Concentrating the Sun's Light a 1000 X

June 21, 2007 - 08:29:25 AM

Professor David Faiman and his collegues at the Jacob Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev are convinced that the feasibility of realizing concentrator photovoltaic (CPV) systems on a very large (state-wide) scale is imminent. They outline the economic feasibility of an approach using existing proven technology in a paper referenced below. Their economic financing model developed by D. Raviv was based on technology developed by Amonix Inc.

But Faiman has developed what he considers a breakthrough technology that utilizes a 400 square meter (4,300 sq ft) solar concentrating dish capable of achieving 1000 suns (concentration the intesity of the suns energy by a factor of 1000). The dish is lined with 216 triangular mirrors and is supported by a geodesic structure that rotates and tilts according to the position of the sun.

The dish focuses the 1000 X light on a single 10cm x 10cm PV panel and according to latest test results can produce more than 1.5 kW at 15 % effeciency under field operating conditions without visible signs of damage or degradation.

The big improvement comes not with power conversion effeciency but with power out put per square centimeter. In this case were talking about a power generation increase hundreds of times greater due to the effeciency of the concentrating dish. This ultimately translates to the ability to build a large power plant utiliizing less land for the giant dishes and far fewer of the expensive PV panels.

Below is an excerpt from Faiman's paper that explains the financing model:


Energy Policy 35 (2007) 567–576 "Using solar energy to arrest the increasing rate of fossil-fuel consumption: The southwestern states of the USA as case studies" -- D.Faiman, D. Raviv, R. Rosenstreich --- Jacob Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, 84990 Sede Boqer Campus, Israel MST Ltd., 8 Heil Himush St., 75702 Rishon-Le-Tzion, Israel

Raviv estimated that a production facility with an annual throughput of 1GWp of solar collectors and 0.35GW of storage batteries should cost 850M$ to set up. This figure includes an estimated 220M$ for first-time engineering research and development (R&D) costs. For the purpose of the present example we shall assume that the credit line is called upon to provide these 850M$ in four equal installments, at the start of each of the first 4 years. According to Raviv’s estimates, a 1GWp VLS-PV solar plant including 0.35GW of storage should cost 1135M$. This total is based upon 160M$ for the CPV cells (64-million triple-junction cells, 1 cm1 cm each, with nominal efficiency ¼ 32%); 64M$ for concentration optics; 40M$ for inverters (400 units of 2.5MW rating); 586M$ for the balance-of-system (fabrication of 20,000 collectors, trackers, delivery to site, and physical erection); 285M$ for storage batteries (Vanadium Redox flow batteries with 6 h effective storage, which Raviv estimates could be mass-produced for $850kW1). It is worth pointing out that the cost estimates for cells and inverters are based on written price quotations from bone-fide manufacturers, and the BOS estimate exceeds (i.e. is more conservative than) that of a recently published estimate made by the US National Renewable Energy Laboratory (NREL, 2004).

For our baseline example we shall adopt 10bkWh1 as the starting tariff at which electricity will be sold. (Here, for cosmetic purposes, we have rounded Raviv’s original figure upward slightly, in order to provide a simpler baseline as the starting point for our sensitivity analysis.) From this total Raviv deducts an estimated 0.5bkWh1 for O&M costs, and 5% of revenues (up to an annual maximum of 100M$) for general and administrative (G&A) expenses.

Finally, for the baseline case, it is assumed that the credit line must be paid back with 5% interest and that interest is compounded annually based on all moneys owed at the end of each year. All calculations are performed in constant dollars and, for the baseline example, it is assumed that each VLS–PV plant will operate for 30 years without any degradation. Regarding annual energy production from the VLS–PV plants, Typical Meteorological Year data from the geographically dense set of stations in Israel’s Negev Radiation Survey (Faiman et al, 2004) indicate normal direct irradiance (NDI) figures in the range 1949–2508 kWhm2 year1, depending upon the precise site. At 25% effective collector efficiency (i.e. 32% STC cell efficiency, reduced by various loss-factors (Faiman, 2003)) and a further 10% field loss due to mutual shading (Gordon and Wenger, 1991) (caused by packing the collectors into an assumed aperture/land ratio of 1/3), this reduces to 1750–2260kWhkWp1 year1 for electricity production. For the baseline case we have adopted the rounded mean annual production figure of 2000kWh kWp1, which would correspond to a site with NDI ¼ 2222kWhm2 year1. Under such conditions a VLS–PV plant of 1GWp capacity would occupy 12 km2 or 4.63 square miles of land.

The resulting monetary bookkeeping is straightforward and the following results appear:

As shown graphically in Fig. 2, the credit line reaches a maximum value of slightly less than 7B$ during year 13. It becomes fully paid off (including 5% interest) during year 21. From then onward, the negative values on the graph represent profit (before tax), now effectively earning 5% interest.

From Fig. 2 it is apparent that the credit line and interest become paid off well within the assumed 30-year lifetime of the first VLS–PV plant.

This means that after year 21, the real cost of the electricity is the cost of O&M and G&A. Thus, the tariff could in principle be reduced to a level that just covers these expenses for the existing plants, plus a profit margin.

However, suppose we wish to continue the program of introducing one additional 1GWp plant (plus storage) each year. A simple calculation then reveals that such a VLS–PV plant can be fully paid for if the electricity tariff during year 21 is a mere 4.13bkWh1. Of course, the additional plant will generate additional revenue so that even this low tariff could continue to be lowered each subsequent year and still suffice to ‘‘breed’’ new VLS–PV plants! This ability of the VLS–PV construction program to fund itself after the initial investment has been paid off is what we have termed ‘‘double’’ or ‘‘type-2 sustainability’’ (Faiman et al., 2005b, c).

Now, VLS–PV plant no. 1 reaches the end of its assumed 30-year useful life at the end of year 35. This means that if we want to continue the program of increasing solar electricity productivity by 1GWpyear1, then during year 35 we must construct two new VLS–PV plants: one to replace plant no. 1 when it is decommissioned at the end of the year. Furthermore, for each of the following 29 years we must construct two plants every year. But this turns out to be no problem, because if we raise the electricity tariff from 4.13 to 4.45bkWh1 during year 35, the resulting revenue will fully fund the required two new plants. Note that this higher tariff is still substantially lower than our starting figure of 10bkWh1 35 years previously, and it could, in principle, be lowered annually as each additional pair of new plants comes on line. We thus see the third sense in which CPV technology becomes sustainable: The ultimate revenues can be used, not only to breed new plants, but also to replace old plants by new. This is what we have termed ‘‘triple’’ or ‘‘type-3 sustainability’’ (Faiman et al., 2005b, c). Of course, after the second set of 30 years, triple sustainability will require the construction of three new VLS–PV plants each year, and so on and so forth. However, the present study need not look more than 40 years into the future in order to achieve its stated ends.

Photo Caption:University of Nevada in Las Vegas Site

One Amonix 25 kW unit was installed at the Center for Energy Research at the University of Nevada in Las Vegas in March of 2004. This project is a joint effort by UNLV, Amonix, and Arizona Public Service under the direction of Mary Jane Hale of the National Renewable Energy Laboratory (NREL) and funded by the Nevada Southwest Energy Partnership.Fig. Its 225m2 aperture consists of 5760 Fresnel lenses with optical concentration = 206 X each of which illuminates a Si CPV cell of STC efficiency = 25%.


posted in: