Reasons to be cheerful, Part 2

by John Quiggin on June 16, 2011

There are plenty of reasons to be gloomy about the prospects of stabilising the global climate, but there are also some promising developments, so I’ve started a series on this topic.

I’ve been meaning to write this post for a while, but Stephen Lacey at Grist (via David Spratt on Twitter) has done much of the job for me, and better than I could have. The crucial point is that the cost of solar photovoltaic electricity has fallen dramatically and is almost certain to fall further. In particular reaching the point where it is the cheapest large-scale alternative to carbon-fuelled electricity generation, and competitive (at reasonable carbon prices and in favorable locations) with new coal-fired power.

This makes for some fundamental changes in the debate over climate change and mitigation, even as it reaffirms the central point that advocates of mitigation have made all along, namely that, with an appropriate policy response, the costs of drastic reductions in carbon emissions will be modest in relation to national or global income.

Until very recently, solar PV was just a promise, with a share in total electricity generation so small as to be negligible. As Lacey observes that changed in 2010, with 17GW of peak capacity installed. Allowing for availability, that corresponds to something like 4GW of coal or nuclear (standard plants are typically about 1GW). But that’s overly conservative because solar output is such a good match for peak daily demand.

The growth in solar PV has been driven by subsidy schemes like those in Australia. As in Australia, the decline in costs has produced massively more demand than expected, leading to succesive rounds of cuts. But while individual markets have bounced around, the market as a whole has grown massively, even as subsidies have been scaled back. At least so far, that’s true for Australia as well. As I said in the Fin a while back, this is one of the rare instances of an ‘infant industry’ outgrowing the need for subsidies.

Capacity for annual output of new solar modules is now approaching 50GW (peak), at which point solar PV would be one of the main sources of new generating capacity, comparable to wind and gas.

There is no obvious constraint on further growth. Until about 2005, the solar industry depended on offcuts from the semiconductor industry for silicon (the blip marked ‘silicon shortage’ on the graph represents the point where demand outgrew that source). And much of the ‘balance of system’ (installation, inverters and so on) still represents adaptations of devices developed for other industries, with the associated problems of supplies and inventories. But with recent growth, the whole supply chain will be optimised for solar.

At some point the share of solar PV will be large enough (say 30 per cent) that it will change the balance of supply and demand, ending the present situation where the excess supply of night-time power from coal must be sold at a discount. That will entail both changes in pricing structures, most obviously a premium for power supplied in the early evening or for storage technologies. But starting from a zero base, that’s quite a way off. For the moment, the main issue is cost

If the cost of solar PV continues to decline at rates similar to those we have seen in recent years, the whole debate over climate mitigation will be changed. Plausibly, a CO2 price of $50/tonne will be enough to drive a fairly rapid decarbonisation of the whole electricity sector. That means a smaller increase in prices than would otherwise have been expected, and therefore less of a role for adjustments in final demand.

Coming back to the claim of vindication made above, the sensible case for the claim that mitigation could be achieved at low cost was not to identify some particular technology as the anointed savior, but to argue that, with a carbon price (through a carbon tax, emissions trading scheme, or, less desirably, ad hoc measures that produce an effective price) and supporting policy instruments, some combination of options (renewables including solar, wind and geothermal, nuclear, CCS, energy efficiency, changes in demand patterns) would produce substantial reductions in emissions at relatively low-cost. At this stage, it looks as if solar PV and energy efficiency are the most promising candidates, along with wind, while most of the others look less hopeful than they did a few years ago. While this particular outcome could not have been predicted with any reliability, the general pattern could be predicted and was.

Earlier in this series Reasons to be cheerful (Part 1): Peak gasoline

{ 52 comments }

1

Utisz 06.16.11 at 8:56 am

It’s good that the price is coming down, but isn’t a bigger obstacle to it being implemented straight away the fact that with the current state of the art, PV cells/units would need to cover an unfeasibly vast area of land to make any inroads on our energy needs. It’s a while since I read David McKay’s Sustainable Energy Without the Hot Air, where this argument is made and backed up with stats and maybe the science has moved on in the last year or so. Maybe others have more (and more up-to-date) info? I’d be interested to hear.

2

dan p 06.16.11 at 9:44 am

MacKay’s focus on the UK can be a little distorting. Among large population countries it appears to have near the worst combination of high population density and low insolation:
http://en.wikipedia.org/wiki/List_of_sovereign_states_and_dependent_territories_by_population_density
http://en.wikipedia.org/wiki/File:Solar_land_area.png

In much of the rest of the world, space is not really the issue.

In his chapter on importing power I think he does a decent job of showing that it would be a giant project but not unfeasible. Since current power needs of places like the UK, Japan, or especially Hong Kong are met by importing huge amounts of fossil fuels, it’s not really a new problem.

