January 18th, 2013

The endless GOP hostage theatre over US budgets succeeds in one of its main purposes: to distract attention from issues that matter. Like the race against the clock to avoid complete climate breakdown. We already have the trailers.

So here comes a fat Victorian-novel style doorstopper weekend post with as many nutritious links as a Christmas pudding has currants.

Renewable energies are still only a drop in an ocean of climate-busting fossil fuels:

world energy consumption table
Spreadsheet here.

What hope is there for my granddaughters? A change of heart by billionaires? A popular insurgency, a Main Street Spring? A miracle discovery?

The one solid basis for optimism is the prospect of terawatts of solar energy. If you just extrapolate the global trend of the last 12 years – an average growth rate of 43% a year – , the world will reach its first cumulative terawatt of solar PV in 2020. Its expansion would cross the world’s full energy demand soon after 2030. That’s fast enough even for James Hansen.

solar PV projections table
Plenty of experts say it won’t happen. But their reasons are unconvincing.

For some time semiconductor experts have been offering very good reasons why Moore’s Law of transistor density must come to an end. But so far it hasn’t, and looks OK for a good few cycles yet. In comparison, the trend doubters on solar PV are running on empty.

Exhibit A: Steven Davis et al., the credentialed authors of a recent scholarly paper on “Rethinking wedges” – my italics:

Existing solar, wind, biomass, and energy storage systems are not yet mature enough to provide affordable baseload power at terawatt scale.

Baseload? What’s that, Grandpa? Well, sonny, in the days before the Deluge we used to have big coal and nuclear power plants running all the time, that’s baseload, and gas plants running only at times of high demand, that’s peak lopping. Now we have wind and solar running when they can, because they are the cheapest, as primary load; but they are variable, so we need despatchable capacity to make up the difference, peak and off-peak. There isn’t any baseload now.

You can’t believe experts who are still trapped in an obsolete paradigm.

Exhibit B, the IEA, the born-again intergovernmental think tank, and semi-penitent former whore cheerleader of the oil industry. They project that solar PV will flatline around 26GW of Solar PV a year, below its current rate. But their assumptions are plain wrong. They predict that utility PV installation costs will gently fall from $2.6 to to $1.65 a watt in 2035. Terje Osmundsen, author of the blog post I got this from, works for a company that will quote around this today. (Independent confirmation from another solar entrepreneur, Jigar Shah, here.) The IEA shows in-depth ignorance by expecting 130 GW of PV to be «retired» by 2035. Osmundsen points out that 25 years is just the maker’s guarantee for 90% performance; the gear lasts much longer, 30-40 years at least.

The only way I can make sense of the IEA’s absurd predictions (I’m not going to shell out good money on a dud report to check) is that they think past growth has been entirely driven by policy, which is now going into reverse, as in Spain. Neither proposition is globally true. The learning curve (a steady 22% since 1955) is driven by technology; policy just enables or obstructs it. And the number of countries with reasonably supportive policies and targets keeps growing: Brazil, Mexico, Chile (for the policies of these three, see also here), South Africa, Turkey, Saudi Arabia, France, India, Thailand and Indonesia …) These are following in the footsteps of the rapidly growing big deployers of China – its latest target is 10 GW this year -, the USA, Australia and Japan. Germany has stopped growing, but at a very high annual level of 7 GW and a stock of 30 GW. Only Greece and Italy have run out of money and are in retreat. In developing countries, the boom is driven by simple economics, not subsidies.

What’s more interesting is that solar boosters like EPIA and Greenpeace also won’t trust the trend. In a joint report, they predict a steady slowdown in the pace of cost reduction and installation, and their optimistic scenario comes out at half the trend number in 2020. I can’t find any considered argument for this. It’s basically herd caution in the consultaverse, reflecting the risk aversion of its client investors. McKinsey is an honourable exception.

Distinguish between the intrinsic characteristics of the technology, and its deployment. Clearly any given process, like the currently dominant polycrystalline silicon cell, will reach an intrinsic cost asymptote some day. If DARPA and its industrial partners think this is below $1/watt installed, who am I to quibble? The current wholesale module price is 65$c per watt (cyclically ahead of the trend), so 50c/watt is pretty certain in the next few years, and BOS costs will surely follow on the German model.

