Photovoltaics Outlook April 9, 2015 Steve Campbell scampbell@umn.edu University of Minnesota Department of Electrical and Computer Engineering
The Lede We are rapidly approaching an era when the choice to install solar energy will be primarily driven by raw cost, not subsidies. You can expect to see significant distributed and regional utility-scale deployment over the next decade. Very large-scale usage (>~20%), however, requires a solution to the intermittency problem.
Energy Comparison Standard viewpoint: When the cost of solar is below that of conventional power usage will increase sharply: Grid Parity. Levelized Cost of Energy (LCOE) Capital costs: modules (~35%) and balance of system (~65%). The latter includes installation, electronics, land acquisition, etc.) and financing. Fixed and variable (O&M) costs Projected output, utilization rate, and lifetime Projected incentives and penalties
Trends for PV Modules Module cost is dropping as efficiency (now ~16%) and manufacturing improve Modules will be free in three more years DOE Sunshot goal: installed cost of 1$/watt or about $0.06/kW-hr Solar Cell Module Spot prices High ($/W p ) Low($/W p ) Si Module 0.99 0.55 Thin Film Module 0.94 0.52 Jelle et al. Solar Energy Mater. Sol. Cells 100, 69-96 (2012).
The Result of Falling Costs PV production doubles every ~ 2.5 years At the current rate, we will have 1 TW of capacity in 8 years and ~4 TW in 15 years. At peak production, the later would be about 15% of the total energy supply.
Solar Power Map of the US ~3 kwh/m 2 /Day ~8 kwh/m 2 /Day Germany Standard level of solar power: AM1.5 (1 kw/m 2 ). http://www.nrel.gov/gis/solar.html
Cost of Electricity in the US Grid Parity depends on location Hawaii West Texas NJ, CT, MA Total cost is not just generation
Utility Versus Distributed Solar
Distributed Generation and BIPV The patchwork of licensing requirements drives up BOS Unresolved question who pays for local transmission infrastructure? Soft cells lay on top of existing structure. They do not require extensive installation infrastructure. Dow Powerhouse TM shingles are directly incorporated into the house
Examples of Commercial Installations BMW building and distribution center, Munich National Stadium (Kaohsiung), Taiwan
Costco, Richmond CA PG&E Com Rates 1) 0.15 $/kw-hr 2) 0.18 $/kw-hr 3) 0.26 $/kw-hr 4) 0.32 $/kw-hr Grid parity is much easier to achieve for Tier 3 or 4 usage
Utility Scale PV Dozens installations of 20 to 300 MW. Number is increasing rapidly due to falling PV costs Between Yuma and Phoenix 5.2 million modules Currently rated for 290 MW
Problem 1: Utility Scale Storage Store the capacity of 24 hours of Agua Caliente at AM1.5? Energy ~ 2.5x10 16 Joules or 25 Petajoules Existing storage options: Technology Energy Density (MJ/kG) Requirement (kg - tons) Lithium-Ion 0.875 26 B - 30 M Alkaline Battery 0.67 38 B 43 M Lead-acid Battery 0.17 150 B 160 M Supercapacitor 0.018 1400 B 1540 M For Pb-acid this is a 10 story building ~4 km on a side
Problem 2: Distribution
United States Power Grid
Regional PV Utilities 170 W/m 2 7.1 kw-hr/m 2 -day 5.8 kw-hr/m 2 -day 140 W/m 2 In spite of the latitude, southwest Minnesota receives ~80% of the irradiance of the desert southwest Regional PV utilities makes a great deal of sense
PV Technologies
Silicon PV Problem: Cost Si needs ~100 mm to absorb the light Performance depends strongly on crystal quality and purity Cutting Si boules and polishing the wafers leads to a loss of 50 to 70% of the material Multi-crystalline Si
Thin Film 1: Amorphous Silicon Unlike c-si, a-si has a large a Stability issues prevent high efficiency
Thin Film 2: CdTe Leading thin film in manufacture (First Solar) Glass Au Contact Low resistivity n-type CdTe Oxide coating window High resistivity p-type CdTe Oxide coating absorber Cd has environmental concerns
Thin Film 3: CIGS CuIn x Ga 1-x Se 2 has the highest efficiency of the thin films, but it is a complex material that is difficult to manufacture
In spite of thin film inevitability Silicon remains king in PV Fraunhofer ISE, Report July 2014, page 18
Clouds on the Horizon
Efficiency (%) Issue 1: Efficiency Theoretical limit for simple PV c-si mc-si CdTe CIGS
Issue 2: Materials It is hard to imagine a way to scale Si to $0.50 per watt installed. Will TF mfg survive until then? Cells with Cd cannot be deployed in some parts of the world. Opening for CIGS. Material cost and availability for TW PV 1 TW of CIGS requires 55 years of Indium production, but In is heavily used in touch screens, flat panels, etc.
Summary PV is becoming a significant portion of the energy production portfolio Adoption will be primarily market driven Expect to see much more BIPV Utilities will adopt solar for regional needs Long term issues: increasing efficiency, material concerns, equitable cost distribution