Utility Scale Photovoltaics

If you can’t make rooftop photovoltaics pay financially without feed in tariffs, tax credits, accellerated depreciation, rebates, and subsidized loans – and even with all that it’s still barely better financially than just sticking to natural gas or coal fired grid electricity – how on earth can something like this succeed at the utility scale?

Solar fields in Ontario’s vast open spaces.
(Photo: Optisolar)

One reason solar energy still cannot compete financially vs. conventional energy is because the value of future energy output from a photovoltaic system is discounted when calculating, for example, an internal rate of return. But economic models that put a time-value on money – making receipts in the future not worth as much as receipts today – cannot necessarily be applied to energy.

Traditional models of economic analysis for an energy system lasting 50 years treat the free energy in years 11 through 50 as nearly worthless. The underlying assumption when discounting returns beyond 10 years is that BTUs are as fungible as currencies; something that is arguable but not certain. If a society as a whole desires energy independence, a solar energy system’s return on investment in year 50 is no less valuable than the return on investment in year one. (Ref. “Solar Energy Heats Up in India”) You could even argue, since energy intensity improves every year (the amount of real economic value enabled per unit of energy), future units of energy are worth more than present units. So photovoltaics may not yet make compelling microeconomic sense, but they already have real long-term macroeconomic value.

A fairly stealthy, fast growing, vertically integrated photovoltaic company who is staking their strategy on utility scale applications is Optisolar, based in Hayward, California. Owning everything from the manufacturing (and the underlying thin film technology), to the solar fields they build, they have begun construction on what will be the largest photovoltaic field in the world to-date.

It’s interesting that the world’s largest PV array currently is the utility-scale 12-megawatt Erlasee solar park in Germany, and this new 50 megawatt plant built by Optisolar is going to be Ontario, Canada. Interesting because Germany and Canada aren’t necessarily considered the sunniest places on earth.

Also interesting is the differences between these two solutions – the Erlasee fields employ a variety of technologies and vendors. Most of their systems use two-axis tracking devices and crystaline photovoltaics, such as from Sunpower. By contrast, Optisolar’s Sarnia solar farm in Ontario will use thin film photovoltaics. Thin film technologies are only now coming into volume production, with First Solar, Uni-Solar, Nanosolar, and Optisolar all opening manufacturing plants within the last two years. Applied Materials supplies tools to manufacture both crystaline and thin film photovoltaic plants, and they have at least a half-dozen major orders to supply equipment for thin-film manufacturing plants. Thin film manufacturing, less than 10% of world manufacturing PV capacity in 2007, could make big gains in the next few years.

Proponants of high efficiency crystaline photovoltaic solutions, or ultra-high efficiency photovoltaic concentrator solutions, correctly point out that limited rooftop space makes the lower efficiency thin film solutions impractical. Even lower installation costs, which makers of flexible thin film solutions such as Solar Integrated have pioneered, where the photovoltaic material is literally unrolled on a rooftop and plugged in, are not a sure thing. Thin film’s lower efficiency not only means it requires more space, but that there are far more connectors, wiring and maintenance surfaces, all of which can completely offset the easier installation. But thin film, which uses far less silicon than crystaline PVs, can be manufactured for less than $1.00 per watt – exactly how much less varies from company to company and with private companies is a closely guarded secret.

Where thin film has surprising promise however is in areas not specifically related to the “name plate” efficiency, wherein, for example, a typical thin film PV panel delivers 5.0 watts of DC current in full sun per square foot of photovoltaic surface, compared to a high-efficiency crystaline PV panel that delivers 15.0 watts per square foot. Clearly crystalline PV has a higher efficiency when pointed directly at full sun. But thin film PV is able to capture sunlight more efficiently in overcast conditions and in conditions where the sun isn’t striking the collector as directly, such as early morning and late afternoon. For these reasons – thin film collectors, in terms of kilowatt-hours produced per day per “name plate” watt output – are about 20% more efficient than crystaline photovoltaics. Thin film is also more efficient than photovoltaics in turning DC current into usable AC power – about 10% more of the DC current that goes into the inverter from thin film photovoltaics turns into AC power when compared to typical crystaline photovoltaics.

These advantages, especially the ability to operate in marginal sunlight without requiring tracking devices, make thin film an especially appropriate choice for solar fields at latitudes such as Ontario, Canada, where Optisolar is building their field. At 902 acres, or 365 hectares, which equates to power output of 11 watts per square meter, or 28 megawatts per square mile, Optisolar obviously isn’t constrained by land availability. Put another way, the Sarnia solar farm will require 22.5 acres for every megawatt of output. The recently commissioned solar thermal field, Nevada Solar One (operated by Acciona North America) is designed to require 6.25 acres per megawatt of output. But what would Nevada Solar One’s space efficiency be if they were located in Ontario, instead of super sunny Nevada?

Most interesting is what Sarnia actually costs. Estimates quoted in a report last year “Ontario Goes Solar” in the Toronto Star suggest the Sarnia 40 MW facility cost about $300 million (US), which equates to $7.5M per megawatt. Not cheap, but comparable to most photovoltaic solutions. Ontario likes solar – since placing their initial 40 MW order with Optisolar, they’ve upped the size of the Sarnia field to 60 MW – which should ensure its place as the biggest photovoltaic installation in the world for a bit longer. And with a feed-in tariff of $.42 per kWh, even in the northern latitudes, $7.5M per megawatt should deliver a good ROI to Optisolar. And if the collectors last the projected 30-50 years, as noted at the beginning of this post, these utility scale photovoltaic solutions should generate long-term economic benefits as well.

4 Responses to “Utility Scale Photovoltaics”
  1. Mike Davis says:

    So in Ontario they will turn 902 acres into a waste land for this field of solar. At Nevada Solar One I am familar with the land scape at that location and it was not much more than waste land to start with. From an Ecology stand point I feel that a nuclear plant surrounded by forest would have been more logical in Ontario.

  2. “Thin film is also more efficient than photovoltaics in turning DC current into usable AC power – about 10% more of the DC current that goes into the inverter from thin film photovoltaics turns into AC power when compared to typical crystaline photovoltaics.”

    Is this at STC or some lower radiation like 600w/m. Has this been verified to the point that you can sell this feature?

    I do not live in Manitoba, but where the solar is capital, Western Colorado. Am disbelieving that would get 10% more output here.

    Would appreciate email on this.

  3. Ed Ring says:

    Fred: The source for that comment is from the California Energy Commission’s PTC ratings (Photovoltaic USA Test Conditions). In their table “eligible PV modules” there is a column listing the nameplate output, then there is a second column showing the PTC output. Thin film consistently outscores crystaline in terms of the ratio of PTC output compared to nameplate output. Here are two examples of thin film:

    First Solar’s FS-60D panel, nameplate 60 watts, PTC output 57.2 watts; 95.3%.

    United Solar Ovonic’s US-60, nameplate 60 watts, PTC output 56.8 watts: 94.7%.

    Compare this to two crystaline PV panels from reputable manufacturers:

    Schott Solar’s ASE-100 panel, nameplate 100 watts, PTC output 89.7 watts; 89.7%.

    Sharp Corporation’s NE160-U1 panel, nameplate 160 watts, PTC output 140.9 watts; 88.1%.

  4. Rob says:

    I stopped reading after 2 paragraphs. Your article is riddled with errors and I hope no one takes it seriously.


Leave a Reply

You must be logged in to post a comment.