November 23, 2008

New developments

This blog provides a reference - a FAQ. As I learn things, I add them to the individual articles, whose posting dates don't change as they are updated. People need a way to know what's new, so, they'll be here in this post. It will be at the top unless I add an entirely new post.
Is solar being slowed down by the current credit crunch?

Only slightly. For instance, one of the largest suppliers has cut it's 2009 forecasted growh from 75% to 58%. (posted 12/22/08)


EV technology, while more than adequate, will continue to improve. Regenerative braking eliminates one source of waste (or finds a new energy source, depending on your perspective) by capturing vehicle kinetic energy: another source is the vertical kinetic energy now lost to shock absorption, which appears to have been solved. (posted 11/23/08)

October 20, 2008

Was Malthus right to forecast starvation?

No. As a practical matter I think we're going to have a hard time juggling all of our resource crunches (especially energy), but I think we should be clear on what the theoretical problems are, and are not.

The most important refutation of Malthus is on the population side: world population as a whole has clearly stopped growing exponentially (or geometrically, if you like), due to the demographic transition (it's roughly arithmetic at the moment). This is a key point: in many ways, growth is generally self-limited, and follows a logistic (or sigmoid curve), generally referred to as an S-curve. For instance, US car sales peaked about 35 years ago, and the US has a clear over-supply of vehicles, due to increasing vehicle longevity.

In fact, in most of the world population growth is on a long-term negative path, due to fertility rates well below replacement, including Western Europe, China, and the US (excluding immigration). Russia, Eastern Europe, Japan and Italy are starting to show absolute declines in population. This is detailed at the UN site below, where we see that the growth rate was as high as 2.19% back in 1963. The total population peak is currently expected to be at about 9 billion around 2075 and population is expected to drop after that.

See especially Figure 3, page 6:

Let's explore a couple of key concepts: the difference between arithmetic and exponential growth, and the difference between high fertility and "bottom line" growth.

Consider the following series of numbers: 10, 11, 12, 13, 14. There is growth of 7.7% at the end, but this is arithmetic growth: the change from number to number is constant, not growing with the base. Malthus assumed exponential growth for population, and arithmetic growth for agriculture. The growth rate of the world's population has been falling for 50 years, from 2.2% (in 1963) to 1.1% (in 2012). . We see the current absolute number is about 78M per year,  and decreasing: exponential growth for population has ended.

Growth varies enormously by country - in Japan, for instance, absolute growth will be negative next year. Italy and Russia will follow soon after. These alone are sufficient to refute Malthus's general rule in a simple, clear fashion.

For many more countries, the fertility rate is below replacement. If every couple has less than about 2.1 children (the definition of the replacement fertility rate), the population is very young, and the death rate is low, there can be a lot of children and "bottom line" growth in the population, but in the long run the population will stabilize and decline, as every generation is smaller than the one before. So, if we clearly have fertility rates below replacement, we clearly have in the long run stable or declining population growth.

Now, are there still parts of the world growing pretty quickly? Sure, but they're in the minority. Just as importantly, the parts of the world that aren't growing clearly refute Malthus's idea that population always grows until it hits a resource limit - he couldn't conceive of voluntary birth control.

Do we still have huge, basic sustainability problems? Sure, but it's important to know that the broad, simple framework that Malthus proposed is just plain wrong.

This kind of logic also applies to energy. Like population, US car sales, and many other examples, energy markets (at least renewable ones, like those for wind and solar electricity - so no, I'm not talking about oil) will naturally mature and flatten out long before we reach theoretical limits.

For far too long we've been talking about a false dichotomy between "infinite exponential growth" and collapse. In fact, with a little luck, growth in resource consumption will gradually come to a stop, while humanity switches it's desire for improvement to what are generally known as "services": health, education, art, etc.


Just for the value of good trivia, let's note that Malthus believed that population growth would continue forever because he believed that contraception was morally wrong, not that it couldn't work.

Further, some argue that Malthus’s argument was principally a class one, designed to rationalize why the poor must remain poor, and why the class relations in nineteenth-century Britain should remain as they were.

His greatest fear was that due to excessive population growth combined with egalitarian notions “the middle classes of society would . . . be blended with the poor.” Indeed, as Malthus acknowledged in An Essay on Population, “The principal argument of this Essay only goes to prove the necessity of a class of proprietors, and a class of labourers.” The workers and the poor through their excessive consumption, abetted by sheer numbers, would eat away the house and home (and the sumptuous dinner tables) of the middle and upper classes.

He made it clear that the real issue was who was to be allowed to join the banquet at the top of society.

Charity simply encouraged early marriage and larger families, he concluded, inspiring an 1834 Poor Law that effectively limited public relief to people toiling in workhouses. Charles Dickens, appalled, rebuffed Malthus with the tale of a flinty miser who had a change of heart. "A Christmas Carol" is still in print.

update: here's comments by "relhager":

There are two relevant points: One, as Nick noted above, is that the UN predicts that the human population will never double again, and will start decreasing during this century. That’s critical.

The second is just as important: Malthus was simply wrong. Or at least half wrong. He was right about population growth (up until now, at least) but very wrong about food production. During the twentieth century, global population increased sixfold, and food production increased sevenfold, thanks to advances in genetics (the Green Revolution) and especially chemical fertilizers (see my recent book on the history of the subject, "The Alchemy of Air," as well as Vaclav Smil’s excellent "Enriching the Earth.").

Yes, we will have big issues handling the population high-water mark (+ another 4 billion or more) before we start going down in numbers, especially because of pollution. No, this does not answer questions about energy resources (making chemical fertilizer currently eats up about 1 percent of all the energy used on earth). But if we ate a reduced-meat, high-vegetable diet, we should be able to avoid the mass starvation predicted by Malthus — forever.

October 3, 2008

Would an energy policy fix our financial crisis?


First, the US's basic problem is a chronic trade deficit, which has lately been financed by mortgage borrowing by US households from other countries. This household borrowing reached excessive levels - the current bailout is intended to fix this.

2nd, the bailout is intended to transfer excessive household mortgage debt to the Treasury. This moves debt to where it can be better carried, but raises the risk of a currency/debt crisis in which foreigners decide the US itself isn't creditworthy.

3rd, the only longterm solution is to fix the chronic trade deficit. Such a fix would greatly improve the US's creditworthiness in the eyes of foreign lenders, and simultaneously reduce the need for such borrowing.

4th, the primary source of the US chronic trade deficit is oil imports, and has been since the 70's.

Conclusion: the US urgently, and as it's first national priority, needs an aggressive plan to reduce oil imports in the short-term and long-term. Such a plan should include short-term measures such as carpooling and telecomuting; in the medium term an aggressive increase in automotive CAFE requirements, fuel taxes, fuel substitution (PHEV/ErEV/EV's, CNG, ethanol, CTL with sequestration) and in the long-term domestic drilling.

September 25, 2008

Can we replace oil in general?

People who are pessimistic about dealing with Peak Oil wonder: which processes happen to use oil today, because of historical accident, and which truly have to do so? What part of manufacturing, transportation etc, is specifically reliant only on oil?

So many things run on oil - can we possible replace oil in all of these applications?

The answer is yes, primarily through electrification of surface transportation and building heating. Aviation and long-haul trucking can be replaced with electric rail and water shipping, and aviation will transition to substitutes.

This will proceed through several phases. The first is greater efficiency. The second phase is hybrid liquid fuel-electric operation, where the Internal Combustion Engine (ICE) is dominant - examples include the Prius and, at a lower price point about $20K, the Honda Insight. The 3rd phase is hybrid liquid fuel-electric operation, where electric operation is dominant. Good examples here are diesel locomotives, hybrid locomotives, and the Chevy Volt. The Volt will reduce fuel consumption by close to 90% over the average ICE light vehicle. This phase will last a very long time, with batteries and all-electric range getting larger, and fuel consumption falling.
The last phase is, of course, all electric vehicles, which are are slowly expanding, and being implemented widely (Here's the Tesla, here's the Nissan Leaf). Electric bicycles have been around for a long time, but they're getting better. China is pursuing plug-ins and EV's aggressively. Here's an OEM Ford Ranger EV Pickup, and a EREV light truck (F-150).

Here are electric UPS trucks. Here is a hybrid bus. Here is an electric bus. An electric dump truck. Electric trucks have much less maintenance.

Kenworth Truck Company, a division of PACCAR, already offers a T270 Class 6 hybrid-electric truck. Kenworth has introduced a new Kenworth T370 Class 7 diesel-electric hybrid tractor for local haul applications, including beverage, general freight, and grocery distribution. Daimler Trucks and Walmart developed a Class 8 tractor-trailer which reduces fuel consumption about 6%.