When the current scheme stalls, there is an extraordinary variety of potential relays being investigated, ranging from 3D light-trapping geometries to 44% efficient multijunction cells (on sale now for satellites and military backpacks, at a price) to the ultimate long shot, artificial photosynthesis mimicking the near 100% efficiency of leaves as solar cells achieved by evolution using weird quantum effects. I have no more idea than you which of these will make it to market success. But it would be very surprising if none of them were able to take over the torch, and we were stuck as it were with 5¼” floppy discs for computer memory. In any case, we don’t need any of them to get to tens of terawatts.

The other aspect is the environment for deployment. Successful game-changing innovations like TVs and mobile phones don’t follow the technological learning curve, but a steeper sigmoid: slow acceptance by pioneers, then explosive growth, then a slowdown to saturation. Since solar PV has now hit grid parity in many places, I predict growth will actually accelerate, for example when cheap consumer-friendly plug-and-play AC-PV modules become widely available from the likes of WalMart, this year or next. This market effect, plus ordinary economies of scale and learning, should mask any slowdown in underlying innovation for much of the decade. So I’m predicting 1 TW solar by 2020, without big policy changes. (An early disconfirmation test: if solar PV doesn’t exceed 30 GW in 2013, I’m wrong.)

Beyond that, the environment will begin to impose constraints, and policy will become critical to a successful transition. It’s most unlikely that 50 TW of solar PV will meet world energy demand in 2031, even if it’s by far the cheapest option. What are the possible constraints?

One: running out of land. A world total of 50 TW of solar panels, the upper bound in my thought experiment, corresponds to 25,000 km², or a square about 100 miles on a side. That’s less than the roofspace of a bundle of megacities. So land take is not a killer. The WWF has just released some pretty maps making the same point.

Two: intermittency. We will need a lot of balancing power for solar in a zero-carbon world. (Actually we will probably have to go carbon negative, and build a few TW for sequestration, but never mind for now.) Some of it should be be wind, which is nicely complementary at night and in winter (Fraunhofer here, slides 15, 23, 58) but also variable. So we must have a lot of more expensive despatchable power. This can be either substitute primary sources, in the forms of hydro, geothermal, CSP, and biomass; or additional storage for time- and mode-shifting, as distributed and utility batteries and heat tanks, synthetic fuels catalysed from the air, and load management with price incentives.

Now all of these except the last are going to be more expensive than solar PV. The optimisation problem is to start from 100% solar PV, which is mismatched to demand, and add the other things so as to match demand at least additional cost. Wild guess for a seed solution: for 10TW continuous demand, the primary energy could be 5 TW solar (installed 33 TW), 3 TW wind (9TW installed), 1 TW geothermal (2 TW installed), 0.6 TW biomass (1.2 TW installed), the 0.4 TW of hydro (0.8 installed or under construction) we have now, and 5 TW (?) installed storage. You can shift between wind and solar, and between storage and the despatchables. (To pacify Brett, we should throw in nuclear power, but in fact it’s not going to be more than a footnote, for reasons explained here and here.) A terawatt is 1000 nuclear reactors, which will never be built.

Even in this very crude form it stands out that the policy problem lies not with wind and solar but with despatchable energy. Load management is a no-brainer, but its scope is necessarily limited. We do not have deliverable technologies for storage, geothermal, and biomass on the required scale. That’s where the funding should go.

Three: transportation, iron, and cement. These are the main areas where solar electricity is not yet a technical solution. Electric cars are slowly on the way to mass deployment, but not trucks, ships, and planes. Absent technological breakthroughs, it will have to be sustainable bio- or synthetic fuels. Cement-making inherently releases carbon dioxide, even with solar calcining. A lot more work needs doing here.

Four: backlash. Cheap solar PV is disruptive because it’s going to be very cheap, and will destroy value embedded in defeated competitors on a heroic scale. The shift will not be controllable by incumbents. Solar PV doesn’t only come in the unthreatening shape of multi-megawatt utility solar farms. It also arrives in distributed form on residential and commercial rooftops – turning millions of passive electricity consumers into an activist army of prickly producers armed with social media and votes. Some rough and noisy politics is in store before the oil and coal industries die.

Germany is the laboratory for the future here, as solar is already below grid parity and renewables are having a major impact on electricity supply. Merkel’s conservative government, pushed by incumbent utilities and heavy industry, is increasingly hostile to its lusty and disruptive solar cuckoo. It favours expensive offshore wind to replace the rashly closed nuclear power stations. In Spain, rooftop solar was strangled in the cradle, so there’s no lobby to defend it, just marginal bloggers. But in Germany the solar producers are a large and strong lobby.

What’s more, the government has lost the technocratic control of the situation it had when the FIT was a subsidy and could be tweaked. The official target for solar PV installation in 2012 was 2.5-3.5 GW; the result was 7.6 GW. The FIT will expire above 53 GW, but why should German households and warehouse-owners stop? The payoff is less and less the miserable FIT, more the saved consumption, valued at the full retail price of 25c€/kwh. Local storage, as low-tech as the venerable storage heaters being dusted off and as high-tech as electric cars, will stretch the self-consumption payoff. What’s sold to the grid will have some market price.

Some German solar people, including Professor Volker Quaschning, are already proposing 200 GW of solar PV in Germany, four times the government target, and a 300 GW yearly worldwide installation rate by 2025.

Can we get there? I’ve said it before: we are witnessing, and in many cases taking part in, a technological revolution. Governments can obstruct it as in Spain or help it along as in Germany, but they are not in the driving seat. Current policies are favourable enough for the revolution to happen.

Will it happen fast enough to save our climate? Even with my optimistic scenario, solar PV doesn’t really vanquish fossil fuels till the 2020s. If carbon emissions have to peak in the next five years, the heavy lifting will have to be done by conservation – meaning carbon pricing – and wind energy. Wind has a lower historic growth rate than solar – 28% over the last 15 years, and slowing as the technology matures; but it has a much higher installed base, around 280 GW nominal to solar’s 100 GW. Since wind’s capacity factor is twice as high, the effective ratio is five to six times. There’s a strong case for differential medium-term support for wind. On conservation, the air pollution crisis in China and coal chaos in India offer more hope than US policy, still in hock to denialism like Noah’s neighbours.
BP parody logo
Credit: Greenpeace BP logo competition post-Deepwater

19 Responses to “Terawatt solar or bust”

  1. Foster Boondoggle says:

    Don’t overplay the capabilities of natural selection. You refer to “the near 100% efficiency of leaves as solar cells achieved by evolution using weird quantum effects”. Actually current silicon is already more efficient than photosynthesis. The wikipedia page cites a theoretical maximum efficiency for photosynthesis of 11% and an actual performance of about 1/4 to 1/2 that.

    On the other hand, a forest full of green plants is much nicer to walk through than a desert full of PV arrays.

  2. Brett Bellmore says:

    “Cheap solar PV is disruptive because it’s going to be very cheap”

    Tautologically so, I suppose.

    Yes, the past growth has been driven by policy, in the sense that only massive subsidies make solar competative any place that’s able to be hooked up to the grid. That may change, indeed I hope it will, but there are all manner of problems with solar PV which can be papered over, (Like using the grid to compensate for intermitancy.) so long as it’s a small fraction of power usage, which will become increasingly important as it starts to push past a few percent.

    • Dan Staley says:

      Yes, the past growth has been driven by policy,

      Fossil still receiving $Bns in subsidies notwithstanding, the past growth – esp. recently – has been due to technology bringing cost down.

    • Richard Cownie says:

      I’ve run through the numbers a few times, and it always seemed to me that solar was
      highly competitive with grid power once you could get it for about $3/peak watt
      installed. The modules and inverters seem to be in that range now – the remaining
      obstacle is installation cost. Some of that may be reduced by better, and more
      standardized, designs of the components. Some of it will go away once there’s a
      mass market. Though sending workers up onto residential roofs is always going to
      be a little dangerous and expensive.

      For residential use in the USA, the intermittency is just fine because the peak
      usage of electricity is on hot sunny days when AC units run full blast – which
      precisely matches the peak output of solar.

      Also note that the economics vary with local climate: I’m in Massachusetts, where
      average insolation is only 4 hrs/day; I’m expecting PV to take off first in the
      sunny Southwest, with 6-7 hrs/day.

      In fact one of the obstacles to rapid adoption right now is that even if the
      economics are marginally favorable right now, you also have to consider whether
      the rapidly falling costs and improving technology mean that it will be an even
      better deal if you wait a couple of years (gritting your teeth and paying utility
      bills in the interim).

  3. Cranky Observer says:

    As an aside, I used to be a bright young man full of ideas in the energy business. Now I’m a cranky old man (not quite a grandpa yet) who occasionally manages teams of bright young men & women. And now I understand that weird quirk of the mouth my elders used to get when I would explain my bright new idea… (I was right about the ARPANET though).

    I don’t disagree with your basic premise, and I certainly know about the ingrained resistance to change in the electricity industry (made worse, not better, by the freshwater-inspired restructuring in 1994+). But I have to ask, have you ever worked in a major load center’s power dispatch control room for a full season (at least six months)? Because it really, really, really isn’t as easy as the wind/solar optimists say – and the people doing it today really aren’t stupid.

