the summer, but that may mean a need for air-conditioning (well, more like a want than a need). The issue is you can see how incredibly hard it would be to capture all of the relevant energy flow data for a network like this. And that includes appropriate allocation of the fractions of those flows that actually end up contributing to the building and operating of the wind turbines.

What should be obvious from this partial network map is that the amount of energy going into the production and operation of wind turbines is potentially much greater than is currently counted as energy inputs. What confuses and confounds our ability to have an accurate accounting is that everything is now denominated in monetary units. Dollars. The problem here is that prices and costs do not accurately represent the value of a product or service. Profit motive gets in the way. Prices are determined by the market not some kind of summation of costs plus a standard markup. Thus the prices paid by company A for a product from company B (say a component for the wind turbine) depends entirely on what B can get away with. This is, in part based on how much company A really 'wants' the product. So according to classical economics, buyers and sellers settle into an agreed price based on utility (whatever that really means). In such a system it is simply too easy to lose proportion as well as perspective. In highly competitive markets prices might understate the value of energy put into the product, while in the case of high demand prices will overstate the value of energy. What to do?

And therein lies a great danger. Note the complexity of the above partial network. If we were seeking to do a thorough analysis of EROI starting at the factory/operations area, setting our boundary around the manufacturing plant, delivery to site, installation, and on-going operations we might come up with a number that makes wind power look good EROI-wise. But what would we get if we start spreading our area? Might there be a real surprise out in that web of connections? A greater energy input than prices would reveal? I contend the risk of this being the case is actually quite high. Note in the lower left-hand corner of the figure I've put in a couple of nodes representing governance services and "government help". This latter could be anything from the subsidies payed to corn growers and ethanol producers to the cost of the war in Iraq (to protect our oil interests). Government 'help' goes to too many of these nodes to simply ignore. Which means some prices are artificially low thus hiding the true energy component. To me this is a real problem with narrow boundary analysis. From an argument of practicality (one Charlie himself advances - see below) it would be extremely difficult to gather data from wider boundaries. After seeing what has been done here, I certainly see the validity of that argument. But what might we be missing if we don't do it? What if the government decides to 'help' the wind industry so much that we start building wind farms like crazy, only later to find out that once we started relying on wind for a significant amount of our electricity, we didn't actually have enough net energy because too much actually needed to be fed back into this network to replace some of those fossil fuel inputs. We could just as easily find a global net drain on energy as opposed to a net gain that would come from a higher EROI.

Charlie is right, of course. It is a nearly impossible thing to contemplate such a massive project as widening the boundaries and collecting the data (if it could even be found). But I and a number of the students have something of a queasy feeling that we're missing something vital in all of this. Several of his students, and one quasi-post-doc on-board for a while, all believe that some kind of "bottom-up" analysis is essential to a better grasp on real EROI. The post-doc has actually done a very interesting study of gas wells, energy costs of finding, drilling, extraction, etc. versus gas delivery rates over time. Without stealing his thunder, I'll just say that all those stories you've been hearing in the MSM of late abut how natural gas is going to be our savior(?), well don't believe them.

I'm proposing that we need to do a project where we define a set of ever-expansive boundaries, like concentric rings around the manufacture of, say, photovoltaic cells and panels. We should choose a domestic manufacturer and hope that that manufacturer obtains most of its materials from domestic companies. The reason is that in the US the accounting rules and accessibility to utility rates would allow us to derive actual energy inputs. We could find out how many actual 'therms' or kWh's companies used to do what they do. We could have access to direct and indirect labor in numbers of employees, which gives us the labor energy (the average worker burns so many calories per day, etc.). We could follow the supply chain backward in stages, checking only the embodied energies in the intermediate products without even considering the prices paid or costs incurred. In essence we would shadow the cost accounting system but look at direct translations of dollars into joules! One side product of this method would be a check on estimates currently floating around for the current joules per dollar average. That conversion number is currently used to make very rough estimates of energy inputs from supply chains. But as I argued above, the price-cost fallacy must distort this ratio and do so dynamically over time.

