[Paleopsych] energy problems
Werbos, Dr. Paul J.
paul.werbos at verizon.net
Mon Aug 9 15:10:08 UTC 2004
I hope that Premise Checker will forgive me for being a bit out of sorts
when I saw his message.
It came at a time when, for the nth time, I "recalculated" some
expectations and got some
feedback, and it was a tad depressing.
Economic theory is essential to a real comprehension of what's going on in
markets. He is certainly right about that. Back in the 70's people did a
lot of stupid klutzy
things to try to make energy problems go away in defiance of economics, and
things themselves caused a lot of problems. (I disagree with the folks who say
the klutzy things caused most of the problems... but certainly they were a
But the klutzy things happen today just as much as they did then. Corporate
welfare now is
as bad as protection-from-the-market was then. Of all the noises of the
weeks in the election itself... Robert Reich's little interview on TV
a whole lot more than all the rest.
But that's a digression.
Since some of you may believe in the magic of the marketplace, I probably
owe you another digression.
Back in 1979, my first real tenured job was at DOE, to do an in-depth
comparative study of the
Long-Term Energy Analysis (Program), LEAP, the official model then used to
generate their long-term
projections. Previous critics had told DOE they needed to respect and learn
from the wisdom
in the private sector -- so LEAP was basically just a massaged and "made
transparent" version of
Shell's Long-Term model.
The model was beautifully exact as a reflection of free-market perfect
market microeconomic theory.
As in Ken Arrow's textbook on the subject. All actors were assumed to have
perfect foresight or
"clairovoyance." (That's one of the required assumptions in the theory.)
They all used marginal cost pricing,
and they all looked ahead in making all decisions. Major "modules" within
represented the supply and demand curves over time faced by a typical firm
in each industry; thus
a module and an industry or economic niche were essentially the same. Basic
types of consumer
But... for a given level of undiscovered resources of oil and gas, and a
given level of elasticity of demand, the model
wasn't all that different from other such models. Even with free markets
hard-wired in, it was perfectly possible
to see projections that would be disastrous. In fact -- a few years later,
when I took over lead responsibility for
long-term analysis, I arranged a model run to see what we could get with a
big worldwide carbon tax,
based on CO2 emission and intended to minimize CO2 emission. (I think this
was the first official
run of that sort, back in the 1980's.) The results were grossly
disappointing -- hardly any real benefit;
what reductions we saw in CO2 emissions were due to reductions in economic
growth more than anything else.
And believe me, we put a HUGE amount of effort into making sure that the
assumptions we put into this were realistic.
We have whole books out there on the key parameters which drove these results.
But now: where are we today, and why is it scary?
I started paying more intense attention to this last summer, when I was
talk at a conference in Mexico.
I highly recommend looking at some of the talks -- particularly the one by
It's not that you would enjoy it. But as the mama said in the old TV movie:
castor oil, and drink it ALL UP. It may taste awful but it's good for you."
(And don't just
spit it out...!)
In his most benign scenario, he projects something like 2/3 of the world
oil coming from the
Persian-Arabian Gulf by 2025. And then he draws out the implications, which
are VERY serious --
the more serious the more honestly you face up to the whole big picture.
benign scenario already includes conservation, hybrids, etc. As I walked
out of there, a bit shaken,
I resolved to try to do what I could to prevent that scenario from happening.
(And I resolved not to be paralyzed by all the econometric theory showing
how hard it would be.)
Al-Shatti's methodology for oil forecasting may be respected in the Gulf
(an area where they do demand
good results but not always Western methodology) ... but it did leave out
certain points I would have wanted
to check. But a month after that, the Policy Office of DOE (with NASA as a
sponsored a workshop in Aspen Colorado, organized by Marty Hoffert of New
York University, to discuss
technologies to address the CO2 issue. Cavallo (formerly DOE, now DHS, but
still in the same office
in New York) presented a paper on future oil trends, arriving at much the
as Al-Shatti regarding OPEC dependency, but using the standard USGS
undiscovered resource numbers.
As I understand it, a full understanding of the latest Shell numbers is
more or less the same. (In many
ways, Shell's "renewable" numbers might be better labelled "gap we need to
fill.") Likewise IEA.
EIA and Exxon are more complicated phenomena, but it would be a digression
to elaborate on them.