3

John Quiggin 06.16.11 at 9:46 am

“PV cells/units would need to cover an unfeasibly vast area of land to make any inroads on our energy needs”

This is a spurious argument, I think. According to this source (PDF)

http://www.nrel.gov/docs/fy04osti/35097.pdf

“We would need only 10 million acres of land—or only 0.4% of the area of the United States—to supply all of our nation’s electricity using PV. “

4

Tim Worstall 06.16.11 at 9:54 am

Well, yes, decline in costs of solar, true, but umm, wasn’t that what everyone shouted at Bjorn Lomborg for pointing out? A decade ago, that the 20% pa reduction in the cost of solar would make, at some indeterminate point in the future but one not too far away, solar PV the generation technology of choice and thus worrying about the electricity industry part of climate change wasn’t all that a productive bit of worrying?

“The growth in solar PV has been driven by subsidy schemes like those in Australia. “

Not hugely convinced that subsidies had all that much to do with it. The decline in costs was partly new forms of solar PV (ie, First Solar’s Cd/Te technology) and partly good old getting better at the old form. As conceptually simple as slicing the silicon more thinly so as to get more wafers from that expensively produced Si ingot (conceptually easy but technically v. difficult).

I certainly don’t think subsidies delayed these things of course: but not hugely convinced that they sped anything up very much either. There’s not been all that much subsidy to the chip making industry these recent decades but price performance numbers improvements have been similar (or possibly better) for that related technology. And manufacturing technology improvements do tend to chug along at their own speed. It doesn’t matter all that much what the subsidy for installation is if it sdtill takes 18 months to go through an iteration of building a new line, producing a new model and then working out how to do it one step better on the next line.

And of course even if the subsidies did bring that magic equal to coal produced at the point of consumption date forward a few years (which at best is what they did) that isn’t quite the point we want to be analysing. Which is, OK, but was it worth it?

Was spending $x worth bringing forward solar PV viability by 5 (or 10 or 15?) years? Would the same subsidy spent elsewhere have done better? Or in JQ style language, what was the opportunity cost of spraying money at solar PV when solar PV may or may not have needed it?

5

Tim Worstall 06.16.11 at 9:59 am

Oh, and on the land area issue. Current cells are about 10% (ish, ish) efficient. We get 10% of the light turned into ‘leccie.

The next technology, multi-junction cells (make layers of different materials, each absorbing/turning into ‘leccie a different part of the spectrum, thus increasing the amount of the light that is useful) can be up to 40% efficient.

Yes, sure, another decade or two before they’re available on street corners but almost inevitable that they will be. And they would rather reduce the land area problem.

6

ajay 06.16.11 at 10:15 am

“We would need only 10 million acres of land—or only 0.4% of the area of the United States—to supply all of our nation’s electricity using PV. ”

Also, the point should be made that PV doesn’t necessarily “use” land in the same way that farms or roads or buildings do. If you put crops on a bit of land, that’s all you can do with it – you can’t put a house on the same bit of land without it ceasing to be a farm. (Unless you are growing stuff on the roof, I suppose.) But you can put PV on top of buildings – so it’s not as though you’d have to get 10 million acres of farmland and turn them into PV farms.
In any case, even this wouldn’t be a financial barrier. At $100 an acre for marginal desert/scrubland, the cost of the land would be dwarfed by the cost of the PV panels you’d put on it.

7

Pete 06.16.11 at 11:22 am

The UK currently has lots of industrial and retail buildings which are effectively ugly metal sheds with flat rooves. Provided they’re not too heavy we could easily cover them in solar panels to power said commerce, whose greatest demand is going to be during the day.

(Mind you, someone might eventually notice that most of the electricity generated on the roof is going into the lights below it, and decide to cut out the middleman)

8

Chuchundra 06.16.11 at 11:56 am

Plus, how many hundreds or thousands of square miles of parking lots could be easily covered with a roof with PV panels placed on top? We’ve got plenty of unused or underused space here in the USA where you could still solar panels if the economics of it made sense.

9

ajay 06.16.11 at 12:03 pm

7: about 15kg per square metre is a good figure for the weight – so for a big-shed-type place 100m on a side you’re talking about 150 tonnes of extra weight on the roof.

Re cutting out the middleman by cutting holes in the roof – the value of having the rain kept off is not to be underestimated :)
The installation above would generate 40 MWh per day in the UK. I have no idea how much this would leave for export after you’d powered your own lights, heating, ventilation, cash registers and so on.

10

TheF79 06.16.11 at 1:26 pm

According to the link, the levelized cost of PV is still 0.22 cents/MWH (even after the drop in cost), which is substantially more expensive than coal (~5 x without carbon pricing), and about double the cost of wind. It would be great if massive amounts of solar were right around the corner, but until that cost halves itself and maybe halves itself again (not to mention resolving the dispatch problem), I’ll be skeptical. Wind strikes me as more promising, is currently about half as expensive, accounted for around 2% of US generation (compared to 0.02% for solar), and has doubled in capactiy every 2-3 years for the last decade. That said, based on EIA’s 2016 estimates of levelized costs, unfortunately neither wind nor solar look terrible good in a no-subsidy/PTC, no-carbon-price world. All the more reason for carbon pricing I reckon.