Volvo is moving toward hybrid heavy vehicles, including garbage trucks and buses. Here is the heaviest-duty EV so far. Here's a recent order for hybrid trucks, and here's expanding production of an eight ton electric delivery truck, with many customers. Here are electric local delivery vehicles, and short range heavy trucks. Here are electric UPS trucks, and EREV UPS trucks. Here's a good general article and discussion of heavy-duty electric vehicles.

Diesel will be around for decades for essential uses, and in a transitional period commercial consumption will out-bid personal transportation consumers for fuel.

Mining is a common concern. Much mining, especially underground coal mining (where ICEs can cause explosions), has been electric for some time - here's a source of electrical mining equipment. Caterpillar manufactures 200-ton and above mining trucks with both drives. Caterpillar will produce mining trucks for every application—uphill, downhill, flat or extreme conditions — with electric as well as mechanical drive. Here's an electric earth moving truck. Here's an electric mobile strip mining machine, the largest tracked vehicle in the world at 13,500 tons.

Water shipping and aviation can also eliminate oil: see my separate post on that topic.

Here's a terminal tractor that reduces fuel consumption by 60%.

Farm tractors can be electric, or hybrid . Here's a light electric tractor . Farm tractors are a fleet application, so they're not subject to the same limitations as cars and other light road vehicles(i.e., the need for small, light batteries and a charging network). Providing swap-in batteries is much easier and more practical: batteries can be trucked to the field in swappable packs, and swapping would be automated, a la Better Place. Zinc-air fuel cells can just be refueled. Many sources of power are within the weight parameters to power modern farm tractors, including lithium-ion, Zebra batteries, ZAFC's and the lead-acid developed by Firefly Energy (before their demise), and others.

It's very likely that an electric combine would be an Extended Range EV: it would have a small onboard generator, like the Chevy Volt. Such a design would be more more efficient than a traditional diesel only combine, and would allow extended operation in a weather emergency.

Most farmers are small and suffering, but most farm acreage is being managed by large organizations, and is much more profitable. Those organizations will just raise their food prices, and out-bid personal transportation (commuters and leisure travel) for fuel, so they'll do just fine. As farm commodities are only a small %of the final price of food, it won't make much difference to food prices. The distribution system, too, will outbid personal transportation for fuel. Given that overall liquid fuel supplies are likely to only decline 20% in the next 20 years, that gives plenty of time for a transition.

Even hydrogen fuel cells could be used, though they're not likely to be cost-competitive soon with the alternatives. PV roofs certainly could be used to extend battery life, though the cost effectiveness of that will depend on how much of the year the tractor is in the field. Electric drive trains are likely to be much more cost-effective than liquid fuels, but locally produced bio-fuels would certainly work. Also, fuels synthesized from renewable electricity, seawater and atmospheric CO2 would certainly work, though it would be rather more expensive than any of the above.

Any and all of these is several orders of magnitude cheaper and more powerful than animal-pulled equipment. One sees occasionally the idea that we'll go back to horses or mules - this is entirely unrealistic.

The easiest transitional solution may be running diesel farm tractors on vegetable oil, with minor modifications. Ultimately, farmers are net energy exporters (whether it's food, oil or ethanol), and will actually do better in an environment of energy scarcity.

Iron smelting currently uses a lot of coal, which isn't oil, but is a fossil fuel which we'd like to eliminate. Iron used to be made with charcoal, and iron oxide can be reduced either with direct electrolysis, or with hydrogen from any source.  Eventually smelting will become much smaller - most of the steel used in the USA is reclaimed from scrap (and when industries mature, essentially all of their steel can be recycled);  an electric furnace can re-melt it, and the electricity can come from anything. About 30% of world steel production recycles scrap with electric arc furnaces ( ).

The US Navy plans to go reduce it's 50,000 vehicle fleet's oil consumption by 50% by 2015. They plan by 2020 to produce at least half of its shore-based energy requirements on its bases from alternative sources ( solar, wind, ocean, or geothermal sources - they're already doing this at China Lake, where on-base systems generate 20 times the load of the base), and it's overall fossil fuel consumption by 50% by 2020 with EVs and biofuel.  Here's a base that replaced on-base vehicles with EVs -

Some question the stability of the electrical grid, in an environment of expensive fuel. Utilities like the idea of "eating their own cooking". Here's an electric utility boom lift. Here's a consortium of utilities considering a bulk purchase of plug-ins (and a good article). Here's an individual utility buying electric cars. Similarly, utilities are buying hybrid bucket trucks and digger derricks. Here's a large commitment by two major utilities .

Here's a good quote from the Governor of Michigan: "For automakers, replacing the internal-combustion engine with an electric powertrain is both revolutionary and daunting. In a world where economic Darwinism threatens slow adapters with extinction, U.S. automakers know that they can either lead this historic transformation or become history themselves. Even today, as they engage in a struggle to survive, the Big Three are leading the way: General Motors, Ford and Chrysler are scheduled to introduce electrified vehicles next year."

France is planning for a market share for EV's of 7% by 2015, rising to 27% in 2025.


What if our current system is less like a train running out of power, where it will just slow down and stop, and more like a jetliner running out of power, energy which it crucially needs to have a safe landing? Do we really have the resources to build out an alternate energy infrastructure?

Well, at least in the US, there's so much energy used for things with very marginal value that we have a very big cushion. We have an enormous surplus of energy (used for single-commuter SUVs, for example) , so we have quite a lot of flexibility.

EVs don't require significantly more energy than ICEs to manufacture. Wind turbines have a very high E-ROI.

Even if PO reduces the energy we have available, we currently have such a large surplus that we have plenty of leeway to reduce consumption in some places to free up the oil needed for such an investment.

Isn't this a tricky transition, with fragile balances between politics, communications, labor, logistics, public-calm, etc?

It's true - a transition away from oil will put stress on a lot of institutions. On the other hand, this isn't any bigger than similar transitions, like going from coal to oil, or from mules to tractors. And, isn't it good to know that there technical solutions?

Where will the needed electricity come from?

From wind, mostly. Wind has a very high E-ROI, and is plentiful. Solar, nuclear, geothermal, etc will also be important. Coal is extremely abundant, but we have to hope that we don't use it.

Aren't we going to have to live within the limits of our environment?

Sure. Fortunately, energy isn't one of those limits. I'd say that climate change and species extinctions are much larger problems.

What about the invested-in infra-structure for our oil-based life style and what it will take to tear down the old infra structure and replace it with an entirely different one? Won't we have to tear down the suburbs, and similiar infrastructure?

Yes, we'll have to toss out some ICE trucks and cars before the end of their natural lifetime. On the other hand, we do that all of the time: the average US car/SUV/pickup gets 50% of it's lifetime mileage by the time it's 7 years old. They could last 25+ years, if we wanted them to, but we throw them away. The premature retirement of commercial trucks will hurt investors in some trucking companies, but that's a sunk cost.

The real question is, can we afford to build new infrastructure, and the answer is clearly yes: new rail tracks and rolling stock aren't that expensive, and EVs are no more expensive than ICEs.

We won't have to toss out housing - Kunstler is just wrong, completely wrong. A Nissan Leaf will allow a 50 mile commute, or 100 miles with workplace charging.

EVs can be built with the same factories - for instance, the Volt shares a factory with 2 other cars. They drive on the same roads.

September 22, 2008

Is Chinese oil demand immune to prices?

No - Automobile sales in China in August 2008 shrank 6.3% year on year to 629,000 units, the first fall in about two years, due to higher fuel prices.

Chinese GDP growth has dropped by about 1/3 recently - see .

Chinese are much more aggressive than the US about replacement of oil-based electrical generation with coal and nuclear; energy efficiency (especially automotive); and PHEVs/EV's.

Regarding competing with China for imported oil: the US produces at least 40% of it's own oil, so a 20% reduction of overall consumption is a 33% reduction in imports. I would note that the US reduced it's oil imports by about 15% recently, even before this credit crunch hit.

China, already a global center for lithium-ion battery component production and battery manufacturing, is ramping up its research and development efforts in the field, both within the private sector and with government support.

September 13, 2008

Can we replace oil for shipping?

Sure - Long distance land shipping can go by rail (which is 3x as efficient, and can be electrified): local can go by plug-in hybrid truck1.

Water transport is even more efficient, and can find substitutes for oil.

Substitutes for oil for water shipping? Pshaw, you say.

No, really. Substitutes include greater efficiency, wind, solar, battery power and renewably generated hydrogen.

Efficiency: Fuel consumption per mile is roughly the square of speed, so slowing down saves fuel: in 2008, with high fuel costs, most container shipping slowed down 20%, and reduced fuel consumption by roughly a third. For example, Kennebec Captain's ship carries 5,000 cars from Japan to Europe (12,000 miles) and burns 8.5 miles/ton of fuel at 18.5knots, for a total of about 1,400 tons of fuel. At a 10% lower speed of 16.6 kts, the ship burns 21% less fuel (about 300 tons).