Now Charlie, who has been doing this stuff for many , many years has adopted a pragmatic philosophy about the value of the EROI estimates we have today from narrow boundary analyses. He has found that the EROIs of oil, gas, and coal are declining at a rather alarming rate. While the numbers may not be exact, its the trend that counts — that's the scary part. Moreover, the early EROI estimates for alternative energy technologies aren't really that hot (pun intended). He is working on efforts to come up with what a minimum EROI would be required to maintain our current civilization (USA standard of living). Remember EROI translates into net energy available to do useful work. Thus there is some minimum average value that, properly distributed through the portfolio of energy production technologies, would allow civilization to continue (assuming we could magically convert to it before the oil runs out). Having a reasonable estimate of what that minimum should be might be considered by some thoughtful people as a really good idea. It could, if our politicians actually knew what they were doing, translate into rational policies on energy. But as the climate scientist found out you can lead a policy maker to knowledge, but you can't make them use it!

There is much more to report, but this one has gone on too long as it is! I will write more in the future. Suffice it to say that I am in a state of ecstasy for the opportunity to work with world-class scientists on one the most important topics that will have a tremendous impact on humanity in the near term. I'm realizing the dream of a perfect sabbatical and feel extremely fortunate to have this opportunity.

The Second Annual Biophysical Economics Meeting is scheduled for Oct. 16/17 here at SUNY-ESF. Joseph Tainter is giving the plenary talk Friday night. We have Nate Hagans and Gail Tverberg (Gail the Actuary) from the Oil Drum giving talks as well as many very knowledgeable folk presenting their works. They will even let me make a presentation on Sat. Once it's done I'll post a link to the slides. And, I am happy (and a little proud) to say that I will be giving a pre-conference workshop on the basic relation between energy and economics for those participants who may not have quite gotten the connection. So I'll be busy for the next few days getting ready for that meeting. Very exciting times here at SUNY-ESF. Wish you could be here!

Comments

[Neven]

Man, I'm jealous! Have fun over there, George!

[Derek]

Really jealous does not do it justice, I’m on the edge of crazy! I mean really I would give up both my bicycles for a chance to meet the TOD crowd and hear both your and Charlie's cutting edge EROEI work. I’m stuck in DK, so I will be checking back for your posting up slides from your (and others if you can snake them) presentations. Agreed on all fronts; EROEI is critical; Trend of 10:1 is 'danger zone', 4:1 time to learn mule driving as a profession. Your posts don't run long enough. I say the scientific community call for energy labeling, like nutrition labels but in 'complete accounting'(TM) of energy networks, due yesterday! Personally I would like to take sabbatical at the SITropical Research station in panama better suntans and snorkeling than in NY.

[GaryA]

Sounds great George! Its a fine counterbalance to the dispair we all feel-new discoveries & ideas clearing a path through our own cerebral jungle growth.....How many of us can recall our 'Eureka' moments?- when we have struggled with a nagging intuition then suddenly- bam! Its so obvious. My personal ones were Dunbars number, Ruddimans hypothesis and Brig Kylce's demolition of selfish gene neo-Darwinism...EROI and eusapience are in there too!

[George Mobus]

Nevan,

I am having fun, but it's really hard work (I'm generally here till 7 or 8 pm 6 days a week!)

Derek,

Why didn't you mention the Panama thing before I did this? I love snorkeling in tropical waters.

OK, GaryA, you will have to give me some references. Or I guess I can go to Wikipedia. Just when I think I have a handle on things, somebody has to mention the fact that there is something I haven't heard of, and ...

George

[GaryA]

Which one George? I dont think any of these theories impinge on your work too much they are personal interests of mine. I can give more detailed info via e-mail but you will have to allow me a few days to reply as I'm busy at (hawk,spit) work at the moment.

BTW I'm just beginning to realise the implications of this EROI methodology shift- it could (or could not) be the final nail in the 'green technology revolution' idea..(without reducing our lifestyle >10x)

[George Mobus]

GaryA,

It's just that whenever someone mentions someone or something that I haven't heard of I get curious! No need to respond any time soon. But if you want to explicate in an e-mail I think you know how to get in touch.

George

[Neven]

WRT Dunbar's number, google "monkeysphere" for a very funny explanation (first hit). :-B