    Think very hard before you advocate Yglesiasizing a service as fundamental as electricity.


  4. Jake says:

    What was it that Mark just posted about intellectual honesty and people hoping to persuade rather than inform?

  5. SamChevre says:

    I’d note that the “the full retail price of 25c€/kwh” ($0.33)is almost four times the current retail residential price here in Virginia ($0.097/kwh). Rate filing PDF here

  6. Dave J says:

    I am with Cranky Observer above. I worked in high tech for many years before transitioning to the power industry to save the world. The optimism displayed here is actually hard on the eyes.

    Wimberley starts off with a dig against experts who are stuck in an obsolete paradigm, but in reality most of those experts are not stuck in a paradigm so much as they are stuck operating with the real constraints to manage a real power system safely, reliably, and cost-effectively. It’s not about “baseload” vs “peaking” so much as it is about putting together a portfolio of resources that can provide the level of service to which we’ve all become accustomed.

    As the latter part of the piece points out, PV obviously cannot solve all our power needs and even an infinite quantity won’t keep the lights on at night. Wind is somewhat complementary, in some places, most of the time, but in most places not most of the time, and in all places not all of the time. Storage, of course, can solve this, but the technological problems to provide bulk storage that can shift across minutes, hours, days, seasons, and even perhaps years are not even close to being solved. Price-sensitive demand response can also help, but let’s not kid — making a system that requires DR is a form of lowering the quality of the product — so don’t be surprised about the backlash.

    So, getting back to that portfolio, it turns out that you need a lot more PV capacity than you can use most of the MWh from, you need a lot more wind capacity than you can use the MWh from, and you’ll need a substantial amount of storage to boot. Dropping energy from wind and PV is equivalent to raising its price, and storage is already pricey.

    My advice to RE boosters (and I are one) is to

    - DROP the commodity thinking
    - DROP LCOE, grid parity and other complete nonsense
    - get serious about the SYSTEM — which includes seriously learning a lot about power markets and why the system is at it is.
    - focus on providing value

    When this happens, PV will take off for real.

    – dave j

    PS — I’ll give the PV fanatics this. If someone cracks the right price level for retail level energy storage enabling complete grid disconnect (not leaning on it for anything at all) then we can see something truly disruptive and exciting happen.

    • Dan Staley says:

      My advice to RE boosters (and I are one) is to

      - get serious about the SYSTEM — which includes seriously learning a lot about power markets and why the system is at it is.
      - focus on providing value

      I think all these are right, and the other part of the value issue is the old grid that is still bumbling along. A few more Sandys and millions of freezers of ruined food, plus the continued price decline and there will be real movement.

      And the leasing model.

    • “…putting together a portfolio of resources that can provide the level of service to which we’ve all become accustomed.” The one prediction I can make with 100% certainty is that the world in 2030 will not tbe the one to which we’ve been accustomed.

  7. Dan Staley says:

    Well, I’m in the middle of writing a handbook for solar site analysis for several trades, so there are enough of us out there who think this can ramp up in cities with the technology coming out of the labs within the next decade.

    As to whether we can keep ecosystems from flipping with this sort of technology, well, at this rate I think all we can do is hope. Especially when you can see the new gas flaring on the Bakken from space, we find that VOCs and CH4 from fracking are more than they told us (and Obama may have suppressed a report stating such), rainforest damage from drought is already bad, and and and…

    • CharlesWT says:

      “…rainforest damage from drought is already bad, …”

      For the pass three decades, the planet has been getting greener.

      • Dan Staley says:

        For the pas[t] three decades, the planet has been getting greener.

        The preciousness of someone actually citing Craig Idso notwithstanding, trying comically to play off CO2 as plant food has nothing to do with the latest news about Amazonian rainforests. But thanks.

      • JMG says:

        This strikes me about the same as when some guy claims, based on some random bit of disinformation gleaned from Fox, that the observed global thermodynamic disruption is somehow actually negative if you squint just right, choose your data carefully, and wear just the right blinders.

        We all better hope the world has not been greening for decades, because if that’s true, then man are we ever screwed, because we’ve only seen about half the effect of the co2 already emitted; if the disruption we’ve already enjoyed was this bad and occurred in the face of a great period of high biological storage of carbon (that’s what the claim of a greening planet is), then when the greening rate slows even as our emissions have increased, whamo, we’re toast.

Post a Comment

Logged in as James Wimberley. Log out »