Size brings efficiency: the Emma Maersk uses about 320 tons of fuel per day to carry 220,0002, tons of cargo, while Kennebec Captain's ship uses about 60 tons to carry about 23,000 tons (see here ), so the Emma Maersk uses roughly 60% as much fuel per ton.

Other substantial sources of savings include better hull (I've seen mention of "axe cleaver" designs - anyone seen details?) and engine design (very large (3 story!)marine diesels can get up to 50% thermodynamic efficiency), and low friction hull coatings (the Emma Mærsk saves about 1.3% with special paint, and bubbles work too).

Container shipping fuel efficiency rose 75% from 1976 to 2007, in an era of very low fuel costs.

Finally, reduction of oil consumption brings a kind of reverse-Jevons efficiency. Currently, some 34% of shipping tonnage worldwide is devoted to transporting oil [source , p.16]. If we reduce oil consumption, we reduce the need for shipping. Similarly, world coal trade was about 718Mt in 2003 [source , p2], at the same time as total world trade was 6,500Mt, so that coal was 11% of world seaborne trade by weight.

Wind: kites mounted on the ship's bow have been shown to provide 10-30% of ship's power - this is cost effective now. See an early article the leading company, Skysails, a followup article showing a commercial implementation, and the Skysails website. These are retrofits: it is likely that far more wind power could be harnessed if the ship were designed to accommodate kite assist (stronger more integrated ship structure to tug upon) rather than merely retrofitted with it.

It's astonishing what can be done with modern materials, computer-aided design, and electronic control systems, to turn the old new again.

Solar: The first question is: is it cost effective? Sure - it's just straightforward calculations: PV can generate power for the equivalent of diesel at $3/gallon (40KWH per gallon @40% efficiency = 16 KWH/gallon; $3/16KWH = about $.20/KWH, or $4/Wp, which large I/C installations have already surpassed.

Ships, trains and planes are outside all of the time, so they'll have a decent capacity factor. Grid tied systems have to deal with Balance of System costs, but a panel in a vehicle should be able to eliminate most costs: it's manufacturing, which is far more efficient than grid-tied systems that require field installation; redundant support structures; and dedicated power electronics. If a vehicle can add a panel for $1 per Wp, and get just 5% capacity factor, it could achieve $.15/kWh.

Let's look at the Emma Mærsk . With a length of 397 metres, and beam of 56 metres, it has a surface area of 22,400 sq m. At 20% efficiency we get about 4.5MW on the ship's deck at peak power. Now, as best I can tell it probably uses about 10MW at 12 knots (very roughly a minimum speed), 20MW at 15 knots, and 65MW (80% of engine rated power) at 25.5 knots (roughly a maximum). So, at minimum speed it could get about 45% of it's power for something close to 20% of the time, for a net of 9%. Now, if we want to increase that we'll need either higher efficiency PV, or more surface area from outriggers or something towed, perhaps using flexible PV. You could add a roof, or you could incentivize 10% of the containers to be roofed with PV - they could power ships, inter-modal rail, inter-modal trucks...

Here' a fun example of a boat that's 100% PV powered, here's a company selling a general approach, and here's a nice pure-electric

"Solar-powered sails the size of a jumbo jet's wings will be fitted to cargo ships, after a Sydney renewable energy company signed a deal with China's biggest shipping line.

The Chatswood-based Solar Sailor group has designed the sails, which can be retro-fitted to existing tankers.

The aluminium sails, 30 metres long and covered with photovolatic panels, harness the wind to cut fuel costs by between 20 and 40 per cent, and use the sun to meet five per cent of a ship's energy needs.

China's COSCO bulk carrier will fit the wings to a tanker ship and a bulker ship under a memorandum of understanding with the Australian company, which demonstrates the technology on a Sydney Harbour cruise boat.

"It's hard to predict a time line but at some point in the future, I can see all ships using solar sails - it's inevitable," said the company's chief executive, Dr Robert Dane.

Once fitted, the sails can pay for themselves in fuel savings within four years, Dr Dane said. They don't require special training to operate, with a computer linked in to a ship's existing navigation system, and sensors automatically angling the sails to catch a breeze and help vessels along." Source

Batteries: Large batteries could provide most of the remaining power needed, to be recharged at frequent port stops, as used to be done with coal 60 years ago (that's why the US wanted the Philippines' military bases, and why they're not needed in the oil era). Let's analyze li-ion batteries: assume 20MW engine power at a cruising speed a speed of 15 knots (17.25 mph) or 20MW auxiliary assistance to a higher speed, and a needed port-to-port range of 2,000 miles (a range that was considered extremely good in the era of coal ships - the average length of a full trip is about 4,500 miles (see chart 8 ). That's 116 hours of travel, and 2,310 MW hours needed. At 200whrs per kg, that's 11,594 metric tons. The Emma Maersk has a capacity of 172,990 metric tons, so we'd need about 7% of it's capacity (by weight) to add batteries.

So, li-ion would do. Now it would be more expensive than many alternatives that would be practical in a "captive" fleet like this - many high energy density, much less expensive batteries exist whose charging is very inconvenient, but could be swapped out in an application like this. These include Zinc-air, and others. It should be noted that research continues on batteries with much higher density still, as we see here and here, but existing batteries would suffice.

Here's a hybrid car carrier from (who else?) Toyota.

Refrigerated storage/transport could go via electric rail, which can certainly be low-CO2; refrigeration can always be made more efficient with better insulation; and there are "reefer" units that are "charged" on land, using grid power, that don't need any inputs from the ship during transit.

Hydrogen fuel cells: they can't compete with batteries in cars, but they'd work just fine in ships, where creation of a fleet fueling network would be far simpler, and where miniaturization of the fuel cell isn't essential. If batteries, the preferred solution for light surface vehicles, can't provide a complete solution, a hydrogen "range extender" would work quite well.

Hydrogen has more energy per unit mass than other fuels (61,100 BTUs per pound versus 20,900 BTUs per pound of gasoline), and fuel cells are perhaps 50% more efficient, so hydrogen would weigh less than 1/3 as much as diesel fuel.

Electricity storage using hydrogen will likely cost at least 2x as much as using batteries (due to inherent conversion inefficiency), but will still be much cheaper then current fuel prices. Fuel cells aren't especially heavy relative to this use: fuel cell mass 325 W/kg (FreedomCar goal) gives 32.5 MW = 100 metric tons, probably less than a 80MW diesel engine.

Hydrogen would have lower upfront costs versus batteries, and a lower weight penalty, but would have substantially higher operating costs. The optimal mix of batteries and hydrogen would depend on the relative future costs, but we can be confident that they would be affordable. Here's a forecast of affordability in the most difficult application, automotive.

Here's a demonstration project on a small boat.

Are shipping lines working on this?

Yes. Here's an example:

"The Auriga Leader, operated by NYK Line, was launched in December 2008 and can transport up to 6200 vehicles. NYK Line has set a goal to reduce car carrier energy consumption by 50 percent by 2010 through solar power generation, ship operation improvement, redesigned hull form, propulsion systems energy savings and improved cargo handling."

I suspect that container shipping will be able to out-bid other uses for FF, like personal transportation, for quite some time. We'll see the gradual addition of direct wind propulsion, like the Skysails, along with engine electrification and the addition of PV.

What about nuclear propulsion?

It would work, though I would be skeptical that it could beat the alternatives on cost or speed of deployment.

Don't forget that commercial nuclear plants are built as large as possible to maximize cost-effectiveness. The US Navy doesn't have to worry about cost-effectiveness - it chooses nuclear not on a cost basis, but on an operational effectiveness basis (maximum range without refueling).

The US Navy maintains a rigorous, labor intensive, costly safety program. Per Wikipedia, "A typical nuclear submarine has a crew of over 80. Non-nuclear boats typically have fewer than half as many." The Emma Maersk, the largest container ship in the world, sails with only 13 crew members!

My litmus test for nuclear proposals is their effect on weapons proliferation, especially relative to the complete fuel enrichment cycle. Per Wikipedia, "reactors used in submarines typically use highly enriched fuel (often greater than 20%) to enable them to deliver a large amount of energy from a smaller reactor." This doesn't seem encouraging.

What about the NS Savannah?

It was designed as a show vessel, not a workhorse, but it was only a few years after it was decommissioned as "uneconomic" that oil prices shot well above its parity point.

That parity point compared operating cost (excluding 1950's era capital costs, maintenance and disposal, etc) of nuclear to conventional operating costs, including fuel oil at $80/ton in 1974 dollars. Non-oil alternatives will be more competitive.

What about air transport in this age of just in time supply chains?

