Over the years numerous organizations have projected the potential impact of photovoltaic technology on both domestic and international energy supply. The primary difference between these assessments was the timing of the cost reduction necessary for the impact to occur, and the resultant rate of market penetration. In essence, all of the studies assume the basic tenet of the photovoltaic program-an energy cost of approximately 60 to 120/kWh is necessary for economic operation in the utility environment. However, in almost every case, the timing of this cost reduction and resulting market penetration is more optimistic than projected in this report.
In a 1984 assessment'101, the Edison Electric Institute (EEI) described a "PV breakthrough" scenario that included a rapid buildup of manufacturing capability to 10 GW in the year 2000 and very low system cost. In this analysis, PV accounted for 17.7% of the U.S. installed capacity in 2010, with further penetration dictated by plant retirement rates.
In a 1985 publication the Office of Technology Assessment, predicted that photovoltaics would be primarily a centralized generation source capable of providing more than 4.5 GW by 1995 nationwide'111. The cost of the photovoltaic system and the resultant cost of energy were identified as the key challenges facing the technology. A market assessment published by Frost and Sullivan, Inc., in the same year predicted annual domestic PV sales of 306 MW in 1990'121.
EPRI performs periodic assessments of photovoltaic technology development status and costs but does not analyze potential market penetration. EPRI's technology assessments are generally in agreement with those of the U.S. DOE National Program. The Japanese have analyzed the potential for photovoltaics in Japan and predict a potential market of "several ten thousand MW," with the annual demand increasing from 100 MW/year to a few hundred MW/year in the 1990s'131. In his analysis of the beneficial effects of using wind and photovoltaics, Hohmeyer projected the contribution of photovoltaics as high as 6% of the total energy production for Germany'141.
The greatest impact for photovoltaics is predicted in a recent publication of the World Resources Institute'91. In this analysis, photovoltaic systems are used to generate hydrogen. The hydrogen is stored and either converted to base-load electricity or distributed for use as a transportation fuel. The authors project that for very inexpensive photovoltaic systems, characterized by energy costs in the 30/kWh range, the potential market for photovoltaics in U.S. base-load electric power (with storage) is 600 GW. They project that the market for photovoltaics to produce hydrogen fuel would exceed this value, reaching 1100 GW.
Two representatives of the U.S. PV industry who reviewed this white paper believe that the levelized cost of PV will drop substantially faster than projected here. In addition, input we received from other reviewers suggested the same thing, i.e., that our expectations about future reductions in PV prices are too pessimistic. In fact, two cost studies from PV manufacturers predict $1/W PV module costs by 1995-equivalent to about 120 to 150/kWh at the system level. The reviewers made cogent arguments in favor of more rapid price reductions, especially in the case of intensified R&D (Table G-l). As one reviewer stated directly, PV is a research-driven technology--perhaps more so than any other renewable technology~and should be expected to respond with great sensitivity to an enhanced R&D budget.
As a response to the reviews, we developed a scenario of faster cost reductions within this section of the paper. The reader may consider it a minority opinion stated as part of the "Other Perspectives" section. We do not propose it as the most likely scenario, but it is one that would be consistent with the expectations of many within the PV community. In contrast to the figures given in Table G-l, we suggest the following: Under an enhanced RJD&D scenario, we could see a drop in PV prices to 130/kWh in 1995 (instead of 150); 80/kWh in 2000 (instead of 100); and 50/kWh in 2010 (instead of 70), recognizing that intensified R,D&D could reduce costs faster than originally assumed. In addition, an intensified R.D&D budget would change the investment and market climate enough to accelerate the time in which 10 MW and larger PV plants would come on line, thus allowing for more rapid economies of scale~(i.e., more rapid price reductions).
Reviewers also criticized our market projections as too pessimistic. One introduced the concept of a period of explosive growth. We had assumed steady, incremental growth. The reviewer felt that instead, PV would experience a period in which numerous companies would begin module production simultaneously, allowing for much more rapid growth. This would occur when PV becomes fully competitive. About 80/kWh was proposed as the price level at which growth would become explosive. After this period (about 10 years), growth would follow more normal trends. Given the modular quality of PV manufacture and the fact that it is amenable to automation, we agree with the reviewer about the possibility of explosive growth in the PV industry.
A period of explosive growth could occur in every one of the proposed scenarios, but at different times. In our original projections (Table G-l), it would occur about 2020 in the BAU Scenario; 2010 in the R,D&D Intensification Scenario; and 2005 in the National Premiums Scenario. In fact, if progress is more rapid in the R,D&D Intensification Scenario, as we assume in this section, then explosive growth could occur even earlier, around 2000. If a period of explosive growth occurs, all the final market penetrations would be higher. If faster cost reductions are realized in the R,D&D Intensification Scenario, and all scenarios experience a 5 to 10 year period of explosive growth (see Table G-4 for comparison with the baseline projections), then we project the following market penetrations:
• For the BAU Scenario, 4 quads by 2030 (up from about 3 quads)
• For the R,D&D Intensification Scenario, 2 quads in 2010 (up from 0.7 quad), 6 quads in 2020 (up from 1.9 quads), and 12 quads in 2030 (up from 6.7 quads)
• For the National Premiums Scenario, 4 quads in 2020 (up from 3 quads) and 12 quads in 2030 (up from 9.3 quads).
Thus, it is clear that a more optimistic set of assumptions such as those proposed by our minority position would suggest a sizable additional PV market penetration.
PV is a new, untried technology. Its future is uncertain. We can project the success of PV based on the physics of PV devices. We know they will work, and we foresee that they will become inexpensive. However, at this point our tools for projecting cost trends and market penetrations 30 to 40 years into the future are too weak. The reader can use the various projections here as an envelope of possibilities. The main text incorporates a mid-range projection in keeping with the overall white paper approach. The alternative R.D&D Intensification Scenario in this Other Perspectives section is based on the judgment of reviewers who viewed our initial cost projections as too pessimistic. This alternative projection is within the uncertainty band of future projections and may well be correct.
Was this article helpful?