Let's just say that there are enough leading forecasts pointing towards a
chilling future that we
need to take the risk seriously.
Anyone who has really run the numbers would laugh -- with a measure of pain
and a measure of contempt --
at claims that more tax breaks for friends of the party would magically
solve all this. There are industry
reps who run around DC telling people that there is no problem.. telling
other nations that they should
sell their oil to Exxon while they can, cheap, before it becomes cheaper...
even as the financial Times reports
huge shakeups and terror at majors' inability to replace proved reserves,
and the failure of desperate plots
in Russia and Nigeria, etc. Today's paper even reports on the oil majors
who want to keep Chavez in power
in order to keep Venezuela's oil.
In fact... to make Al-Shatti's dire forecast not happen... we really would
want half the world's cars to be
able to run without gasoline, in a crisis. Roughly. By 2025. To reach that
goal -- we need to remember that
the average car stays on the road for 15 years. (This is not a random
number. Maybe DOE still
has copies of the Transportation Energy demand Model I build in the
mid-80's, then the model
used to generate the forecasts in the Annual Report to Congress, etc. It
was simple but beautiful,
and VERY well grounded. We had very solid data on how long cars stay on the
road, for example,
albeit not all public data.) Thus we would ROUGHLY need all new cars sold
by 2010 to be capable
of running without gasoline, in a pinch.
That may sound nearly impossible... but it's not.
There is one and only one really known and proven and affordable way to
achieve the subgoal
I just stated: fuel-flexible vehicles (FFV).
FFV have a long history. Roberta Nichols of Ford had a long crusade (with
full support form her company) to
try to make FFVs happen. In 2003, she published her history of what
happened and why, in
a local journal. (I have pdf copies, about a meg.) From other sources, I
hear it would take
only $200/car to make new cars able to handle any mix of gasoline, ethanol
("GEM" flexibility), and -- more important -- only two years to retool the
sold to a typical nation. (The US, however, would require an extra two years
to meet unique regulatory barriers against doing anything new. Our lawyer
system really is
unique, though I hear Europe is trying hard to catch up and surpass us...)
People in Brazil are happy that their free market is ALREADY selling
growing numbers of GE cars.
And Kerry is talking about incentives which, among other things, would
a nice harmonious gradual introduction of GE flexibility here too.
Sounds nice... but nice and gradual is not 2010. And GE is not GEM. There
is a time
when someone needs to bite the bullet, when, after all, the costs are not
(No government spending; the equivalent of a 1 percent tax in the cost of a
which should fall as automakers use new technology which makes flexibility
THE BEAUTY OF GEM FLEXIBILITY IS THAT WE DON'T HAVE TO BE CLAIROVOYANT.
The effect of GEM flexibility is to give us more of a free competitive
market for fuel
IN PRACTICE, DE FACTO. It is sweeping away historic BARRIERS to competition
in the fuel business.
We don't have to choose between biomethanol, biowaste methanol, coal-based
gas-based methanol, ethanol or gasoline itself. We let the market itself
And what if we believe that electricity, hydrogen or direct natural gas
will be better car fuels
than alcohol OR gasoline in the long-term?
No problem. The law could say:"Any car registered in our country WHICH was
placed into service after Jan. 1, 2007, and which has a tank to hold
have a tank which can hold gasoline, ethanol or methanol, and must be able
any of these three fuels." (Some technical details would be needed -- e.g.
E85 and M85 is good enough.) That allows anyone to buy a dedicated electric
car on the same
terms as a GEM car. It leaves the decision up to the marketplace. I
that we would see so many dedicated electrics on the road by 2010 -- but if
I don't see the US doing this any time soon -- and, for many reasons, the
US is not the best
place to start anyway. But I have started to hope that someone might. That
gives us some hope.
And, yes, there are sources of ethanol and methanol out there enough to
make a real difference,
at least for some nations -- and any loosening of the world oil situation
right now would be
very welcome, even if only worth a million barrels a day or so.
OK -- but that is only the start of the game. The early start.
The three main pillars of energy economics are the market, the regulations
(all-pervasive in the real world) and the R&D system which generates new
Back in the 80's, the key lesson we had to learn was that "oil shortages
are not energy shortages."
But GEM flexibility will immediately change that. It makes energy more
GEM flexibility is essential to pulling our "airplane" out of a tail
spin... essential, necessary...
but very far from sufficient.