I would estimate less than 5% of plane transport is represented by the kind of small industrial components that go by air. Air freight transport uses surprisingly little: fuel is only 10% of Fedex's budget, so a doubling of fuel costs would only raise Fedex costs by 10%. The ratio of fuel cost to product cost is probably .1%. If Fedex fuel costs were to go up by 3x, it wouldn't have any significant effect on the affordability of sending such a part by air. Prices won't exceed an inflation adjusted $150/b anytime in the next 30 years in a sustained fashion - other things would change to prevent prices going over that level, including reduced fuel consumption by personal transportation & commuting, and US economic stagnation. The kind of small industrial components that go by air will be able out-bid other forms of fuel consumption.

What about passenger aviation?

Aviation will use liquid fuel for quite some time, as there will be some oil for many years, some biofuel will be available, and it will always be possible, though perhaps expensive, to synthesize fuel. Eventually substitutes like liquid hydrogen will be substituted if necessary - aviation will have to be dragged kicking and screaming to it, of course.

Won't the transition from oil take a long time?

Let's take trucking: first, it has some time for the transition to rail: trucking's consumption is only 27% of surface non-rail transport. Personal transportation is by far the big user, and personal transportation is mostly optional consumption which will be out-bid by commercial users (optional includes anything not essential, such as commuting that could be replaced by carpooling, albeit with great inconvenience).

Let me say that again: the food-and-goods freight transport network of the modern world uses about 1/4 of oil consumption in the US. Light vehicles overall account for 45% of oil consumption: their utilization could be doubled with carpooling in a matter of months, freeing up whatever fuel was needed by the freight network.

2nd, the transition is already underway: intermodal shipping is replacing trucking, and the trucking industry is under a lot of pressure.

Finally, truck efficiency can be greatly increased: here's a report by the US National Research Council that finds large truck fuel efficiency increases are technologically possible and cost effective.

"The report also estimates the costs and maximum fuel savings that could be achieved for each type of vehicle by 2020 if a combination of technologies were used. The best cost-benefit ratio was offered by tractor-trailers, whose fuel use could be cut by about 50 percent for about $84,600 per truck; the improvements would be cost-effective over ten years provided gas prices are at least $1.10 per gallon. The fuel use of motor coaches could be lowered by 32 percent for an estimated $36,350 per bus, which would be cost-effective if the price of fuel is $1.70 per gallon or higher. For other vehicle classes, the financial investments in making improvements would be cost-effective at higher prices of fuel."

Regarding water shipping: it's cost advantage will allow it to outbid other uses for a very long time. Here's a back of the envelope calculation:

1400 tons of oil to ship 5000 cars - 1400 tons is about 9,800 barrels. At $80/bbl that's $784,000. That's $157/car. If the average car sells for $20,000, that's 0.8% of the cost of the car. That's not much.

Jeff Rubin says that the Chinese have already lost their advantage in manufacturing steel for export to the US because total shipping costs, of both the ores and the finished product,are now so high that domestic American producers are now in a very competitive position again, despite higher labor costs. Doesn't this show that higher energy will make shipping infeasible?

No, it means that in selected areas it will be uncompetitive, which is very different. Commodities like iron ore and coal are very low cost per pound, so that the cost of transportation is a significant fraction of it's value. In competitive industries, a small change in cost makes a large difference, and a rise in shipping costs can tip the balance between regions and manufacturers. Moving high value added products will be relatively unaffected even if energy prices increase by a factor of five or even ten, as the shippng is only a tiny fraction of the cost of a camera or a computer.

On the the other hand, where all manufactures are affected, perhaps because there is a single source on which all are dependent, a change in cost of commodities due to shipping will raise all costs slightly, but have little effect on demand.


1The truck comes with a fast charger, which takes it to 80% charge in about 1 hour. The 6 hour charging time is for the remaining 20%. It might be fast charged during lunch, and slow charged overnight, giving a daily range of perhaps 145 miles ((72+90)*90% of max range).

The battery pack of 280kWh gives about 3.1 kWh/mile. With dense city driving a heavy diesel truck probably doesn't get more than 20% engine efficiency, which would give about 2.6MPG.

Here's a breakeven cost analysis:

miles (AM) 81.0
miles (post-lunch charge) 64.8
total miles per day 145.8
days per year 286
MPG 2.57
gallons per year: 16,216
10 years: 162,162
Fuel cost $1.43
Cost per day $81.20
10 yr cost $232,227

battery cost premium: $100,000
10 yr amort: $142,378

kWh/mile 3.1
pwer/day 453.6
cost/kWh-night (I/C) $0.037
cost/kWh-day (I/C+demand charges) $0.110
Cost per day $31
10 year cost $89,850

Net cost: $0

2 this figure conflicts with the wikipedia figure - IIRC it came from the Maersk line fact sheet, which is temporarily unavailable.

July 16, 2008

New developments

New developments

This blog provides a reference - a FAQ. As I learn things, I add them to the individual articles, whose posting dates don't change as they are updated. People need a way to know what's new, so, they'll be here in this post. It will be at the top unless I add an entirely new post.
EV technology, while more than adequate, will continue to improve. Regenerative braking eliminates one source of waste ( or finds a new energy source, depending on your perspective) by capturing vehicle kinetice energy: another source is the vertical kinetic energy now lost to shock absorption, which appears to have been solved. posted 11/23/08

July 7, 2008

Are plug-in's economically justified?

Yes. Plug-in's cost the equivalent of about $3/gallon with lithium-ion. As the price of gasoline rises above that level, plugins become more and more compelling.

First, let's simplify, and assume we simply add additional battery capacity to a Prius (and a plug, which is trivial). I would argue that a serial hybrid (like the Chevy Volt) is less complex than a parallel (and therefore less expensive in large, mature volumes), but that's not necessary to demonstrate the point.

OK, at 45MPG and $3/gallon, a Prius costs 6.7 cents/mile.

Now, good quality cobalt-based small-format batteries, as used in the Tesla, cost $400/KWH. Iron-phosphate is less expensive, and large formats are less expensive. The plug-in Prius is planned by Toyota for 2 years from now, which gives us another 2 years of the normal 8-10% annual cost reduction seen with li-ion's. Large scale PHEV battery production will instantly raise the volume of production for these 2nd generation li-ions to very large levels compared to conventional li-ion, reducing costs further. That gives us a reasonable forecast of $300/KWH (if this seems too aggressive, perhaps you'll grant that this is very likely several years later, when PHEV's have gone beyond early adopters, are ramping up to much higher production volumes, and batteries are that much more mature).

A123system's batteries can handle 5,000 discharges at 100% depth of discharge. If we assume 250 per year we have a battery that will last the life of the car. At a 10:1 capitalization rate (to account for interest, depreciation and obsolescence), we're paying $30/KWH per year, for 250 discharges, or $.12 per KWH discharged.

At .25kwh/mile, that's $.03 per mile, less than half the Prius cost. If we double the battery size to account for GM's conservative decision to only use 50% of the battery capacity (this is similar to the Prius, and almost certainly unnecessary, but GM's taking no chances at all), we're still at $.06/mile.

Now, charging will be done almost exclusively at night. Utilities are required to offer time-of-use power pricing by the 2005 energy statue, but many don't publicize it. OTOH, PG&E, SCE and Exelon are pushing it. PGE&E's night time pricing (like most CA electricity) is more expensive, starting at $.08/KWH, OTOH gas is also more expensive there.

So, add $.04/KWH for night time electricity (and 4 KWH/mile) for a cost of $.01/mile for power, and we're at $.07/mile, or rough parity.

Of course, taxi's and other fleet operators are likely to recharge more than 250 times per year, dramatically raising payback. For the average driver, add in CO2 costs, and other external intangibles like independence from the ME, and the ability to weather gas shortages and you have a compelling case.

And that assumes $3 gas, and a Prius as a benchmark - $4.10 gas, and a 22MPG vehicle (the US average) would make the case that much more compelling.

How about battery reliability?

Every report indicates that both batteries are doing extremely well. Could there be problems? Well, A123systems batteries have been in the field for years. It's competitor is an extremely large, competent company. Finally, we have redundancy in the two competititors. GM wants a very high level of assurance (as they should), but there's no question that these batteries come very close to their highly demanding specifications, and very little chance that one of them won't meet those specs entirely.

What about charging all these cars - do we have the power?

Pacific National Lab's study, showing that we don't need new power plants, is here.

Using a wide boundary analysis, including materials, employees, marketing via dealerships, and spending of profits, how much oil is used 'per GM-Vol' (not in driving it, but in making it)?

Not a lot.