11

ajay 06.16.11 at 1:33 pm

Back of the envelope time – looks like power consumption of roughly 1 to 1.6 MWh per year per square metre of sales area for supermarkets, in very general terms. And less than that for the storage rooms (those open chiller cabinets must use a lot of power).
PV generates 1.46 MWh per year per square metre in Britain (4 kWh per day). So a supermarket would easily be able to power itself and have a bit left over to sell.

12

Oliver 06.16.11 at 1:54 pm

Still, they don’t work in darkness and work badly in bad weather. At some point you need dependable energy.

13

jon 06.16.11 at 3:15 pm

The future is bright! This article is relevant.
http://ieet.org/index.php/IEET/more/naam20110317
“The cost of solar, in the average location in the U.S., will cross the current average retail electricity price of 12 cents per kilowatt hour in around 2020, or 9 years from now. In fact, given that retail electricity prices are currently rising by a few percent per year, prices will probably cross earlier, around 2018 for the country as a whole, and as early as 2015 for the sunniest parts of America.

10 years later, in 2030, solar electricity is likely to cost half what coal electricity does today. Solar capacity is being built out at an exponential pace already. When the prices become so much more favorable than those of alternate energy sources, that pace will only accelerate.”

But the right conclusion is not to set the climate change mitigation target lower. We should aim at a more demanding target at a faster pace.

14

Barry 06.16.11 at 3:41 pm

ajay 06.16.11 at 12:03 pm
” Re cutting out the middleman by cutting holes in the roof – the value of having the rain kept off is not to be underestimated :)”

If only there were some materials which passed light but not water!!!!!!!!!!!!

15

Kenny Easwaran 06.16.11 at 3:42 pm

TheF79 – a carbon price would certainly help the transition happen earlier, but if you look at the graph above, the cost of solar energy has dropped by 75% in the past 15 years. Unlike coal or even wind, solar seems to develop at a pace comparable to electronics, probably because it is a piece of electronics while the others are just big turbines.

Oliver – that’s completely right that a source of dependable energy is needed, but as long as there’s some wind power, and smarter energy transportation and storage, the problems for solar might not be as bad as they are now. The daily cycle of solar energy nicely matches the daily cycle of electricity demand (though probably with some excess morning capacity and excess evening demand), but the annual cycle looks a bit more problematic. That annual cycle effect will be hemisphere-wide, and therefore can’t be solved in the way that cloud cover can be, by piggybacking on excess capacity a few hundred miles away.

16

mpowell 06.16.11 at 4:08 pm

@14: This only becomes an issue once solar power is a significant source of energy, in the 30% range. By that time it may be cheap enough to just install excess generation capability which should be able to handle seasonal variations.

A lot of people don’t understand that power consumption habits could be shifted pretty easily if solar becomes a dominant source of energy. For example, if you start running electric cars, you could have people charging them during the day.

17

ajay 06.16.11 at 5:42 pm

If only there were some materials which passed light but not water

Listing the possible disadvantages of working in a giant greenhouse in summer is left as an exercise for the reader…

By that time it may be cheap enough to just install excess generation capability which should be able to handle seasonal variations.

Conveniently, by the way, winter tends to be much windier than summer, which might serve as at least a partial hedge…

18

dsquared 06.16.11 at 6:04 pm

Listing the possible disadvantages of working in a giant greenhouse in summer is left as an exercise for the reader

There’s a sort of “effect”, of which greenhouses are an example, isn’t there? I know they’ve got something to do with carbon pricing somehow.

19

leederick 06.16.11 at 7:40 pm

“the sensible case for the claim that mitigation could be achieved at low cost was not to identify some particular technology as the anointed savior, but to argue that, with a carbon price … and supporting policy instruments, some combination of options … would produce substantial reductions in emissions at relatively low-cost.”

I really don’t agree with this. We’ve very narrowly escaped an absolute disaster. Carbon costing isn’t neutral between technologies, it’s actively pro-nuclear. That is what happens when carbon is the only externality priced when other externalities (like spent nuclear fuel and other nuclear hazards) are actively subsidised, we’ve come extraordinarily close to having an extremely high upfront cost committed for nuclear reactors. These would then have been used over a 50-year lifespan rather than a better technology because of sunk capital costs. The policy was idiotic and we’re extremely fortunate external events mean it’s logic won’t be followed through.

I also don’t understand the case against backing particular technologies, solar’s success was totally predictable and entirely predicted.

20

Lee A. Arnold 06.16.11 at 8:13 pm

The case against solar subsidies is a form of mysticism based on the idea that the market will produce solutions in enough time to prevent disaster. Lomborg was shouted at for being a dunce. There is no a priori logic proving that case. Quite the opposite, there is a full litany of analogous system catastrophes which suggests a great danger to be avoided.

Solar technology research is suggesting a lot of new possibilities which should be fully encouraged. See for example this list of recent real news, from MIT’s Technology Review, and hit the pages at the bottom to go further back:
http://www.technologyreview.com/tag.aspx?id=115&aid=37715&mod=Tag_solar

21

Omega Centauri 06.16.11 at 8:45 pm

I try to follow the area, and it seems to me that rapid progress (as measured by manufacturing price) have only come about in the last two or three years. I strongly suspect this is a critical mass (of industry effect), there just wasn’t enough prospect for making high profit from investment in improving the state of the art, so those improvements ran at a desultory pace. I also suspect serious price competition from China also has a lot to do with the recent turnaround in the rate of progress. So those susbsidies (mostly paid for by German utility customers), may have been crucial in getting the scale of the market big enough to get this industry over the hump.

There’s still tons more work to be done. Especially in the less high tech areas that get lumped under the term balance of system. But significant further cost reductions look highly probable.

BTW, most current PV being installed has an efficiency in the upper teens. The exotic triple junction stuff, that can deliver near 40%, is really only good with complex (light) concentrating systems. There are finally some decent sized CPV projects getting approved, but I think it questionable whether fancy optics plus trackers can compete with cheap flat panels. I wouldn’t be at all surprised if the efficiency for panels with reasonable cost performance stalls out around 20%.

22

Lemuel Pitkin 06.16.11 at 8:57 pm

Not my area but I had the impression that, despite the attention photovoltaics get, in fact (1) wind power is way ahead of solar, being already cost-competitive and supplying a nontrivial fraction of commercial energy (and about half of net new capacity in the US in recent years); and (2) even within solar, high-temperature solar thermal is a more plausible candidate for commercial-scale power generation than photovoltaics.

Can anyone who actually knows something about this say if either of those impressions is correct?

23

Omega Centauri 06.16.11 at 9:15 pm

Lemuel: Certainly in terms of installed capacity wind is ahead. Solar’s compound growth rate is higher than wind, so if you extrapolate out several years it will catch up. Siting requirements for solar are pretty easy to meet. Also the power per square meter of land from which the public must be excluded is much higher in solar. Wind turbines require a large area with no public access, because of the danger of flying turbine blades (given that particular failure mode).

Several proposed solar thermal plants have been outbid by photovoltaics, which is cheaper on a per watt (or per Kilowatt-hour) basis. The real advantage for solar thermal is the possibility of storing the heat for later usage, making the power at least weakly dispatchable. However current contracting quidelines don’t give any credit for (partial) dispatchability. I suspect solar thermal may become a niche technology, either in hybrid form, reducing daytime demand for fuel of thermal plants, or for cogeneration, where low grade industrial heat is needed, and the waste heat from the plant may be more valuable than the electricty.

ajay: Your annual PV numbers are way to high. I suspect they assume 100% conversion. My roughly 22 square meters of panels provide about 4MWhours per year (in sunny Califonia). Current grocery stores are just plain huge energy hogs. Lots of open to the air freezers require ridiculous amounts of power to operate. But whats a few extra zillion KWhours, compared to the chance a consumer may not make an impulse buy if he has to open a freezer door to select an item?

24

Matt McIrvin 06.16.11 at 11:00 pm

we’ve come extraordinarily close to having an extremely high upfront cost committed for nuclear reactors.

In what country? In the US, for reasons John has already explained, the nuclear renaissance was dead on arrival even before Fukushima Daiichi. Much of it is because natural gas is still really cheap here.

25

jafd 06.17.11 at 12:19 am

Re Ajay’s comment – mayhaps recent story of the lightpole-mounted solar panels of New Jersey is relevant.
http://www.nytimes.com/2011/04/28/science/earth/28solar.html

26

John Quiggin 06.17.11 at 1:40 am

Looking at the picture for the NJ story, it’s hard to imagine what kind of person would live happily with the poles and wires that dominate the street, then jib at the panels.

27

Yarrow 06.17.11 at 1:52 am

Looking at the picture for the NJ story, it’s hard to imagine what kind of person would live happily with the poles and wires that dominate the street, then jib at the panels.

Yes but people don’t see the poles and wires — they’ve been there forever. The panels are new, so they’re visible. In few decades there will be “save our solar panels” groups pleading to keep them even though they’ve been obsolete for years.

28

David 06.17.11 at 4:23 am

Given the incredible levels of direct and indirect subsidization that coal, oil and nuclear have received over the last century and the ongoing emergent disaster at Fukashima, I think that the case for alternatives is looking quite strong. Of course, Entrenched Interests and Conventional Wisdom are formidable obstacles to overcome.

29

Matt 06.17.11 at 8:10 am

According to the National Renewable Energy Laboratory, with 13.5% efficient modules in the USA you could produce about 819 terawatt hours annually simply from rooftop systems. That doesn’t require any new land to be used. By way of comparison, the USA produced about 800 terawatt hours from nuclear plants last year.

Plain old silicon cells (no rare elements, concentrating systems, or multi-junction fabrication required) can hit nearly 25% conversion. Over 20% is already commercially available. If those rooftop systems work at 20% you get over 1200 terawatt hours per year, again with no new land consumption.

What to do with those 1200 TWh? If you used it to power electric vehicles with the efficiency of the Nissan Leaf (200 watt-hours per kilometer on average), you could drive about 6.1 trillion vehicle-kilometers. In 2010 Americans drove about 4.8 trillion vehicle-kilometers. To a zeroth-order approximation, rooftop solar alone can provide enough energy to eliminate American gasoline consumption and then some.

If you want to go beyond the zeroth order I will gladly concede that many vehicles cannot be replaced by a Nissan Leaf or comparable EV, and that simply totaling up the annual watt-hours available ignores many important details, but in terms of a handful of numbers I think it’s highly suggestive and encouraging.

What would give me a real feeling of relief is an affordable flow battery system to store the electricity that we can’t stuff into car batteries or usefully shuffle around the local microgrid at time of generation. There was some puffy news release a few weeks ago about researchers with a new approach based on iron instead of vanadium that was supposed to take storage costs down to $30/kWh for a sufficiently large capacity (several megawatt hours) flow battery. It sounds too good to be true and unfortunately none of the news pieces said anything about the actual chemistry other than “it’s safe, abundant iron instead of expensive, foreign vanadium.” If you could really get storage costs down that low, hell, who needs large scale grids at all anymore?

30

ScentOfViolets 06.17.11 at 1:21 pm

Against stupidity the Gods themselves contend in vain?

31

French Uncle 06.17.11 at 2:14 pm

“because solar output is such a good match for peak daily demand.”

I don’t know for Australia, but in Europe it is quite true in summer, but fdefinitely wrong in Winter, which on top of that has an quite higher peak demand than summer.

The same intra-year balancing should be taken in account when speaking of the electric autonomy of the retail shop.

32

ajay 06.17.11 at 3:58 pm

31: Here’s the British DECC talking about daily variations in demand.
http://www.decc.gov.uk/assets/decc/statistics/publications/trends/articles_issue/560-trendssep10-electricity-demand-article.pdf

Chart’s on page 2. It looks like a rough match with insolation in summer. In winter, though, demand peaks at 5.30 pm, by which time it’s dark in Britain, according to the National Grid. The main culprit seems to be the British fondness for making tea (with electric kettles) immediately after any great emotional or spiritual experience.

http://www.nationalgrid.com/NR/rdonlyres/1C4B1304-ED58-4631-8A84-3859FB8B4B38/17136/demand.pdf

33

ScentOfViolets 06.17.11 at 4:08 pm

Ajay, you’re not suggesting that Brits change their habits to accommodate solar energy, are you?

34

ajay 06.17.11 at 4:14 pm

That’s not really going to happen unless you can persuade them not to want to turn the heating on when they get home from work in winter. Tricky habit to change, that.

35

Pete 06.17.11 at 4:18 pm

Looking at that chart in 32, it appears there’s still plenty of headroom for adding renewables to displace the easily-variable gas consumption. The issue of storage is one that doesn’t have to be solvedyet.

(Anyone done the maths on what kind of peak/trough electricity spot price spread would be necessary to make battery facilities feasible?)

36

Pete 06.17.11 at 4:18 pm

Also, you’re not going to get people to change their habits, but you might be able to get somewhere with smart loadshedding on industrial uses.

37

Theophylact 06.17.11 at 5:33 pm

And a promising technique offers a significant reduction in the cost of silicon wafers for PV panels.

38

Paul 06.17.11 at 6:04 pm

Solar thermal offers the interesting possibility of making hydrogen directly from water. A process based on ferrite redox cycling could achieve 20% efficiency to hydrogen (and also would produce electricity from a steam bottoming cycle, not counted there.)

39

Tim Worstall 06.18.11 at 10:31 am

“What would give me a real feeling of relief is an affordable flow battery system to store the electricity that we can’t stuff into car batteries or usefully shuffle around the local microgrid at time of generation.”

There are other ways than batteries to do that. Solar or wind/electrolysis/hydrogen/fuel cells for example. Whether it’s economic or not depends upon the price difference between that cycle and batteries. An unknown as yet. Although I hope this technology succeeds:

http://www.bloomenergy.com/

For it uses my beloved scandium…..