The cheapest way to make alcohol fuel today is to convert remote natural
gas to methanol,
which is far easier to distribute to the car driver than natural gas
proper. It costs $250/ton
using new technology in large-scale use today.
(Google "canaccord methanol" to see a business plan complete with vendors.)
competitive -- and it's great news for a new "marginal" consumer.
HOWEVER -- the USGS now estimates usable conventional natural gas around
as slightly less than oil!!! In the past few years, gas estimates have been
lowered and oil raised
(both by modest amounts, but enough to switch the story -- and, more
disprove what once seemed like a strong trend towards gas). What surprises
me -- as I do
a quick web search even of the Gas Salesman Optimists ... they still
publish all the rhetoric about gas lasting forever, but their numbers
aren't much different from USGS.
They state that there have been major advances in technology lately,
which have reduced the range of disagreement. (In fact, some parts of the
new technology can
actually be traced back to an algorithm I developed back in 1974, and I
have seen first-hand
how things are qualitatively different now.)
If the $250/ton technology is only 50 percent efficient, and if we assume
we can use ALL the
natural gas to make methanol for cars, and we assume all the methanol is
used in GEM cars
which have the seem efficiency using methanol as with gasoline... The net
only be a 50% increase in the world's supply of car fuel. That may buy a
little time... but it
is very discouraging. It is especially discouraging since gas prices have
been going out
of sight near pipeline systems in North America (and Europe?), due to excessive
use in electric power generation (above all), but also industrial heat and
And so... to stop the hemorraging, it will be critical to stop the waste of
natural gas for
electricity generation as soon as we can.
Bear in mind... if you ask folks like Entergy why all this gas, and why no
use of cheaper fuels like coal,
they will say it is all a matter of regulation forcing people to use
natural gas. I will refrain from saying
more here and now. Clearly we need more efforts to change this situation.
There is a huge amount of rhetoric
about the "intelligent power grid," but we need action to get it out of the
space of pure hot air and pork barrel.
And there are the oxygenated coal gasification technologies, able to
produce electricity and methanol
profitably as coproducts, which we somehow need to get inserted into the
in China if not in the US -- but certainly we could use it here! And there
are some new
LARGE-SCALE distributed generation technologies we need to mainstream and
proliferate as soon as we can
(though the best will require about 3 years to tune and shake out for
optimal efficiency and mass-production cost).
And... good old space solar power is suddenly much more relevant and urgent
and even timely than I would have thought,
back when we first joined NASA in funding a small research topic in that
area. (Search on "JIETSSP"
at www.nsf.gov). We have a lot to do to make that real, but we do see a
realistic path now for doing it.
(Though I keep being reminded of Dan Brown's Deception Point which, while
towards some real barriers.)
Other areas where R&D is critical and lagging to a worrisome degree --
pursuit of direct methane-to-methanol technology, which has real potential
to get us
close to 100 percent efficiency, thereby making natural gas a potential 100
percent extender of oil;
carbon-tolerant alkaline fuel cells, which can get twice the mileage on
methanol as a hybrid car,
thereby doubling again the value of the natural gas;
depressurization technology for methane gas hydrates, which, COMBINED with
the two other technologies,
might provide an economic technology for massive amounts of advanced car fuel;
technology for "beneficiating coal," using solar thermal or space solar
to allow us to produce one molecule of methanol per atom of carbon;
and, most valuable of all, batteries good enough to be used in "heavy hybrids"
that can be plugged in so as to reduce their use of GEM fuel -- and
gradually phased into
full-fledged electric cars of adequate range and cost.
There is no one magic bullet. We will need them all -- especially given the
"decision tree stochastic" nature of the game. And we need mechanisms to
so that oil producers and car drivers can cooperate with less fear on both
sides -- yea
even the nation-states which sometimes represent them.
P.S. Some of you might wonder: Is there a way to calculate shadow prices
correctly, for use
in a distributed agent system, for the STOCHASTIC case, in a way which is
feasible to compute or approximate?
Can we get away form the requirement for clairovoyance?
Yes, we have figured that out. See chapter 1 of the Handbook of Learning
and Approximate Dynamic Programming,
Si et al eds, published by IEEE and Wiley this past month. In particle,
note the discussion of the stochastic
Pointryagin equation and Dual Heuristic Programming.
More information about the paleopsych