First, an ErEV (extended range EV, aka plugin/PHEV) like the Volt won't cost any more than an ICE vehicle, or require more energy to manufacture. The high initial prices you hear about are a combination of R&D (which is very low energy - think high paid engineers in front of LCD displays - their biggest energy consumption is the junk food they eat as they work 18 hours per day) and GM exaggerating costs to justify an initial early-adopter premium, and lobbying for tax credits (or possibly they're just being very conservative - they're now saying that they're figuring in the cost of a replacement battery under warranty, which is pretty silly given the over-engineering already in the battery design - only 50% depth of discharge for a li-ion chemistry which can handle 100%, unlike, say, the NIMH being used by the Prius).

2nd most manufacturing (with the main exception of smelting, like PV silicon or aluminum) takes relatively little energy, and very little oil - what energy is consumed is primarily electricity, making irrelevant the claims one hears about BOE's (barrels of "oil equivalent") consumed by car manufacturing.

Employees don't take much oil: I suppose it would mostly be commuting gasoline. 1st, there aren't as many hours as one would expect in cars, as they're pretty high-wage jobs even now - perhaps 100 hours of assembly time, for instance, which is only 12.5 shifts of work. 2nd, they could drive a Volt...

Materials: Steel is the major material input. It's mostly recycled: more than 50% of inputs, and 95% of scrapped cars are recycled, so that energy isn't wasted. Even so, the energy input isn't enormous, and finally, it doesn't come from oil: it's a combo of electricity and coking coal (I have a steel mill a mile from my home: it operates at night when electricity is cheap).

Marketing is high paid labor and TV, which is powered electrically.

Profits? What profits?? Don't forget, this is GM we're talking about - the external costs of the spending of profits are the very least of their problems...

Will the Volt be too expensive?

No, not in the long-term.

GM says their latest price estimate of $40k for the first year model includes the cost of 2 batteries! They're including a warranty replacement of the battery, just to be ultraconservative (see ). Realistically, it's just a way to exaggerate the price in order to capture as much money as possible from early adopters, given that the first year demand will greatly exceed supply, yet GM doesn't want to appear to be gouging customers. Also, they want to encourage tax rebates and discourage increases in the CAFE regulations.

Lutz said the following: "We're being conservative on battery life. For our cost calculations we're assuming each car will need a replacement during the warranty period." original source:
GM is being ultraconservative on design - they're using 50% depth of discharge, where Tesla is using 100%, even though GM's cell chemistry has about 10x the cycle life in bench tests (at any given depth of discharge). They're assuming 2 battery packs during the life of the warranty, where Tesla is assuming 1 (of course, the warranty is longer, but that's GM's (conservative) choice).

There's no way this car can cost $40k to produce, unless they're using very, very unusual ways of applying R&D overhead and warranty costs. Heck, the Prius also has 2 (more complex) power-trains, and it costs roughly $20k less to produce than the $40k figure. The battery, in volume, should cost far less than $10K, so where does this premium come from?? The answer: GM is front-loading R&D costs, and exaggerating warranty costs, for the reasons I gave above.

Is the Volt development still on track?

"the company is "happy" with the capacity and performance of the batteries. GM also knows what the cooling system will look like and has physically integrated the pack into the vehicle. What's more, the entire propulsion system, including the battery pack, the electric motor, and the generator, was incorporated into a test vehicle and delivered to the company's Milford, MI, testing grounds at the end of August, just two days behind the schedule set last year.

"I wouldn't say that the battery is ready," Cesiel says, "but we're right on track." source:

What about lithium supplies?

Lithium is pretty abundant, and can be found in a lot of places, including China and Australia. Here's an article about the world's largest producer of one form of lithium ore:

Here's an article about li-ion battery costs - it's a bit outdated, but it provides a lot of detail.

What about hydrogen fuel cell vehicles instead?

A short answer: a hydrogen vehicle is an electric vehicle, like the Chevy Volt, which uses a smaller battery and a fuel cell to generate it's electricity. It will always be better to use a larger battery which is charged from the grid, combined with a small, cheap ICE engine for occasional backup. Here and here is the same information in much greater detail.

Why are some manufacturers, like Honda, still pursuing them?

Here's a good answer, posted by "Pangolin" as a comment on my first reference above: "I suspect that Honda's strategy was to develop electric car components and systems using hydrogen-vehicle research subsidies and have them on tap for fleet conversion to all-electric or plug-in vehicles. Toss the fuel cell and the hydrogen tanks and install larger batteries and you have an electric sedan. Throw in a small Honda generator and you have a plug-in hybrid. Off the shelf, every bit of it.

The nasty bit from Honda's point of view is that they will sell fewer vehicles. The only thing that could make a Honda more reliable would be to give them electric drive trains. That makes cars more of a long term investment as components would become swap-able. It was best to get all the bits right and wait until they were forced to make the shift. "

What's the future?

There are going to be a serious surge of PHEV/EVs by 2010, including GM's Volt, Toyota, Nissan, Chrysler, and other large car manufacturers. They're available now, in small but growing numbers from several small manufacturers, like Think and Tesla. Hybrid (the transitional form in the evolution to EVs) are now available from almost every manufacturer, including producers that would appear to be less vulnerable to customer demand for lower operating costs, such as BMW.

Electric vehicles (either PHEV's or EV's) faced serious barriers to entry in the form of very large investments (capital, emotional, career, etc, etc). This period of prolonged high oil prices will provide the impetus to push through this barrier. Once the barrier is crossed, costs will come down due to economy of scale, and PHEV/EV's will be forever entrenched. They are likely to follow ever falling cost curves, and largely replace fuel-based transportation.

How this will play out for the whole world (and the race against global depletion curves) is a tough question, as the US is the clear leader in adoption of hybrids, PHEV's and EV's (Japan sells them, and China is developing PHEVs and EVs), but I see fundamental change ahead for the US.

Here's a nice article about coming plug-ins. I'm also curious about retrofits - this appears to only cost about $4,000, and provide around 15 miles of electric driving. It isn't available yet (the promise fall 2008), and they haven't addressed regulatory issues, but the tech looks plausible. Hybrid taxi's are growing quickly, and Electric taxi's are coming.

EV technology, while more than adequate, will continue to improve. Regenerative braking eliminates one source of waste ( or finds a new energy source, depending on your perspective) by capturing vehicle kinetice energy: another source is the vertical kinetic energy now lost to shock absorption, which appears to have been solved.

July 5, 2008

What's needed, nationally and personally?

On the national level:


We can create policies that mandate or promote carpooling, telecommuting and videoconferencing. Telecommuting in particular could save 10B gallons of gasoline per year (7% of total consumption, or much more than ethanol) or more, with no pain at all. Sadly, business culture isn't quite ready for it - a rapid expansion of telecommuting will take a real push. Paradoxically, everyone would be better off. Similarly, telecommuting can be better than solo driving.

One painless and important change: we can immediately end policies that favor free parking, which effectively subsidize driving. One study estimates that up to 25% of commuters who receive free parking would stop driving alone to work, if their employers allowed them to receive the value of their free-parking in cash. Cities should charge more for parking (while reducing other business and citizen costs, such as sales or other taxes or fees, to avoid becoming un-competitive) and switch from requiring businesses to provide free parking to requiring businesses to charge for parking. This would encourage mass-transit and ride-sharing, and immediately save fuel (as much as 50% of city driving can be looking for parking!). This won't harm consumers: customers will find shopping much more convenient, and businesses will have more money to encourage shopping in other ways.

More transparency for futures trading would be helpful, and slightly reduced leverage (greater capital requirements) would help reduce any bubbles. Most of the speculation-related proposals would do more harm than good. For instance, elimination of the ability of institutional investors to trade futures would drive some to less transparent foreign markets. Elimination of trading by investors (those not taking delivery) would do a great deal of harm - at minimum it would make hedging much more difficult for many energy consumers.


Obviously, we need to dramatically raise fuel taxes - the best way would be a gradual increase over 5 years, with revenues rebated per capita to individual and industrial/commercial consumers. This will be very difficult to do before a true emergency, but the rebate would help sell it - like social security, it would spread the benefit in a reasonably progressive way.

Fuel taxes are much better than emissions-trading. Emissions-trading requires complex legislation and bureaucracies, and endless expenses in brokering. Sadly, carbon and fuel-taxes are much harder to sell, precisely because they're simpler, more transparent and more effective, which makes them seem more painful.

We need to raise the CAFE dramatically and provide rebates for plug-ins and electric cars (assistance to Detroit for retooling, and socialization of Detroit's pensions and healthcare costs wouldn't be a bad idea - the alternative for Detroit is bankruptcy, which will achieve the same ends in a much more painful way), and expand non-fossil fuel electrical generation ASAP. Obviously, the wind and solar federal programs need to be extended, with a 7 year gradual phaseout. Expanded mass transit is a good idea, though it's far from a silver bullet. National model building codes should be revised for much greater efficiency, and pushed into local government.

