Google's RE<C: Making the Perfect the Enemy of the Good?

futurelab default header

by: Michael Hoexter

I’ve generally applauded or appreciated Google’s initiatives in the area of climate and energy. Among large technology firms, Google has seemed to have gotten the basic outlines of the future renewable electron economy.

For one, they embraced carbon neutrality pretty early in the game. For two, they have located at least one of their server facilities so as to take advantage of clean hydro and purchased one of the larger photovoltaic arrays in the world for their Mountain View HQ. Thirdly, they’ve assembled some, what from all reports are, pretty energy efficient “blade” servers for internal use and have been leaders in the Climate Savers Computing Initiative. Fourthly, they have been supportive of electric drive transportation, in particular plug in hybrids. Furthermore Larry and Sergey have put their own money into some of the Valley’s more promising energy plays, including Nanosolar and Tesla Motors.

However, when Google announced RE<C in late
November, the clean energy industry and climate policy community was
delivered something akin to the Apple of Discord or Pandora’s Box, a tempting, shiny but divisive and potentially destructive gift (in fact the German word for poison is “Gift”…coincidence? I think not!). While
Google had received almost universal applause for its previous
initiatives, here was a still splashier and more ambitious clean
energy/climate project that leaves a slightly bitter taste in the
mouths of many of Google’s current and would-be allies.

To those who are not aware of RE<C, I can give you a short, perhaps biased summary of its intent but you can find the official version at the website. RE<C means “renewable energy cheaper than (less than) coal”. The
intention of RE<C is to make renewable energy so cheap that
everybody in the world can afford it, thereby out-competing
climate-altering coal on price. Cheaper than coal means
from one to three cents per kilowatt-hour, which is compared to the
price of energy from old, “paid-for” rather than new coal plants. The timeframe for this transformative vision is a period of years rather than decades. Google has already invested $20 million dollars in RE<C, $10 million each to the solar thermal electric start-up eSolar, an Idealab company and Makani Power, Inc, a high-altitude kite-based wind turbine company. In addition to these technologies, Google has expressed an interest in Enhanced Geothermal Systems technology, a way to scale up geothermal from specialized hot spots to almost every location. RE<C is an initiative that is housed in Google’s venture philanthropy,, which is led by Larry Brilliant, the medical doctor who among other things, helped in the effort to eradicate smallpox.

Replace Coal?

One of the ambiguities of the RE<C announcement
is whether the new Google renewable power plant will function as a coal
replacement. Coal-fired power plants
take a while to start and stop and they are cheap to run so they are
used for always-on baseload power that operates day and night. Sometimes
pulverized coal plants are used to follow the load, meaning gradually
increasing and decreasing power output to meet the daily rise in power
demand. RE<C is to produce one gigawatt of power
generating capacity at a price cheaper than coal yet it is not clear
whether this will be an “always-on” solution that replaces the
functions of typical coal plants. One of the challenges
for most renewable energies has been the lack of controllability and
variations in the strength and availability in the renewable “fuel” for
the electric generator; the sun shines during the day, wind blows
intermittently. So, while it is implied that the new
Google renewables will replace the function of coal plants on the grid,
this goal is not stated in what has been made public so far.

RE<C and the Cheap Energy Contract

RE<C assents to the continuation of what I called in my latest post in the Renewable Electron Economy series, “The Cheap Energy Contract”. The Cheap Energy Contract is an implicit social contract that requires that non-food energy be cheap, really cheap. Not
only does this contract require natural and technological conditions
that allow energy to be cheap but also it is a “contract” because
political leaders are on the hook to ensure that energy remains cheap. This
contract is in particular force in the United States and Canada but, by
dint of lower per-capita income and high demand, it can be said to be
in force in China and India, two of the new industrial giants. One
can say that industrialization requires at least moderately priced if
not cheap energy; otherwise food-powered human labor or feed-powered
animal work would be able to economically compete with the output of
machines powered by electricity, gas or coal.

The Cheap Energy Contract has come under fire
recently from a number of quarters, most notably governments and
activists concerned about global warming. The measures
proposed, cap and trade or carbon tax, are meant to raise the price of
fossil energy to allow clean energy sources, mostly renewables, to
compete. In some countries and proposed here in the US,
feed-in tariffs function as a New Energy Contract in which renewable
electric generators receive higher wholesale electric rates to support
their growth, adding somewhat to electricity prices. Finally, climbing
fossil fuel prices may endanger the Cheap Energy Contract as oil and
natural gas move towards or pass their peak.

RE<C conforms to the existing valuation of energy, be it clean or dirty: it is good and right that it is cheap. Given
that one would think Google’s leaders would be allied with the movement
against climate change, the position of RE<C is surprising. Furthermore, it appears that Google’s policy guru, Dan Reicher has also been spreading the cheap energy love lately, so RE<C has become the Google energy mantra. Nordhaus and Shellenberger in their book BreakThrough,
concur with the cheap energy requirement but suggest that government
subsidy plays a very big role in keeping energy cheap, as well as
helping spur new innovation. Google’s RE<C seems to
suggest that the government spending will not be as crucial, as heroic
inventors will make it all happen on their own.

RE<C as Aid to the Developing World

One of the emphases within the initial
announcement and now in recent public discussion at the 2008 Economic
Forum at Davos, has been the motivation of Larry Brilliant and the two
Google founders to make clean energy affordable for the developing
world. In the, supposedly inspiring, words of Larry Brilliant:

“You can’t succeed just out of conservation because then you won’t have economic development. Find a way to make electricity — not to cut back on it but to have more of it than you ever dreamed of.”

The Davos Forum is a place where the glitterati
and technorati like to emphasize the charitable aspects of their
activity, but the theme of “the China price” has cropped up in most
news conferences on RE<C. At Davos, Larry Page also talked somewhat dismissively of clean energy that is produced at 10 cents per kilowatt-hour, saying that this did not change much in the world.

Techno-pessimists vs. Techno-optimists?

Apparently at Davos, preceding the session with the Googlers, Al Gore and Bono had just been talking about putting a price on carbon. At one point, in Jeff Jarvis’s account of the Google RE<C session, a kind of shadow dispute occurred between Gore in the audience and Brilliant about the nature of the opposition to clean power. The more substantive policy dispute about the acceptable price of clean energy remained unspoken.

At least in Jarvis’s account, but also apparently
in the Google guys’ own thinking, they see themselves as technological
optimists or just plain optimists, fighting the good fight against
those who are obsessed with regulation. Jarvis in his account of his own contribution to the session

“I say from the floor that I see a
cultural difference between the movement and Google on this. Google has
the positive message of the potential for change through technology. I
ask about how they are going to get this message out to encourage
investment from government and the public.”

The framing of
the debate in these terms was very friendly to the Google folks’ point
of view, as Gore and “the movement” appeared to be standing in the way
of progress, while the Google folk were simply putting themselves in
the position to bring energy to the masses.

RE<C: A Reality Check

The orgy of
self-congratulation that seems to be accompanying RE<C has helped
obscure some key weaknesses in its plan and also its destructive
potential. It is shocking to me that a bunch of well
intentioned, well-educated, extremely rich, and intelligent folk can,
in this most crucial matter, have made so many careless statements and
put in place an effort with an unnecessary minefield of expectations.
The potential of a Google renewable energy initiative WAS huge but
unfortunately this one may be hamstrung by good intentions and a dose
of techno-fantasy.

let’s take the attitude betrayed by Larry Page in his somewhat
dismissive remark about start-ups selling 10 US cents a kilowatt-hour
green power. Apparently his goals are still loftier: 3-cents or less. But we will circle back to what you are getting right now for 10 cents a kilowatt-hour in the area of carbon-free energy. We are talking about in the arena of solar thermal electric,
power that is available while the sun is shining and pretty much at the
lowest end of the current price spectrum; some of the very newest
generation of thin film photovoltaics like Nanosolar and First Solar under good light conditions will be close to this number when the sun is shining. For less than 10 cents in high wind areas you can have clean wind power when the wind is blowing. If
you are lucky to live near a strong river or conventional geothermal
resource, you can have clean power round the clock for most of the year
for ten cents or less per kWh. These are in their current
configurations not scalable coal-replacements either because of the
limited availability or inconsistent energy flow. To
scale to the needs of the grid in industrial societies you are still
going to need a lot of new additional energy storage or coal or another
polluting form of energy, which either means carbon pollution or
substantial additional costs per kWh. So we have no 10 cent per kilowatt-hour renewable coal alternative now.

The technology
that currently has the greatest chance for scaling rapidly to be a coal
replacement is solar thermal electric a.k.a. Concentrating Solar Power/CSP with enough thermal storage and solar collectors to bridge day and night. In the 1980’s one CSP plant was built with the capability to generate power around the clock, but informal estimates put costs for a commercial version of these plants somewhere in the area of 25 cents/ kWh in their first generation in the sunniest areas of the world. By
contrast a plant with no storage, i.e. clean solar power only during
the daytime, is now somewhere in the area of 10-12 cents/kWh. In
either configuration, these are very large power plants costing at
commercial scale some fraction of a billion dollars or more.

If Google is
putting itself in the business of putting coal out of business (or at
least on the run) within a period of years it is saying that as a power
business and manufacturing novice it will create a scalable power plant
at 12% of the current estimated cost ($.03 to $.25) for the first
generation of baseload solar thermal plants. Therefore,
unless there is something magical about computer engineering,
charitable intention, or great wealth that transfers to power plant
engineering and manufacturing, Google would be very lucky and skillful
to create a coal-replacement solar (or wind) plant at 10-cents per
kilowatt-hour within a period of years.

such a scalable 10-cent per kilowatt-hour renewable energy baseload
power plant would not only represent an enormous advance but also would
seriously compete with natural gas generation and for the climate
conscious populations of the world stimulate huge demand as a baseload
coal replacement. Ways could be found through carbon-sensitive policy to bring such power plants online very quickly. As
manufacturing techniques improved, those power plants would become more
affordable for the developing world, especially as those governments
start to assign a pricetag to local air pollution let alone GHG
emissions. If this segmentation in time is the “environment vs. development”
argument that the Davos organizers wanted to spur, it is a provocative
conversation starter but may not be doing either issue justice.

Reality Check: Power Density, Accessibility, Storage, and Construction Costs

As I have been emphasizing in my series on the Renewable Electron Economy, fossil and fissionable fuels are notably energy dense. They represent a lot of stored energy per unit weight and volume. This
energy density leads to a high power density, in the form of the
radiation of heat and expansion of gases, of the reactions that convert
these fuels into usable energy. The energy density of the
fuels and the power density of the energy conversion process means
that, while large, the fossil fuel or nuclear power plants can generate
a lot of power with a relatively small footprint, meaning less
(expensive) architecture and machinery per unit power. While
mining the coal, uranium or natural gas, increases the footprint of the
entire power system, these stamps on the landscape are holes or
penetrations into the earth rather than relatively more expensive
pieces of architecture and machinery.

By contrast,
renewable energy has in most of its accessible forms a lower power
density than a coal fire or a nuclear chain reaction: organisms would
be destroyed by the high energy flux in the form of heat, so the
biosphere flourishes more under moderate energy flux conditions. The highest naturally occurring power density can be found in limited or inaccessible areas of the world: high-volume
rapidly flowing rivers (one of the few hospitable high energy areas for
life), the hottest geothermal wells reaching deep in earth’s crust or
mantle, and high in the jet streams. For the highest
power density streams of renewable energy, there are then either
limitations of scale or notable physical and therefore investment
challenges to accessing it.

The more
accessible and easily scalable renewable energy with lower power
density requires larger power plants to capture the same amount of
energy as is found in a coal pile, natural gas well or enriched uranium. Power
plant engineers can correct me on this but back of the envelope
calculations indicate that the average power density at the heat
exchange surfaces of a coal fired power plant is about 3,000 times the
average power density of the desert sunlight, averaged day and night. Furthermore most of this energy flux is intermittent or periodic. Therefore additional storage needs to be constructed, an additional expense.

The point I am
getting at is that power plants and particularly scalable renewable
power plants that can replace coal, are, for the time being, very large
construction projects that require substantial initial investment and,
potentially, a fairly long innovation curve to reduce costs. Unless
someone shows me an argument based on physics rather than wishful
thinking that this scale issue can be circumvented, I think it is
prudent to continue down the path that assumes that they will start
being more expensive until we get some more experience building them;
it is not for a lack of “moxy” or imagination that clean power plants
are currently at this juncture. Therefore the per unit
energy cost will be high until there is an evolution or revolutions in
manufacturing to the point where they are affordable for most of the
world’s people. We have only recently decided that coal
is not “cool” so it is highly unlikely that the first generation of
coal replacement plants will suddenly dip below the energy costs of
natural gas plants, let alone coal.

Furthermore, Google has not had as its business focus the manufacture of objects that live in the outdoors in the elements. Its
main products on the market are valued for their ability to process
signs and symbols in ways that are useful to people though they are
based on energy efficient server architecture that needs to manage
power and heat as a energy consumer. In fact, some of the
superheated price expectations behind RE<C may be fueled by the way
Google’s and other software/Internet businesses have been built on the
platform of Moore’s Law, which has allowed computing energy to go down
as computing power has gone up: few businesses in history have lived in
an environment where the development and efficiency curve has been so
rapid due to this happy intersection of factors. In their
own “Google engineering” self-image they may be downplaying the work
done by many materials scientists, physicists and chip designers at
Intel, AMD, AMAT, etc. Yes, the wealth that Google and the Google
founders have harvested from their business enables the purchase of or
construction of large physical objects and of intellectual property but
this is not the same thing as making cheaply the big and/or remotely
situated things that we need to generate clean power. They might do wonders but it is not a short leap which they are proposing.

Google vs. the Clean Power Industry

In the Davos
session, there was a sense that Google sees the evolution of the
renewable energy industry as prologue to its own efforts, especially
with its insistence on the low price point. Beyond the
particulars of that session, RE<C sets it up to diminish rather than
promote the existing efforts of companies in that industry as it hoping
to leapfrog by a really significant margin, the price-points that seem
to be fair and reasonable to that industry in at least the developed
but also in the richer parts of the developing world. The implication is that

  1. Existing
    renewable energy companies and their engineers are not bright enough;
    Google engineers can step in and do a better job practically from day

  2. There
    hasn’t been the motivation in the renewable energy industry to become
    more efficient and cheaper until Google has come along and declared
    this goal

  3. Scalable
    coal replacement via renewable energy is a trivial technical and
    economic problem compared reducing the cost of renewable power

  4. Google’s
    investments will be qualitatively different than those of venture
    capitalists who have already been investing in cleantech for a number
    of years

But an insulting
or arrogant attitude is really the least of the damage that Google’s
RE<C can do to the renewable power industry that it is at the same
time hoping to join

importantly, Google is confirming the price expectations of the public
that electric energy SHOULD be $.03 or less per kWh without ever itself
being on the hook for producing a clean power generator that competes
on the power plant market. It doesn’t yet know how to run
a sausage business and it is saying that all sausages should contain
truffles at the same price as the cheapest sausage. As
Google has developed a reputation, to this point, of being smart and
successful, and possesses a huge megaphone, it reinforces popular
expectation that high prices for clean energy mean someone is making
out like a bandit, is intellectually lazy, or that inefficiencies are
rife within the clean power industry. In this way,
Google, itself a large consumer of power, is talking as if it is a
power producer even though it may more accurately be viewed as a power
super-consumer. As there are often divergent interests in
negotiating the final price of goods between buyer and seller, Google,
which has never marketed power, appears to be functioning as a fifth
column within the industry it has just dipped its toe into.

In some of its
public statements, Google has expressed concern about the “Valley of
Death” that faces new power generating technologies in the early
commercialization stages. The “Cheap Energy Contract” and
Google’s reinforcement of it only deepens this Valley of Death; in the
current investment environment, most clean energy companies are selling
proportionally more of their products in European countries where there
are premium payments for cleanly generated electricity as well as
stronger government commitment to renewable energy than in the US

Power plant
investors, utilities, policymakers and public utilities commissions, if
they take Google’s RE<C bluff seriously, are going to be sitting on
their hands when a renewable energy plant builder comes up with a
perfectly good scalable renewable coal replacement plant at somewhere
in the area of $0.10 to $0.25/kWh. They’ll say to
themselves, “why should we commission this plant when the Google guys
are going to come up with something at a fraction of the cost in a
couple years time?” Thus yawns the Valley of Death even wider.

in the absence of actual prototypes with transparent cost accounting or
without a good physical, technological, supply-side rationale for their
assertions, Google may be reinforcing the position of the coal industry. It is the coal industry that has specialized in cheap but dirty energy supply. Applying
the criteria of cheap (and controllable output), utilities and public
utilities commissioners are likely to build and permit more coal plants. Contrary
to what Google representatives seem to feel, there is little news value
in their assertions that energy should be cheap; yes linking this with
the needs of the developing world puts a charitable twist to it but
otherwise it is nothing that renewable energy technology companies
haven’t heard before from utilities and regulators.

