Thursday, June 17, 2010

Solar Scale, revisited

I wanted to look at the solar scale issue again with a few more fundamental numbers. If we start with (round numbers) 4 PWh per year consumed in the US (EIA), a rough estimate for max solar flux of 1kW/m^2 (also rounded). Typical solar cells today are delivering 10-18% efficiency. However, that doesn't mean a panel deployed will generate 150W/m2. There are issues related to latitude, weather, day-night cycles, etc. This multiplier is called the capacity factor and varies in the US from 10-20%.

So lets be OPTOmistic. 20% efficiency, 20% CF (say Arizona), storage systems and grid efficiencies that allow unity transfer to load. Just doing that math gives us 4500 sq miles of this manufactured material.

Now lets add numbers for oil, as a proxy for conversion of transportation to electric. Roughly 12PWh of power is consumed in oil. Roughly 12% of that is aviation, which we'll assume can't be electric for awhile. So lets say 10 PWh. That would entail an additional 11,250 sq mi for a total of ~16,000 sq miles.

This is better substantially than my previous estimate, but also is a very idealized number. Realistic impact of grid efficiency, real distributed nature of sources, not to mention energy storage issues in order to realistically deploy the technology and you can easily get to double this number. Ultimately, this speaks to a more limited use of solar as a distributed, point of use, supplement at peak demand periods (mid day) and use other technologies for 'base load' technology.

Tuesday, May 19, 2009

Great talk by Dave Irvine-Halliday

Here was a great example of technology making a difference in lives of families in the developing world. Simply a couple of LED lamps, a battery, and a small solar panel. What happens is that the families who get these units:
  • Save up to $400/yr in kerosene costs for lighting
  • Safer environment, no fires!
  • Lower pollution, less exposure to combustion by-products
  • Lower CO2, in the trillion ton range if the whole world were switched
  • Lower crime, particularly applicable to Afghanistan and other trouble spots
  • Higher education rates
  • Higher productivity
  • Higher ownership of property
What can this mean? For a family, this can mean a way out of poverty. The savings alone can mean purchasing a home in 2-3 years instead of living in a wood shack. Children can study.

For more information, see the Light up the World Foundation.

Monday, May 18, 2009

Gerry Fine speaks...

Schott's largest business unit is solar. He agrees that price/performance is not there yet. 4 segments: off-grid, residential, commercial, industrial. CSP is a big part of the business-thermal, because heat can be stored. Thermal systems are at 6-8% ROI, low risk, 20 year old tech.

PV: thin film and crystalline. It is very hard to differentiate. Cost/W differences are small and arguable. Reliability of supply and product are key--low risk important in the market.

Germany, largest market for solar in the world, has the solar flux of Alaska...due to subsidies and feed in tarrif. But that tarrif is going away.

Clearly a world-wide market dip, FIT reduced in Germany, cap on subsidies in Spain, and slowly developing American market.

What do the customers really need? Solar is expensive and non-competitive without subsidies. BP stated last week that solar power will never be competitive for BP. PV shows a 78% experience curve. DRAMS are 62%, LCDs are 65%, Limestone mining is 78%...not good enough. Roadmaps will keep industry on that curve, but how to make a discontinuous jump? Needed: order of magnitude type improvements in Si cost, efficiency, designs, radical integrated manufacturing. This is the conventional path. But solar must be much better than grid BECAUSE to use PV, other stuff needs to happen:
  • Smart grids for fluctuating, distributed inputs
  • Metering and Ren Port. Stds
  • Storage
  • Consistent federal policy
  • Installation process
  • Credit
78% isn't good enough.

Vinod speaks...

The solution to energy issues lies in a "Black Swan Event", check it on Wikipedia. He does not expect us to reduce usage, while we may be more efficient in what we use, usage will go up. This is more a question of taking more shots on goal. How do we get the financial system to fund more "science experiements"? This fits the VC model better than other investment models. Government money can be an accelerator, but shouldn't be the driver.

Solar: 1st Solar owns the business, give up on low cost low efficiency. Look into getting up to 50% efficient is the remaining investment niche. Also look into storage, wind an solar have problems. Solar thermal is still the dirty secret. Thin film will get to ~20%, but no higher expected. Concentration may still be important, balance of system cost important.

Integrated systems will make an impact when someone realizes efficiencies by doing so. Upgraded high voltage DC grid is better for us than a smart grid, even though we need both. These approaches get to 15% ish of the national grid on peak demand, the issue is really on the 85%, how do we get low carbon in a big way?

Short term, government policy will drive changes, but long term the sources will become more efficient and pay for themselves without support. Interesting derivative of that statement, if carbon taxes are used to offset other taxes, then the revenue from carbon tax is a short term increment in governmental time scales.

International: even though there is a lot of interest in Europe, Asia, India, there is little innovation; most innovations come from the US and migrate. So how do we extract the ideas? Innovation ecosystem is best in the US, again the government tries to fill the role elsewhere but doesn't do as well. The big markets will be in India, China, and they are unsubsidiezed. So you have to win without subsidies.

What about Nukes? Allow it to compete, and it will not be successful. Nukes depend on low cost of capital, loan guarantees etc. This gives a huge advantage and subsidy. Innovation cycle takes too long, 15-20 years, and you have to pay 100m in fees to qualify the system. Therefore, there won't be innovation at nearly the rate of other technologies, which will surpass it.

In the end, the diversity of opinions and approaches is the key to making innovation work. This is the problem with the government support. Throw out anything the pundits say, don't be afraid to try something, we are all too conservative (according to Vinod).

LEDs for lighting vs. flourescent

Cree presented strong energy results for street lighting and replacing incandescents, but what was interesting to me is the replacement of T8 flourescents with LEDS still leading to a 30% reduction in energy. Why? It turns out the optics of a point source lambertion radiator (LED) allows much more efficient optical design. Cylindrical radiators or even HTS lights have radiation patterns that lead to losses and ultimately make them less efficient than LEDs. Now, the 30% savings is not sufficient alone to drive deployment, but start to combine the better quality of the light, the better color control, and other potential advantages and it is a deployment option even in the 'walmart' scenario.


OIDA is holding its 1st OPTOmism conference in Santa Clara this week. I'll be doing some posts of interesting datapoints picked real time.

The first speaker, Greg Kats from Good Energies, talked significantly about Green Buildings. He makes a point with which I concur that a lot of the discussion is about marketing. The energy payback, with the exception of the issues I've mentioned before, is straightforward. While 'Energy Efficient" is not going to get you any dates in college, "Green Building" is becoming a sexy term. This changes perceptions allowing a more open adoption environment. Already, in some areas, Green SQ Ftg is unavailable.

New data shows that LEED buildings also have
  • Lower vacancy
  • Higher lease rates
  • Higher productivity
  • Higher sales prices
And if you doubt what can be done with efficiency, with California regulation and utility participation, the state has actually held electricity consumption flat since the 70's, not per capita, total!

Buildings, including embedded energy (used to make the materials in the building) consume about 45% of US energy, so the target is huge. Obama administrations is to reduce CO2 net emmissions of buildngs in the US by 80%. Current programs can generate some 50% of that savings, but even more innovation needed to get to 80%.

Thursday, November 6, 2008

Green Financing

Green projects, especially for individuals, have a cashflow problem. Essentially, the upfront expense, whether for new lightbulbs or a solar roof, are large and these projects only pay off over the longer run. In some cases, you are also taking on a maintenance liability with the new system.

There have been in the solar industry arrangements called PPA (Purchased Power Agreements) for several years. In these deals, a third party installs and owns the facility on a rooftop, secures financing, and collects a guaranteed revenue stream from the building owner as he purchases power. Furthermore, any subsidies and tax benefits accrue to the PPA provider, who is also responsible for system maintenance.

There is a similar model being proposed in the lighting industry. Solid state lighting has the potential to greatly exceed the efficiency of existing lighting and could save 15% of national electrical use eventually. But the 'light bulbs' (LEDs) are substantially more expensive than traditional lighting. However, they will last for 20 years. Furthermore, the efficiency of the technology is still improving rapidly. This presents three problems: lighting companies today are in the light bulb replacement business. They will need to change business models if they are to succeed when replacement cycles reach 20 years. And facility owners are not willing to up-front the costs of SSL, especially if two years later even better technology becomes available. Enter a new business model: 'leased light'. Instead of purchasing light bulbs and fixtures, the lighting manufacturer or provider enters a contract to provide light at certain specifications for a certain price (perhaps tied to electricity prices). The lighting provider collects a revenue stream for light, and decides when to install the next generation based on long-term payback of the improved technology vs. the view on electricity prices. Efficiency improvements then accrue to the provider in reduced electrical costs.

Both of these models, while useful for businesses, break down in the residential market. The problem is that homeowners (need to) sell their houses with relative impunity. So any long term agreement or financing can't be guaranteed. In essence, what is needed is an agreement attached to the house, not the owner. Enter the Berkely First project. In this scenario, the city floats a bond to finance solar roofing for residences. The owners install the roofing with city money, and the owner pays for it long term by an additional line item on his property tax bill. If the owner sells, the new owner will continue to pay on the property tax line item until the bond is repaid. So the financing is attached to the building, not the person. The owner is still responsible for maintenance, as with the rest of his house, and it is in his interest as the benefits of reduce electricity bills depend on it.

I've often thought that the biggest problem in going green is going to be financing, not technology. Creative solutions like these will provide relatively painless avenues for people to install the technology.

h/t The Vine