We need greatly expanded research into bio-fuels from cellulose and algae. Algae in particular has a great deal of technical potential. Cellulosic fuels will always be limited in scale, but improved processes could reduce their cost and environmental impact. Coal To Liquids with CO2 sequestration should be offered loan guarantees, to allow greatly expanded private investment without the fear of another Synfuel-type disaster (caused by the 80's oil price crash ).

I suspect that reliance on diesel is a blind alley - it's more efficient than gasoline, but only a little (a gallon of diesel has more energy, and carbon, than a gallon of gasoline), and it's success in Europe is dependent on subsidies. The US hybrid->plugin->ErEV->EV path seems much better in the longrun.

Here's a good article on these questions by Andy Grove, the ex-CEO of Intel:

What can we do personally?


Better driving can save 20%. Try to take the train, carpool, or ask about telecommuting. The next time your work requires travel, ask if this can be done by teleconferencing. If not, ask if there are any plans to make it possible. All of these can make life easier, and at the same time cut costs dramatically.

Reduce home energy consumption: we all know the simple and effective things, like CFLs, plugging air leaks (windows, doors, attics, electrical outlets, roof, joints, chimney and walls), and programmable thermostats. Spot/room heating with electric space heaters is very cost-effective. Space heater infrared radiated at people (say, at one's desk or kitchen counter) is even more efficient, since it allows one to feel warm even if the air is cool. Comforters allow lower winter temperatures, and fans (ceiling and desk) allow higher summer temps. Simple, cheap, kits allow you to add a layer of insulation to windows with a sheet of transparent plastic.

How does saving electricity and natural gas help with oil/gasoline prices (for those of us not heating with fuel oil)? In the US, natural gas is used for both transportation and electrical generation, so NG connects the electricity and oil markets. Outside the continental US, oil is used for electrical generation.


High-mileage drivers can buy used high-MPG cars easily and quickly, such as a Corolla (40MPG highway, but as little as $3K for a reliable 10 year old model), or a Honda Insight (gets 60-70 MPG, but still only $10k). Get rid of an SUV, and rent it back when needed for heavy cargo. Look into carsharing.

At home:
-install better windows (we added two more layers to our thermopane windows, and now don't need the furnace until outside temperatures are below freezing),
-replace appliances, including refrigerator, furnace and A/C, with higher efficiency models.
-Consider a heat-pump - new air-based heat pumps work very well, and are very cost effective, compared to any other form of heat.
-If you heat with fuel oil, simple resistance electricity may be cheaper for whole-house heating, and is relatively inexpensive to install. For instance, $4.50 #1 fuel oil is more expensive than 13 cent electricity (the national average is about 10 cents). There are about 35 KWHs per gallon (at 90% combustion efficiency): divide your cost per gallon by the $/KWH - if it's higher than 35, you'll save.
-Natural gas is almost always cheaper than simple resistance electricity (the necessary price ratio is 264, and we're only at about 130, on average).

Don't forget that these things help reduce household costs, CO2 and oil/gasoline prices (a threefer!).

When looking for a new job, give preference to closer jobs - this will pay many dividends beyond reducing fuel costs. Look for homes close to rail, or to work, if possible.

July 3, 2008

What will oil do in the short-term?

There are many dynamics operating in oil markets, including geological limits, capital expenditure lag-time, geopolitics, currency valuations, technology, consumer psychology, institutional resistance to change, and speculation. It is reasonably accurate to say that supply and demand got us to $100 oil in the last year. This is well explained at Econbrowser.

$100 oil was sufficient to flatten out oil consumption (the US DOE calls this "demand", though this word usage drives economists crazy) for the last several years, despite the overwhelming media message that high prices were temporary. Lately supply (and consumption) has increased slightly, which should have helped keep prices at the same level. Instead, prices are jumping - from the viewpoint of supply and demand, it makes no sense. For instance, David O'Reilly, chief of Chevron, is baffled by the extremely high oil prices seen recently. "We're surprised. We can see how you can get to $100," he says. "At $140, I just don't know how to explain it." (NYT 7/5/08) .

Could speculation raise oil prices beyond $100?

Sure. Fundamentally, futures markets are a marketplace to order commodities for future delivery at a fixed price (even if they're used as a casino). If many new participants place new orders, we have new demand, and prices will go up, so there's no question that speculators can raise prices by "going long" - in other words, betting that prices will rise. Of course, they can also lower prices by going short. In the most benign scenario, speculators provide a socially productive service by "price-finding": pushing prices up (or down) until supply and demand are balanced. If speculators see a looming shortage, they bring that shortage forward to the present, and accelerate the necessary economic adjustment. This is good, though the process is by trial and error, and can easily overshoot and oscillate around the ideal price, causing unpleasant volatility. On the other hand, it is common for speculators to "herd" - IOW they mostly choose one side or the other of such a bet (this is often what is called a "momentum play", aka "the greater fool" theory), and this can create a much larger and very unpleasant overshoot, and an excessive adjustment afterwards, such as we are seeing in the housing market.

An analysis at Econbrowser suggests that a small bubble is possible. Some people say that the effect has to be temporary, as eventually delivery must be taken or the contracts sold, but new paper bidders (speculators) can surely raise futures prices temporarily (IOW, create a bubble), and spot prices will rise in tandem due to arbitrage. Even if it is only temporary, most futures contracts are measured in years, so why can't we see a bubble for a year or even more? Further, if existing investors roll over their bets, and new investors continue to arrive ("greater fools") the bubble can grow for a significant time.

So, are we creating a bubble?

It seems to me that, that fundamental supply and demand are the problem, but there's no question that there's a lot of speculative money in the market. Even though price-finding is probably a healthy thing from an economic viewpoint, that can easily overshoot. Further, there's a lot of money that is herding (mostly, apparently, in the form of Exchange Traded Funds), so a bubble due to momentum-play speculation is also reasonably likely, and these investors seem to be mostly going long.

I suspect that current prices are unsustainable: higher prices stop consumption, even in China. We see that China had to cave in, and raise price controls, even before the olympics, which is very likely something that they really didn't want to do. Despite a constant linkage in media reports of Chinese and Indian consumption, India's consumption has been flat for several years. Brazil has become an exporter, due to increasing oil production (with a greatly exaggerated assist from sugar cane ethanol). The highest prices are being seen by the US, and countries whose currencies are tied to the dollar, including China and Japan.

Short-term demand elasticity is much smaller than long-term. In other words, people don't reduce consumption if they think high prices are temporary, and reductions are much easier over time, as people make routine capital expenditures such as car purchases. It's a serious mistake to think that people and businesses don't respond to oil prices. Overall global demand is unsustainable at this price level, and will fall until prices decline substantially. The level at which demand matches the current plateau of oil production is probably around $100-$120. I expect to see more stories like this describing a short-term decline to more sustainable prices.

Wouldn't we see accumulating oil storage in the form of rising inventories?

First, that assumes increasing production or falling demand. Oil export show no sign of responding strongly to prices (outside of a small increase from Saudi Arabia), and it would take some months for declining demand to have a strong effect (there are a number of sources of delays, including the time required for oil product distributors in China and India to run out of money because of price controls, the time required for pundits in oil-consuming countries to admit to the public that prices are going to stay high, and the delay in reporting production and consumption statistics), so a bubble measured in months wouldn't have a visible effect. Second, the speculators don't have to store it (and refiners currently consider oil far too expensive to store - they're minimizing inventories just to save money). The logical place for storage is by the sellers who have promised future delivery. AFAIK we have little info about National Oil Company storage, and in the final analysis NOC's can store oil in the ground. The King of Saudi Arabia has recently talked publicly about doing just that ("saving our oil for future generations").

Are NOC's (like Saudi Arabia) manipulating prices?

They certainly could - they have more than enough money. I think that Saudi Arabia is smart enough not to do that, as it would hurt them in the long-run. Sadly, not all oil exporters are as smart as Saudi Arabia. The leading regulator thinks manipulation is possible : "The acting CFTC chairman told the panel that it is imperative that strong enforcement actions be taken in order to prevent illegal manipulation of the commodities markets, noting that the markets are "ripe" for such manipulation. "

Do we know for sure?

No, there's inadequate data on trading and and inventories outside the US. We have to rely on our analysis of fundamentals, and read the entrails of snippets of information about the activities of traders. Here is a good discussion of how little we know.

Note: I remember seeing the heads of the commodities exchanges quoted as saying that oil speculation is a productive price-finding process, in effect bringing future shortages to the present. I can't find a reference - has anyone seen that?

June 25, 2008

Medium-term, is this another 80's oil bubble?

No, not if you define medium-term as less than 10 years.

The oil crisis of 1973-79 was caused by OPEC - the US had lost it's ability to increase oil production (as "swing-producer"), and OPEC tried to corner the market. Unfortunately for them, they tried too early - when consumption fell, and non-OPEC production increased, OPEC couldn't reduce production sufficiently to support prices (it's interesting to note that OPEC tried before, in 1967, to choke off supplies, and was completely unsuccessful - it was much too early).

With the benefit of hindsight we know that the 70's oil crises were "geo-political", but at the time that wasn't so clear - even where this was recognized, it was assumed that very high prices would continue (which they would have, had OPEC been able to manage it). Paradoxically, the 70's "energy crisis" caused an influx of new supply, and a great deal of efficiency and conservation, precisely because people thought this was a long-term problem. If everyone had thought prices would crash in the 80's, no one would have acted as decisively as they did.
As a result, many people got burned. Oil companies invested in projects that didn't pay off. Individuals "relocalized" and forewent children, in anticipation of resource poverty that never arrived.

It's simply not realistic to dismiss the question "what about another 80's style bubble and crash?" by saying that people should have known back then that things weren't that bad, and arguments that "this time is different" need decent proof.

On the other hand, I think it's clear that the current problem was caused by demand exceeding supply, starting very roughly in 2005, rather than primarily by 70's style top-down decisions by OPEC. Whether this is a classic commodity boom-bust cycle caused by lagging capital expenditures, or is caused by Peak Oil, it isn't going to be over quickly: large oil ventures take 5-10 years to take from the inception of exploration to discovery to production, and the oil services industry has been decimated by historically low oil prices in the last 20 years, and will have a hard time catching up to demand.

Now, we don't know if OPEC is contributing somewhat to this shortage, or if they're simply covering up their inability to meet demand, but there's very little question that they're pretty happy with prices above $70 (at minimum), and would defend them. These days it looks like they're happy with $100 oil, and would very likely defend that higher price level. The gradual loss of a surplus supply cushion means that OPEC would be successful in such a defense.

There is a real risk that it won't be over for 20 years: Peak Oil theory (which is explained here in a PDF), and the Export Land Model (as explained by Jeffrey Brown and "Khebab" in their blog) suggest that quantities of oil available to importers such as the US may fall quickly, accompanied by dramatic price increases. See figure 17 in the second website for a range of projected exports for the top 5 oil exporters - these 5 currently account for about 50% of all oil exports (this doesn't include smaller exporters, some of which, like Canada, are likely to increase exports, but these big exporters are important). The chart suggests that there is a real risk that oil exports by the biggest oil producers could drop in half by about 2019.

Here is a recent mainstream perspective.

We could also face a sudden loss off all oil exports from the Persian Gulf due to, say, someone bombing enrichment facilities in Iran, or civil insurrection in the region (monarchies aren't the most stable of governments).

What could be done?

Well, if US imports were to drop by 50% in 11 years that might reduce overall US oil consumption by 30% (we import 57% of consumption - about 11.6M bpd vs consumption of 20.2M). In 1978-1982 the US reduced it's oil consumption by 19% while growing slightly, for a reduction of about 4% per year. At that rate a 30% reduction would require 8 years. So, we could handle it, though possibly with stagnating GDP for some years.

If oil imports were to fall more quickly, people would at some point be ready for real emergency measures. Well, emergency measures could easily reduce consumption by 25% in 6 months by conservation (just make all highway lanes HOV, strictly enforced), and drilling (in ANWR and off the coasts) and large-scale CTL could both be done in 3 years under truly emergency conditions.
Many analysts project conditions that reflect a true emergency, and assume Business As Usual responses. That makes no sense.

On the other hand, none of these are fun scenarios. It's urgent that we, at a national and personal level, start planning ahead, and start reducing consumption ASAP, as well as trying to increase production.

What if China's buying power continues to rise versus that of the US?

That's likely to continue, but how large will the effect be? Keep in mind that Chinese consumers have a lot of alternatives for their buying power, and expensive oil is more expensive than substitutes everywhere, not just in the US. higher prices stop consumption, even in China. We see that China had to cave in and raise price controls, even before the olympics, which is something that they surely didn't want to do. Because of those price controls the effect on Chinese consumption was delayed, but I expect to happen just like anywhere else. Don't forget, much Chinese oil consumption is for diesel electrical generation - we have to think that's unlikely to continue for long in these volumes at these prices.

What about domestic production, which depends on when ANWR and OCS get opened up?

I suspect that ANWR, OCS, as well as CTL, could be done quickly (in very roughly 3 years) in a true emergency, in which environmental precautions are thrown to the wind, various processes are done in parallel, and cost is no object. Perhaps we'll continue to do our frog-boiling thing, but one way or another I think we'll reach a breaking point on domestic production, including CTL, which will both start these projects, and accelerate their development.

Here's an interesting take on OCS. It refers to an EIA study that says that there is quite a bit of oil (41B barrels) in the OCS that is already available to oil companies. That suggests that low oil prices prevented production in these areas in the past, and that scarce resources (rigs, manpower, etc) are now preventing oil production that is likely to be arrive eventually as the oil-services industry expands. The EIA indicates that there are another 18B that would are now legally unavailable, and that require high prices to be viable.

Aren't existing Alaska, onshore and existing offshore going to go down?

That's not at all guaranteed - I'm including new wells in old fields in "existing". Tertiary methods, especially CO2 injection, have a lot of promise. There's an enormous amount of oil to be extracted in the US. The Bakken alone has upwards of 200B bbls. Only about 4B are economic now, but that's a $30T incentive - there's a likelihood of more to be extracted there, with a significant chance of a great deal more.

Hasn't Hubbert's projection for the lower 48 been very accurate?

That's been exaggerated a bit. Keep in mind the domestic price controls before and during the price peaks, and the crash in prices immediately thereafter. Sure, drilling went up, but how much earlier, higher and longer-lived would that drilling increase have been if the domestic oil industry had received a strong and persistent price signal? Also, Hubbert was far from prescient: he made a similar projection for natural gas that was completely wrong (he projected a crash roughly in the 1980's, while NG production is still fairly stable).

how much do you think US production will decline in the 2010s?

I'm not sure. It will increase when Thunderhorse goes into production, and the EIA projects an substantial increase in the medium (though we don't trust EIA projections much). What I do know is that as discussed above there is a lot of oil to be extracted in the US (a lot of decent prospects are known and just waiting to be exploited), and that the drilling services industry will ramp up with time. I would be very surprised if production fell, and I expect at least a small increase.

Didn't the 1978-1982 period included declining oil prices in the last couple of years?

If prices had been higher there would have been an effect on both oil consumption and GDP - how much we don't know, but we have pretty good evidence that a 4% annual decline can be done without declining GDP.

Didn't we also have a much better trade balance?

Actually, if we exclude oil we have a pretty good balance right now - oil is the problem. I agree it's a big problem - right now we're selling the family silver and hocking our furniture to pay for it. OTOH, if oil exporters are sensible, and recycle petrodollars, we'll have some mix of stagnating domestic consumption (transferred to exports) and go into debt, but not suffer declining incomes.

Won't oil prices go far higher, in inflation-adjusted terms, in the 2010s?

I think current prices are the ceiling for several years, and that $200 oil is a maximum. Roughly 50% of world GDP has their currencies tied to the dollar (including the yen and yuan), and price signals will indeed work.

Don't efficiency increases get harder as you go along?

Not really. Telecommuting will have benefits for all, once we get past the cultural barriers (it's kind've like the Norsemen in Greenland who starved rather than eat fish). Carpooling is an inconvenience, but very, very cheap.

Don't hybrids cost thousands of dollars more per car?

The Honda Insight was cheap, and got 60-70MPG. The Prius (at $24k) is cheaper than the average US light vehicle (at $28k). The Volt will be below $30k easily, with volume production. NEV's can be $15K.

Didn't it take over 20 years for the US to improve by 34% in the 80's and 90'?

Keep in mind that this was in an era of low prices, with efficiency an afterthought. Now it's a high priority .

Won't the oil availability drop be too steep for orderly and timely adjustments?

Probably. The real question is how disruptive this will be. We know that there will be premature obsolescence for a lot of capital (like SUV's, and hopefully coal plants). I just think that a depression is not likely. It's a risk that our policymakers should be paying much, much more attention to, but not likely.

Isn't one of the reasons that the energy intensity of OECD economies has gone down is that energy intensive activities have migrated to less developed countries?

Keep in mind that the energy intensity of the whole world, and especially the "oil intensity" of the whole world, have increased at roughly the same rate. World GDP has been growing at 5% per year for the last several years while oil production has been flat. This is much more a problem of recycling petrodollars than it is of a lack of energy or oil.

What about Saudi Arabia - aren't their exports going to fall dramatically, due to increasing consumption and falling output?

Probably not, though it is a real risk. If we take a look at the best known analysis of this type we see: "Our middle case shows Saudi Arabia approaching zero net exports in 2031, within a range from 2024 to 2037." That's only 23 years away, and is entirely unrealistic: it doesn't take into account the differences between KSA and other producers (KSA is much, much more carefully managed from the top-down); it projects out current consumption growth without change; and it suggests that KSA wouldn't do anything to maintain exports even as it lost all export income.

Hubbert linearization is useful, but only as a preliminary guide - it can't be elevated to the status of supernatural authority. As a very good example, Hubbert's 1970's prediction for Natural Gas in the 80's was completely wrong: he projected that NG would fall off a cliff, while it has stayed very stable for another 30 years. The same criticism applies to the ELM model - it's just a preliminary, rough guide, and needs a great deal of finetuning.Look at his use of historical US production numbers without an acknowledgement of the problems with extending this analogy to the world: US price controls, and import competition. Look at how badly Hubbert's prediction for Natural Gas missed the mark.

KSA's per capita oil consumption is just behind the US's, so it's growth is likely to slow soon. A great expansion of consumption would mostly require industrial energy disintermediation by KSA, with continued net energy exports. In other words, they'd start refining oil, manufacturing fertilizer and smelting aluminum. Those things would be largely exported (or displace imports), so that oil & gas exports would fall, but that would reduce energy needed for those products elsewhere, so the net effect would not be to starve the rest of the world of energy.

June 21, 2008

Is solar power a real solution?

In short, yes.

This can be broken down into several questions:

Is solar too expensive?

See this article for analysts who foresee solar becoming as cheap as grid power in 3-5 years.

Also, take a look at First Solar's last quarterly report, which shows PV costs at $1.12 per peak watt (which is very cheap), and that it fell 12% from 1 year ago. I believe their thin-film has efficiency which is a bit below the average for conventional silicon, so Balance of System costs will be a little higher. Their wholesale panels are being sold for about $2.50/Wp, which suggests complete systems, installed, at about $4/Wp. That, in turn, suggests about $.20/kwhr for large, industrial/commercial installations.

Solar costs are now around $.30/kwhr (for retail, rooftop PV) in ideal locations like Southern California, and it looks like large, industrial/commercial installations are indeed achieving about $.20. Given that solar competes with retail electric rates, this is competitive (meaning a total cost that is lower than paying for utility electricity) without subsidies for many customers in S. California. I should think that with continued growth and competition we could expect to see wholesale panels at $1.50/Wp in perhaps the next 3 years, and complete systems, installed, at $3/Wp ( This story suggests costs of less than $2/Wp in several years), which would give us $.15 and grid parity generally in Southern California and many other places. Of course, as long as we have high subsidies we will see very high growth rates, elevated prices and incredible profits - already we're seeing Chinese solar billionaires. It could take a while. OTOH, silicon is also coming down in cost very fast - big producers like Sharp are being very aggressive about this, in order to maintain profits and market share. Ultimately, of course, it depends on the subsidies. As long as Germany is paying around 40 euro cents per KWH, that's how it will be priced in Germany!

At some point, as volumes grow, Germany will have to drop the subsidy levels, and then the rest of the world will begin to reap the benefit of the economies of scale the Germans have paid for.

Solar costs are dropping about 10% or more per year, which puts it at $.12/kwrh in 5 years, and $.06 in 12 (this is a cost-reduction rate which is reasonably well accepted among experts in the area - actually, it may be much faster, with the rate of change in thin-film PV). Solar installation volumes won’t catch up with wind anytime soon. Instead, in around 7-10 years solar is likely to catch up with where wind is now, which is to say that it will be a clearly up and coming large scale power source.

Please note that prices have not fallen quickly, even as costs have fallen. Why? Demand has really gotten ahead of supply. PV supplies are expanding at about 40% per year, but they are only just now beginnging to catch up with demand, especially in Germany. CA has increased subsidies, and France has raised the price they'll pay for PV power, but Germany is gradually reducing theirs. Lately supply seems to be catching up with demand, and prices started to fall in mid 2006.

PV suppliers can still charge a heckuva markup to ration their product, until supplies catch up in the next year or so. For instance, First Solar has a 55% gross profit margin.

Will solar soon be as cheap as coal even when you don't include external costs, like CO2 and mercury?

Yes, when compared to new coal plants - the cost of new coal plants is surprisingly high. Sadly, old, dirty, coal plants are probably unbeatable.

Could solar supply all of our power?

Probably, but it's marginal costs will rise above those of other sources at a much lower % of market penetration, due to intermittency (daily, seasonal and weather-related). The optimal level is around 35% of the overall market as a rough guess.

What solar's advantages?

1st, solar PV is mostly a retail, consumer side technology, and competes with retail pricing. In the US that means that it's competitive at $.10 per KWH, not $.04-.05.

2nd, it provides peak power, which is more expensive for Industrial/Commercial (I/C) consumers, and hopefully will become so for residential consumers.

3rd, most power demand is daytime, especially when you include the I/C demand which DSM has shifted to the night, and which people generally, and erroneously, include as part of "baseload".

4th, PV costs are plummeting - see the last section.

5th, consumers can buy and install PV with very little cooperation from utilities. If they don't care about selling back to the utility then they can cover roughly 75% of their consumption and rarely have unneeded production.

Our energy economy is awfully large, and we currently invest hundreds of billions of dollars in it. That investment just needs to be redirected. Currently PV is labor intensive, and therefore somewhat more expensive than fossil fuels in most places, but that's changing fast - check out

Can solar supply all we need (is it scalable)?

Yes. The earth receives 100,000 terawatts continously from the sun, and humans use the equivalent of 4.5 terawatts on average (15 TW of BTU’s is the standard measurement. That’s equivalent to 1/3x as many electrical BTU's. For instance, in the US 39 quadrillion BTU's are used to produce 13 "quads" of electricity).

Can solar grow fast enough to matter?

Yes, it's doubling more quickly than every 2 years, and manufacturing capital costs are falling, so that growth rate appears sustainable.

Solar PV grew at a 25% annual rate from 1994 to 2000 (doubling twice), and a 40% annual rate from 2000 to 2006 (doubling three times), and the rate of growth is still accelerating (it’s constrained only by the speed manufacturers can ramp up). In 2007 about 3.5 gigawatts worldwide was installed. Solar is definitely here.

Can solar help with Peak Oil?

Yes. First, liquid fuel can be replaced with utility powered plugins and EVs.

It's interesting to note that at current prices PV is cost-effective on any form of transportation that's in use all day - RV's, trucking, bus, rail, water shipping, even aviation. They're all going to hybrid-electric drive trains (or went long ago, in the case of rail), and PV can provide a surprisingly high % of their power (100%, in the case of container boats).

Water shipping is the easiest form of transportation to power renewably. In fact, container vessels could easily run mostly on solar and wind, due to the very low power to surface ratios of these huge boats.

Is it cost effective?

Sure - it's just straightforward calculations: PV can generate power for the equivalent of diesel at $3/gallon (40KWH per gallon @40% efficiency = 16 KWH/gallon; $3/16KWH = bout $.20/KWH.

Could the big container ships that cross the oceans get a substantial fraction of their power this way?

Let's take the Emma Mærsk. With length: 397 metres, and beam: 56 metres, it has a surface area of 22,400 sq m. At 20% efficiency we get about 4.5MW on the ship's deck at peak power. Now, as best I can tell it probably uses about 10MW at 12 knots (very roughly a minimum speed), 20MW at 15 knots, and 65MW (80% of engine rated power) at 25.5 knots (roughly a maximum). So, at minimum speed it could get about 45% of it's power for something close to 20% of the time, for a net of 9%. Now, if we want to increase that we'll need either higher efficiency PV, or more surface area from outriggers or something towed, either of which will increase costs. I suspect that the outriggers would be very cost-effective, but that would involve some design analysis by naval architects.

On using wind propulsion to cut long-distance shipping costs by 10- 50%:
It's astonishing what can be done with modern materials, computer-aided design, and electronic control systems, to turn the old new again..

Large batteries could be carried for the remainder, to be recharged at frequent port stops, as used to be done with coal. Or, the ships could just slow down - a speed reduction of 25% reduces power consumption by 50%. If this is so easy, why don't we do it already? Because bunker fuel has been so cheap. Now, even at PV's currently relatively high price points it would be cheaper than bunker fuel for propelling ships.

As PV gets cheaper, and oil more expensive, more efficient forms of PV become economic - 10% efficient PV is the cheapest right now, but 40% efficient will get there, and that means a high % of industrial transportation energy from PV. The only exception here is aviation, which is probably limited to getting something around 25% of it's energy consumption from high-efficiency PV.

Is solar being slowed down by the current credit crunch?

Only slightly. One of the largest suppliers has cut it's 2009 forecasted growh from 75% to 58%. One of the largest suppliers has cut it's 2009 forecasted growh from 75% to 58%.