At Tuesday’s IEEE Nanotechnology Forum, Phil Metz, Director of Business Development for SolFocus, discussed their technology in his presentation “SolFocus Concentrator Photovoltaics – An Introduction“. Though I enjoyed learning about their concentrator photovoltaic (CPV) technology (the presentation was appropriately focused for the audience), I had a greater appreciation for their integrated system approach including the economics. This was evident in the non-technical details he shared. As an early adopter with a residential photovoltaic (PV) system, I was surprised when comparing systems beyond the core technology.
Both CPV and PV systems convert the energy radiated from the sun to direct current (DC) power. Most “grid tie” systems then use an inverter to convert the DC power to alternating current (AC) power which is then fed into the power grid. Beyond these basic similarities, there are large differences in technology, complexity, and economics between the systems.
Unlike “traditional” flat PV panels – typically silicon wafer or thin film based – that are mounted at a fixed angle on a roof or other structure, CPV systems need to actively track the sun to generate power. The SolFocus solution consists of Power Units which contain a reflective optic system to concentrate 650 “suns” on a high-efficiency type III-V semiconductor solar cell. Twenty of these Power Units form a panel (350 W peak power per panel). Twenty-eight panels (560 total Power Units) are grouped on a two-axis tracker which is in turn mounted on a pedestal base to comprise a single CPV System. The tracker unit needs to keep the Power Units aligned within one degree of the sun for proper operation (a fairly high optical acceptance angle when compared to 0.3 to 0.4 degrees for refractive optics based upon Fresnel lenses). Due to the physical size (approximately 21′ W x 16′ H for the array itself) and power output (~ 9.24 KW) of the CPV System along with operational complexity, these systems are targeted at the commercial and utility scale installations not the residential market.
Dr. Metz told us that SolFocus “didn’t start with a science project and then figure out how to make it” as their priorities were manufacturability, then reliability and lastly efficiency with the goal of lowering the total cost to produce and maintain the systems. This year they expect to ship 50 MW from their existing 20 K sq. ft. factory and expand production to 1,000 MW next year. From a capital expense (CapEx) point of view to manufacture a panel, SolFocus is between $0.18 and $0.24 per Watt versus approximately $2 / W for thin film and approximately $1.25 / W for traditional silicon PV.
When our home system was installed eight years ago, Dollar per Watt at peak output ($/Wp) was the key metric of the system performance. Since then the Levelized Cost of Energy (LCOE) developed by the National Renewable Energy Lab (NREL) is more widely used since it is a more sophisticated metric that estimates the average cost of energy produced over the expected life a system. LCOE is the net present value (NPV) of all the costs to deploy and operate the system over its lifetime divided by the NPV of the energy produced over the life of the system. He mentioned LCOE only briefly, saying that SolFocus is targeting both improvements to reduce cost (in particular manufacturing expenses) and increasing energy output (including increasing solar cell efficiency) to lower the LCOE of their systems.
Dollar per Watt peak is still relevant when comparing similar systems for the same installation location- such as fixed mount flat panel PV – since the maintenance costs are minimal and the panels with similar peak power output ratings produce the equivalent power output under the same light conditions. However, the advantage of LCOE is that permits the analysis of different types of energy systems regardless of system complexity even with significantly different initial costs, operational costs and lifetime. For example, by using LCOE wind power can easily be compared to CPV power or even nuclear power.
Even though the SolFocus CPV System uses a two-axis tracker which has increased the initial cost and operating expense (requiring both power to operate and maintenance), a surprising difference was the increased energy output of a CPV system with a tracker versus a fixed tilt panel over the course of a day. For CPV and fixed tile PV systems with the same peak output the graph to the right shows the difference. [Source “Key Advantages of Concentrating Photovoltaics (CPV) for Lowering Levelized Cost of Electricity (LCOE)” by Nishikawa & Horne.] He pointed out that this can make a significant difference in economics since much of this additional power is produced during peak hours (typically noon to 5 or 6 PM) which can be valued at up to three times that of non-peak power.
As an overall system, the SolFocus product has a peak efficiency of 26 to 27% (electrical energy produced to sun energy input) versus standard PV which currently averages around 19%. However, the amount of solar energy -Direct Normal Irradiance (DNI) expressed in KWh/m^2 per day -varies around the world. The map above shows the DNI patterns for the United States. The greater the DNI, the greater the energy produced by any system. However, SolFocus has determined that areas with a DNI below 5 KWh/m^2 are not economically attractive today. What is surprising about the DNI data is that it segments their served available market (SAM) on the basis of an individual location due purely to solar radiation levels – i.e. it is independent of the topography and shade present at a site. And these distinctions do not follow any political boundaries (state, county, city, etc.). Currently the North East, North West and Coastal California are excluded from their target markets until they can reduce costs significantly.
In the end, alternative energy comes down to the bottom line regardless of how compelling the technology. There are many factors that influence the economics: cost to manufacture, system complexity, installation costs, and maintenance costs. There are also government subsidies and tax credits that impact the financial analysis not to mention the cost of capital. And there are factors that simply cannot be changed such as the DNI and weather of the target location. As with all technology, careful analysis is required to make sure the financial models are reasonable and built upon solid engineering data. Warranties for new technology, solar in particular since they are typically 20 to 25 years in length, pose a risk to both the seller and buyer. As many of us know, a lot of strange and unplanned things can happen in the real world especially when there are only a few years worth of operating data.