Making the Perfect the Enemy of the Good

I learned the phrase “making the perfect the enemy of the good” a few years ago and am glad Voltaire invented this pithy warning of the dangers of perfectionism. RE<C is clearly making the perfect the enemy of the good, as highlighted above. Google wants to enter the energy market as the perfect participant: low price, clean (and I assume capable of replacing coal). On the way, aspersions are cast upon the good (better policy, reasonably priced alternatives).

Spectatorship vs. Action

Following from
the search for the perfect, taking action on climate (part of the whole
point of the maneuver) is delayed as we search for the perfect. We are invited to become spectators as buyers, investors, and activists, as the wizards go to work. We don’t start where we are but wait for the much better solution, which may not ever arrive or may not arrive on time. This
is a trivial matter for people waiting for the newest flat-screen TV
but in the area of climate introduces delay in introducing better
policy and starting with the technologies that are achievable today. These delays will affect all of us, not just Google’s reputation and bottom line.

Google: DOE substitute?

If instead of
increased electric rates, Google is suggesting that massive early
investments by private investors will bridge the “Valley of Death”, I
believe they are overlooking the problem of scale, risk, and return on
investment. To build massive prototype power plants of a
new type requires tens if not hundreds of millions of dollars, at which
point one is not even yet developing the manufacturing base to get to
the low, low coal price for renewable energy. Furthermore, there are very large risks associated with these types of investments. As
an example, think of the risks associated with floating a large
mechanical object like a wind turbine in the jetstream tethered by a
long electric cable; these are not simply soft weather balloons, even
if Makani will use kites as part of the rotors. To absorb
these risks, private capital needs very substantial returns, which in
turn boosts the costs of the plants and the resulting energy.

Historically, investments of this type of scale and risk have been heavily subsidized by or fully funded by the government. With
the exception of a push in the years after the 1973 oil shock, the
Department of Energy has not fully funded clean energy with the goal of
energy independence via using renewable energy. Google
and the Google guys have grown up in an area of lesser government
ability or inclination to fund clean energy, so maybe have not
conceived of a DOE that funds both early research and also facilitates
early commercialization efforts. In fact, they may see
themselves as white knights with regard to government funding of
research, as apparently they have developed a deal where NASA uses
their private jets for research.

I don’t think
Google with all its wealth can afford to substitute for a Department of
Energy that does not have stockholders to answer to and can survive
large losses on exploratory research. Furthermore, private companies cannot substitute for the rule-making and framework creating ability of governments. Maybe
Larry and Sergey’s short historical timeframe and great wealth have
blinded them to the potential of government to help them and the power
industry get clean energy to market. Though I am open to
reasonable counterarguments, energy subsidy has been a part of the
energy game for a very long time and it seems to be more a question of
the goals and amount of that subsidy than a complete substitution of
private for public investment.

Reforming Google’s Renewable Energy Programs

As I appreciate
most of what Google has done so far in the area of Energy and Climate,
I am hoping that they can change the direction of their renewable
energy programs to act as a facilitator of the transformation of our
energy system in the direction of clean energy sources. Investing as they have in start-ups is a good idea and should be applauded. Getting involved in energy policy debates that facilitate the building of clean power plants is a step in the right direction. However,
larding on top of this the demand and insistence that the products be
priced extremely low with no reality-based cost-accounting and
physics-based technological rationale, weakens the effect of Google’s
plans and tampers with the effectiveness of renewable energy technology
firms that can deliver clean power plants at prices within the price
range of energy consumers in more developed societies.

To make the
energy business work, as new demands are placed on it in the form of
carbon-reduction and sustainability, consumers and/or tax-payers are
eventually going to need to help finance the building of a massive new
clean energy infrastructure, through investments in renewable home
energy generators, the spending of tax dollars on clean energy and/or
through electric rates. Google’s RE<C is perhaps a
noble but not particularly well thought-out attempt to circumvent what
is ultimately a reality constraint on our wishes for a better world.

I am in
agreement with the Google folk that this is largely about technology
but the question is what is the acceptable market entry price and what
is the expectable and realistic cost reduction curve that will underlie
the price of achievable technologies. So I am no
techno-pessimist, though perhaps I am less prey to wishful thinking
about the convergence of multiple technological innovations within a
short time frame. Google doesn’t let wishful thinking set the price for Adwords; why should it guide their new energy venture?

Original Post: