Do Photovoltaics have a Future?

Abstract

While there are several commercial businesses selling photovoltaic power supplies, the applications are all in places where conventional electric power is unavailable. The present interest in solar cell electricity for widespread terrestrial use has led to an intense activity exploring the opportunities for significant cost reduction. In this paper obstacles to cost reduction will be identified and the programs addressing techniques for overcoming those obstacles will be presented.

Evaluating the causes of cost reduction in photovoltaic modules

“…Our findings show some similarities to earlier reported results, as well as some differences. Similar to Nemet (2006) we find that increasing module efficiency was the largest contribution to cost reduction in 1980-2001. While Nemet (2006 finds the sources of cost declines to be heavily concentrated in plant size and module efficiency changes, we find cost-reducing effects were spread across a number of variables.…”

“…2012: (Powell et al, 2012), (Mints, 2015). Polysilicon price (p s ) 2015$/kg 126 36 26 1980: (Mints, 2015), (Williams, 1980), (Costello and Rappaport, 1980), (Christensen, 1985). 2001: (Mints, 2015), (Swanson, 2006).…”

“…2012: (Applied Materials, 2011), (ITRPV, 2012), (Mints, 2015). Share of materials costs (θ) unitless 0.69 0.43 0.65 1980: (Williams, 1980). 2001: (Maycock, 1997).…”

“…We derive the value of c from estimated materials costs in the three time periods. Based on the literature, the share of materials costs in PV modules varied between 43-69% (Williams, 1980;Maycock, 1997;Powell et al, 2013). We calculate the materials costs using the total module cost and the fraction due to materials, subtract the cost of silicon, and divide out the wafer area to obtain c.…”

“…This is a round number that reflects more or less the characteristics of the representative module we use (a Solarex module from Block IV generation, which has 72 multicrystallline cells, each of which is 95x95 mm, and an area utilization of about 0.9, in (Christensen, 1985). a) 13.5% for multicrystalline Si cells (Rohatgi, 2003) a) 66 $/kg (Mints, 2015) b) 150 $/kg (Williams, 1980) c) 162 $/kg (Costello and Rappaport, 1980) d) (Side note: In 1975, it was greater than $97/kg (Christensen, 1985)) -Central value used in analysis: mean(66, 150, 162) = 126 $/kg a) 32 $/kg (Mints, 2015) b) 39 $/kg in 2002 (Swanson, 2006) -Central value used in analysis: mean(32,39) = 35.5 $/kg a) 21 $/kg (Mints, 2015) b) 31 $/kg (Powell et al, 2013 (Christensen, 1985) b) 80 cm 2 (Mints, 2015): "1980 was a breakthrough year with wafer diameter sizes moving from 3 inches to 4 inches", which means from 45 to 80 cm 2 . c) (Side note: Later in the 1981-1984 period, 100 cm 2 was typical (Christensen, 1985)) -Central value used in analysis: 90 cm 2 to reflect the characteristics of the representative module we use (a Solarex module from Block IV generation, which has 72 multicrystalline cells, each of which is 95x95 mm, and an area utilization of about 0.9, in (Christensen, 1985) a) 156 cm 2 (Mints, 2015) b) 156 cm 2 (125x125 mm) (Swanson, 2006) -Central value used in analysis: 156 cm 2 a) 243 cm 2 (156x156 mm) (Powell et al, 2012) b) 243 cm 2 (156x156 mm) (Mints, 2015) -Central value used in analysis: 243 cm 2 Silicon thickness (t) µm a) 500 µm (Swanson, 2006) b) (Side note: In 1990, it was 400 µm (Fraunhofer Institute, 2017)) -Central value used in analysis: 500 µm a) 300 µm in 2002 (Swanson, 2006) b) State-of-the-art wafers are in the 250-350 µm range (Sarti and Einhaus, 2002) c) (Side note: In 2004, thickness was around 260 µm for both mono and multicrystalline Si wafers (Green, 2005) unitless 0.20 in 1977 ("About 80% of the purified silicon is left as scrap in the crucibles used to grow large cylinders of the material in crystalline form, or ends up as sawdust when the thin wafers used to make cells are cut from the cylinder.")…”