40

Barry 06.18.11 at 1:26 pm

Barry: If only there were some materials which passed light but not water

ajay ” Listing the possible disadvantages of working in a giant greenhouse in summer is left as an exercise for the reader…”

As you know, Bob, such materials have developed quite a bit from the old pre-space days. There are these things called ‘coatings’ which can selectively alter the distribution of energy pass-through.

41

Barry 06.18.11 at 1:38 pm

Something occurred to me while reading this thread – the reason that solar systems are improving at a phenomenal rate is only partially due to Moore’s Law. The other reason is that solar power is far closer to the free market than, say, nuclear. There are a very large number of individuals and companies working on improvements in the base systems, installation, financing, power manipulation (at the micro level) and analysis/decisions on when/what to buy.

Nuclear, on the other hand, is Big Bureaucracy. It’s too big for the free market, so it’s got to have government support from A to Z. Not just storage, but the R&D, the training of operators and engineers (‘in the Naaaaaaaavy, you can sail the seven seas!’), insurance and financing (making sure that the up-front billions are recoverable, one way or the other.

42

Matt 06.18.11 at 7:04 pm

“There are other ways than batteries to do that. Solar or wind/electrolysis/hydrogen/fuel cells for example. Whether it’s economic or not depends upon the price difference between that cycle and batteries. An unknown as yet. Although I hope this technology succeeds:

http://www.bloomenergy.com/

For it uses my beloved scandium…..”

Sure, I’ll be happy with anything that can provide affordable storage. I’m not completely wedded to flow batteries, I just think they have the best shot at affordable large-scale storage for regions without the option to do pumped hydro. Anything that involves hydrogen I’m wary of because it’s a bitch to store (either large irreversible thermodynamic losses, large equipment costs per unit of energy stored, or both).

I hope Bloom succeeds wildly too, mostly so we can use natural gas more efficiently. If you can say anything technical about the future of scandium supply, I’d love to read it. It’s a surprisingly abundant element, only seeming rare because its chemistry stubbornly resists most geochemical concentration processes. Based on my reading of old Bureau of Mines publications, I’m guessing that expanding scandium production is most likely to come from Bayer process red mud or certain titanium, tungsten, or uranium ores. Apart from the expense, the solvent extraction and ion exchange methods I see in those old publications seem to implicitly carry a high environmental cost from high volume but dilute waste streams. Maybe that’s the big obstacle to overcome now, or maybe it’s overcome by siting production where regulation is minimal.

There seem to be a lot of Chinese and a few Russian suppliers on Alibaba offering scandium oxide. Most say that their available output per year is quite modest, and I suspect that those offering huge quantities are being less than honest. What I don’t know is if there are really a lot of producers recovering a little scandium from other operations or if most of the suppliers are just resellers. Or, hell, most of them could be scammers for all I know; even a one-kilogram “sample order” could yield the equivalent of several months’ wages in China.

43

Paul 06.18.11 at 9:45 pm

Based on my reading of old Bureau of Mines publications,

http://minerals.usgs.gov/minerals/pubs/commodity/scandium/mcs-2011-scand.pdf

Its presence in ferromagnesian minerals is interesting, since those might be processed in extremely large amounts if mineral carbonation ends up being used to sequester CO2.

44

Matt 06.18.11 at 10:29 pm

http://minerals.usgs.gov/minerals/pubs/commodity/scandium/mcs-2011-scand.pdf

Its presence in ferromagnesian minerals is interesting, since those might be processed in extremely large amounts if mineral carbonation ends up being used to sequester CO2.

Indeed. If you look at USGS analyses of Columbia Plateau basalt (USA domestic reserves: approx. 4.7 * 10^17 kilograms), one cubic meter of basalt contains about $500 worth of highly valuable elements at current market rates. More than 99% of the total mineral value is concentrated in less than 1% of the mass. Scandium alone accounts for more than half of the value, followed (in order of value) by gallium, molybdenum, vanadium, cerium, neodymium…

The devil is economically separating the tiny valuable fraction from the interfering and comparatively worthless alkali and alkaline earth metals, aluminum, and iron after you’ve crushed and leached the silicates. If you can simultaneously sequester CO2 it could make a better economic case than mineral extraction or CO2 sequestration alone. The idea is to combine the notions of Schrag’s “Electrochemical Acceleration of Chemical Weathering…” with extraction of valuable metals. If you’re dissolving a million tons of basalt anyway, might as well try to pick up some added value other than the CO2 sink.

In some ways this dream starts to resemble the absurdity of business plans to mine the moon or asteroids. The total mineral value locked up in common rocks is enormous, but the costs to extract that value are even more enormous. You need some “magic wand” like extremely cheap electricity (keeping fingers crossed for long-term renewable energy cost reductions) and/or a substantial global CO2 emission/sequestration tax/credit before it starts to sound plausible.

45

Tim Worstall 06.19.11 at 7:44 am

“If you can say anything technical about the future of scandium supply, I’d love to read it. “

Probably best not to set me off as this is exactly what my day job is all about. I run/own a tiny company that handles some 50% of the world Sc trade each year.

“Based on my reading of old Bureau of Mines publications, I’m guessing that expanding scandium production is most likely to come from Bayer process red mud or certain titanium, tungsten, or uranium ores.”

Uranium is where most of it used to come from. Certain of the Kazakh fields were mined using in situ leaching. There was an Sc capture circuit added to the processing plant (at Aktau) and the stored production from Soviet days has been what everyone has been using these past 15 years or so. The plant’s been closed since 1993 ish: and no one’s willing to develop and roll out technologies based on one CIS supplier anyway.

The last 18 months I’ve been running around the various possible alternative supplies. TiO2 from ilmenite doesn’t work any more as an Sc source. It used to concentrate in the bottom clinker of the chlorinator and there was a process developed to extract it from that. Modifications in the process mean that there’s no bottom clinker any more, let alone any Sc concentration. While we can measure the Sc going in, it’s dispersed through the entire production output, no nice place to collect it.

Ta processing, also tin slags: not enough to make it worthwhile there. Hundreds of kg pa only from the largest processors possible. Tungsten processing? Almost all done in China these days and again, a reluctance to base a process on China/CIS supply lines. Not a great source anyway. Zircon sands fail for the same reason as the TiO2 route. Nb/Y ores, mebbe, still checking but not hopeful.

Red mud is our answer. We’ve proved that it can be extracted in some work in Greece. A rough guide is 5 tonnes from each 100,000 tonnes of red mud. As there’s 40 million tonnes a year produced (dry weight, ish) should be enough for everyone.

There’s a CSIRO paper which explains much of this if you’ve university access:

http://www.sciencedirect.com/science/article/pii/S0304386X11000648

They also mention two Oz junior miners who are proposing to go hard rock mining for it. Extract from nickel laterites. We/I think they’re mad to do this when there’s that huge amount in red mud.

Who is actually right we’ll find out in time. Essentially depends on who gets the development money first. Anyone with a spare $20 million is encouraged to contact me. The problem with the red mud route is that the reagent consumption (mainly acids) makes it expensive to process directly for the Sc content. Entirely, proven to be, possible but expensive. To bring the cost down we want to extract the Fe, Al2O3, TiO2 and NaOH first, then play with acids on the residue. There’s a technique developed to do this (weirdly, from the U of Sheffield). Which is what we want the $ to do, go and build a pilot plant to prove this.

Grant applications have been filed, investors spoken to but no one has bitten as yet.

Current supplies are from here and there. Various people are extracting a few hundred kg pa each from various sources. Almost a cottage industry really, one waiting to be properly industrialised.

As an aside, Sc is the only one of the rare earths where China is an importer, not exporter. And it’s not just the people making the stuff for Bloom either.

Best not to take the USGS reports entirely at face value. Zheltye Voda has been flooded for a decade for example and never produced more than a few kg of Sc anyway.

And moving entirely off subject, here’s Oliver Sacks getting a scandium hamburger.

http://www.periodictable.com/Stories/SacksVisit/index.html

That’s at the Urbana IL company mentioned in the USGS report and made from Sc from that Kazakh uranium plant.

46

Utisz 06.19.11 at 3:37 pm

Just to say I’m (@1) happy to stand corrected on this issue. Promising news.

47

Chris Williams 06.19.11 at 8:20 pm

U of Sheffield, funded by Firth to make metallurgy scientific, sitting next to some of the world’s mist advanced special steel firms, with a wacking great legacy of mining engineering, develop an ore-extraction process? Weird.

48

Matt 06.19.11 at 9:05 pm

@45 Tim: Thanks for writing all that. Do you think that solid oxide fuel cells offer the greatest growth potential for scandium consumption, or do you think that (e.g.) its aluminum alloys could be more widely used with increased supply? Is there much growth potential at all, presuming lower-cost production? I don’t know if it’s more like thallium, where you could double production and the world would collectively shrug its shoulders, or more like indium, where lots of people would like to use more if it weren’t so expensive.

49

ScentOfViolets 06.20.11 at 2:02 am

There’s a technique developed to do this (weirdly, from the U of Sheffield). Which is what we want the $ to do, go and build a pilot plant to prove this.

Grant applications have been filed, investors spoken to but no one has bitten as yet.

But but but you said there was a new process that was way cheap (never mind that you never actually produced any sort of quotes or checkable cites) and that you were in the process of extracting money from the financiers. We’re well into our second Friedman unit since the time you made that claim. I won’t press you for the exact number of Friedman units you will require to pass with no followup before you admit that it was yet another lead that just didn’t pan out; I will note, however, that by your own admission the financial types seem to be distinctly unimpressed so far.

No, I’m not picking on Tim; I’m just pointing out how even in the relatively short time since all the last big breakthroughs were announced (most of which have faded without comment since they were first introduced with great fanfare) it’s possible to get a sense of how slow progress in the field actually is and how much of the industry is just so much buzz and hot air. These sorts of things tend to appear on the market – if they appear at all – only after many long development cycles have passed and much money has been spent doing “useless” – i.e., profitless – research. Contrary to popular mythology, something like a better battery is not usually the work of a lone visionary working out of her garage ;-)

And which anybody should have been able to figure out for themselves if they knew a little basic science, some physics and chemistry and maybe a little bit about how industrial process engineering works. People just don’t seem to realize (or possibly just don’t want to believe) what an advanced, efficient and all-round effective technology is embodied by that oh-so-hoary internal (and external) combustion of chemical fuel.

It’s got solar/battery tech beat every which way because – as I’ve said many times before – of some very good fundamental physical reasons. In fact, I don’t think we’ll see that better battery before what is considered some rather advanced nanotech becomes just another routinely available industrial process :-(

50

Tim Worstall 06.20.11 at 6:56 am

“U of Sheffield, funded by Firth to make metallurgy scientific, sitting next to some of the world’s mist advanced special steel firms, with a wacking great legacy of mining engineering, develop an ore-extraction process? Weird.”

Well, yes, a bit, given that there’s no Bayer Process plants in the UK……we’re hoping to test it on the one in Turkey…..

“Do you think that solid oxide fuel cells offer the greatest growth potential for scandium consumption, “

Yes, by orders of magnitude.

Much of the benefit of Sc in Al is use specific. Those who really want it will pay what it currently costs (or perhaps more exactly, a reasonable long term cost, not the current bubble price). It is, I suppose, remotely feasible that if Sc were 10% of that long term cost (ie, $100 a kg instead of $1,000) that it would replace Zr as a grain refiner but I’m not enough of a metallurgist to really be able to pass judgement.

“Is there much growth potential at all, presuming lower-cost production?”

Our analysis, for what it’s worth, is that there are several technologies (sofcs, certain Al alloys) where they will happily use Sc if the price is right, right being somewhere around that long term price I’ve just mentioned. If the long term price is much higher than this then they’ll not be willing to pay for the performance gains. Y can substitute in sofcs with a performance degradation. But the difference between, say, $5,000 a kg of Sc2O3 and $50 for Y2O3 (just as an example) would mean the Y is used. Similarly with Zr and Sc in Al.

If Sc were $100 a kg then I don’t know of any particular uses that people would flock to it for (leaving aside that very suspect Zr replacement). Then again, that’s not unusual: we often find people coming up with a use for something undreamed of by the originator. Edison thought the phonograph would be a dictaphone at first and the telephone was sometimes marketd as a way to listen to concerts remotely. Nobody had really thought of suburbs and commuting until they’d built the first railways.

“a new process that was way cheap”

Nooo. “Cheaper”.

“These sorts of things tend to appear on the market – if they appear at all – only after many long development cycles have passed and much money has been spent doing “useless” – i.e., profitless – research.”

Eh? You tell us all how long and complex is the development period for a new technology then berate me for not having solved it all in one Freidman unit?

What?

If solid oxide fuel cells are to succeed and if scandium is to be used in them then it’s necessary to find a new source of Sc. Which is the little bit of this long and complex technological process that I’ve been working on.

51

ScentOfViolets 06.20.11 at 10:18 am

“a new process that was way cheap”

Nooo. “Cheaper”.

Funny – as I recall it, you were considerably more optimistic way back when. Do you really want me to quote those old posts of yours? I’d be glad to – especially since I made a prediction that’s apparently been confirmed in spades.

“These sorts of things tend to appear on the market – if they appear at all – only after many long development cycles have passed and much money has been spent doing “useless” – i.e., profitless – research.”

Eh? You tell us all how long and complex is the development period for a new technology then berate me for not having solved it all in one Freidman unit?

Don’t grunt – that’s a privilege you’re even further from earning than you were last year when this subject came up. And as I recall it, I was considerably more realistic about the challenges as well as more cognizant of the history of these sorts of “breakthroughs”. You, otoh . . . well, let’s just say that you were being way too optimistic in your assessments.

So I guess I’m berating you for saying these big new solutions were just around the corner and then not copping to the fact that you’ve been proven wrong. In fact, it’s like those events never happened.[1] They are now “inoperative”.

It’s all about the predictions.

52

piglet 06.21.11 at 7:37 pm

More reasons to be gloomy: global primary energy in 2010 jumped 5.6% to a new record high, after a recession-induced decline in 2009. Global oil consumption also reached a new high. Coal consumption jumped 7.6%. Renewables increased 15.5% but are still negligible in absolute numbers (1.3% of total), and this figure includes unsustainable biofuels. Thew only silver lining is that the US, Europe and OECD as a whole are still below their 2007 peaks.

Source: BP Statistical Review of World Energy June 2011, bp.com/statisticalreview

Comments on this entry are closed.