Thursday, June 21, 2007

At the beginning of the Solar Century.

It may drive some economists and people in the ‘big energy’ business crazy (i.e. fossil + nukes), but solar energy is about more than just dollars and cents. Solar is about independence on an individual scale. In the same way that the automobile changed life in the 20th century, solar panels will become so dominant in the coming decades that by mid-century it will touch everyone’s daily lives in dozens of ways.

Never mind the EIA forecasts that predict less than 1% of our energy supply coming from solar in 2030 (or whatever time frame they adopt), within the next two decades solar will become the cheapest source of energy. And that change will radically alter how we power our lives.

My view differs radically from the way existing big energy firms look at the picture, so let me explain the reasons for it. First of all, we don't have to worry about running out of sunlight. Every fossil source, even coal and nuclear will run out in a couple centuries, if not sooner. Yeah an individual solar panel may die after 20 or 30 years but it can be replaced (hopefully recycled) with a better version. Secondly people love solar power...I think its almost hardwired in our genes. Other than cost (and a few grumbles about the industrial processes used to make it--which is improving all the time), I don't think I've ever heard an objection to solar. A few stick-in-the-muds complain about nighttime and clouds, but a bit of storage (i.e. batteries, thermal, hydro) will generally solve this objection.

I see a direct analogy between solar panels and personal computers (compared to power plants and mainframes). There is the distributed versus centralized aspect which is well understood. And both solar and PCs (especially 20 years ago, in the early days) represent a considerable upfront capital outlay. But both solar panels and PCs also offer convenience and ownership benefits which I believe amounts to independence on an individual scale.

The analogy is not perfect since solar panels simply provide power, while PCs have clearly enabled an information revolution, but 20 years ago computers were really just glorified typewriters/high priced calculators. Even 10 years ago the web of connected PCs that we now call the Internet was slightly more than a curiosity, although PCs had demonstrated their worth in the workplace. Solar panels are well placed to come down in price in the coming decade (indeed they have fallen in price by a factor of 4 over the past 14 years), and as they continue to fall in price their uses (the number of power market niches they cost effectively fill) will multiply much like PCs. Solar panels today cost only 2x grid delivered power.

Solar will also allow developing countries to leapfrog (to a large degree) the developed worlds energy infrastructure. In a dozen years it is likely a citizen in Africa will find it cheaper to buy an electric vehicle and a set of solar panels to power them (possible even a portable set allowing refueling anywhere) than to buy an internal combustion engine and maintain the supply chain needed to get gasoline where s/he needs it.

With every doubling of solar panel manufacturing capacity, the cost of solar has fallen by 20%. [Going from 100 MW capacity to 200 MW capacity is a doubling.] By 2010 the world wide solar panel capacity is expected to increase by a factor of 4 (i.e. two doublings) which means costs should fall by over 35%. Even if the rate of cost decline slows in the future (i.e. drops to 15% per doubling), the cost of solar panels will still fall by 50% within about 4 capacity doublings. If we get 2 doublings in 4 years, we should easily reach 4 doublings in 10 years.

The industry is in the midst of a silicon capacity squeeze (the price of silicon, the main ingredient, and about 50% of cost of a solar panel increased by ~2.5x in the past 3-4 years) which has prevented prices from falling at historic rates the past couple years (indeed prices have risen about 10-15%) but that will be resolved over the next year or so as 2x the current silicon capacity comes on line.

Global capacity is almost 2 GW today and should easily reach ~25 GW in the next ten years. At that point solar panels will reach grid parity. No subsidies will be needed. Solar will be the least expensive power option (power markets with above average costs will reach this point sooner). This will represent a ‘tipping point’ in the power markets and demand for solar panels will go through the roof. Literally.

Companies will form simply to lease roof space from homeowners who can't afford the upfront cost of solar and install it for them. Companies like this already exist, but they pay the homeowner with a stable electricity rate, instead of cash upfront. Imagine when every homeowner can get paid a couple thousand dollars a year (for 20+ years) to let someone cover their house with solar. Within fifty years time I think utilities will cease to exist in their electron broadcast model, they may retain a niche as electron traffic cops shuttling energy around the country/grid as needed, but the revenues they generate from producing non-renewable electrons will evaporate.

Solar panel manufactures will be able to sell as many panels as they can make until solar exceeds 100% of annual capacity additions. Which could happen if old power plants are shut down b/c they are no longer economically competitive. Compound annual capacity growth rates will shoot up from ~50% today to over 100% and stay there for years (hopefully panel makers/silicon suppliers will be able to manage the growth better as a result of our current squeeze). While costs of manufacturing panels continues to drop b/c of capacity doubling, prices to consumers will likely stay at about grid parity for several years while more expensive sources are displaced. Solar panel manufactures will make huge (and growing) amounts of money at this point. Eventually supply will exceed demand and prices could fall even further as panel makers play musical chairs with their margins.

A significant benefit of solar power that will become apparent (rather soon) is their predictable performance over long time periods. Businesses like predictability…and homeowners like predictability too. Even today, despite the relative high cost of solar, because it is a long lived (20-30 year) asset, buying a home with solar panels installed reduces your power/utility bill more than the extra cost of the panels increases your mortgage payment. Remarkably then, if you are buying a new house in many US markets, buying one with solar panels improves your cash flow (i.e. saves you money) from day one. Plus you don’t need to worry about energy prices rising (at least the part your solar panels supply) for the next 20-30+ years.


At 10:32 PM, Anonymous Anonymous said...

Daniel, I like your post a great deal. A few thoughts:

1st, I think PV costs are going to fall faster than you estimate. The 15-20% per year is for silicon, and thin film is going to fall faster.

2nd, PV is only cashflow positive after rebates & credits.

3rd, I think PV is already at a tipping point. It's growth rate is hyperexponential, meaning it's % increase per year is rising, and demand is already well ahead of supply. Fortunately, many governments seem able to understand the reality of external costs...

4th, PV is not going to slow down when it hits the rate of capacity addition, because it's consumer based. Consumers don't care about utility planning: they're just goin to keep buying PV, and the utilities will have to figure out what to do with excess capacity.

At 8:45 PM, Blogger Daniel said...

Thanks for your comments Anonymous.

4) Excellent point about solar's ability to exceed annual capacity additions because it is consumer based.

3)Tipping point--demand already exceeds supply...true but demand is driven by government subsidies today, in 10+ years it will be driven by manufacturing costs.

2) PV--cashflow due to rebates...that is why I point out that in certain US markets you can benefit from lower cashflow. Because of low interest rates in Japan I believe that WITHOUT subsidies new homeowners benefit from better cashflow if PV is rolled into a mortgage.

1) While PV may fall faster than I expect, I'd like to be clear that I think they will fall 15-20% per capacity doubling (for silicon), which is only 7-10% per year. Thin film as you point out could work on a steeper cost curve.


Post a Comment

<< Home