2008 Solar Technologies Market Report

JANUARY 2010

Energy Efficiency &

Renewable Energy

2008 SOLAR TECHNOLOGIES

MARKET REPORT

Table of Contents

Table of Contents ........................................................................................................................... i

Figures ........................................................................................................................................... iii

Tables ............................................................................................................................................. v

Acknowledgments ........................................................................................................................ vi

List of Acronyms ......................................................................................................................... vii

Executive Summary ..................................................................................................................... ix

Notes .............................................................................................................................................. xi

1. Installation Trends, Photovoltaic and Concentrating Solar Power ..................................... 1

1.1 Global Installed PV Capacity................................................................................................ 1

1.1.1 Cumulative Installed PV Capacity Worldwide .................................................................. 1

1.1.2 Growth in Cumulative and Annual Installed PV Capacity Worldwide ............................. 2

1.1.3 Worldwide PV Installations by Interconnection Status and Application .......................... 4

1.2 U.S. Installed PV Capacity ................................................................................................... 6

1.2.1 Cumulative U.S. Installed PV Capacity ............................................................................. 6

1.2.2 U.S. PV Installations by Interconnection Status ................................................................ 6

1.2.3 U.S. PV Installations by Application and Sector ............................................................... 7

1.2.4 U.S. States with the Largest PV Markets ........................................................................... 9

1.3 Global and U.S. Installed CSP Capacity ............................................................................. 10

1.3.1 Cumulative Installed CSP Worldwide ............................................................................. 10

1.3.2 Major non-U.S. International Markets for CSP ............................................................... 10

1.3.3 U.S. Installed CSP Capacity ............................................................................................ 11

1.4 References ........................................................................................................................... 13

2. Industry Trends, Photovoltaic and Concentrating Solar Power ........................................ 15

2.1 PV Production Trends ......................................................................................................... 15

2.1.1 Global PV Production ...................................................................................................... 15

2.1.2 U.S. PV Production .......................................................................................................... 18

2.2 Global and U.S. PV Shipments and Revenue ..................................................................... 20

2.2.1 Global PV Shipments ....................................................................................................... 20

2.2.2 Global PV Cell/Module Revenue .................................................................................... 23

2.2.3 U.S. PV Shipments .......................................................................................................... 24

2.2.4 U.S. PV Cell/Module Revenue ........................................................................................ 26

2.2.5 U.S. PV Imports and Exports ........................................................................................... 27

2.3 CSP Manufacturer and Shipment Trends ........................................................................... 29

2.3.1 CSP Manufacturers .......................................................................................................... 29

2.3.2 CSP Shipments................................................................................................................. 29

2.4 Material and Supply Chain Issues....................................................................................... 30

2.4.1 Polysilicon Supply for PV ............................................................................................... 30

2.4.2 Rare Metals Supply and Demand for PV ......................................................................... 36

2.4.3 Glass Supply for PV ......................................................................................................... 36

2.4.4 Material and Water Constraints for CSP ......................................................................... 37

2.4.5 Land and Transmission Constraints for Utility-Scale Solar ............................................ 38

2.5 Solar Industry Employment Trends .................................................................................... 39

2.5.1 Types of Jobs in the PV and CSP Industries.................................................................... 40

2.5.2 Current and Projected Employment in the Solar Industry, Global and U.S. ................... 41

2.5.3 Labor Intensity in the PV Industry, Global and U.S. ....................................................... 43

2.5.4 Employment and Labor Intensity in the U.S. and Global CSP Industry ......................... 44

2.5.5 Quality Assurance and Certification for Solar PV Installation ........................................ 45

2.5.6 DOE Response to Current Barriers in Workforce Development ..................................... 45

2.6 References ........................................................................................................................... 46

3. Cost, Price, and Performance Trends ................................................................................... 50

3.1. Levelized Cost of Energy, PV and CSP............................................................................. 50

3.2. Solar Resource and Capacity Factor, PV and CSP ............................................................ 51

3.2.1. Solar Resource for PV .................................................................................................... 51

3.2.2 Solar Resource for CSP ................................................................................................... 53

3.2.3 Capacity Factor, PV and CSP .......................................................................................... 54

3.3 PV Cell, Module, and System Efficiency ........................................................................... 55

3.3.1 PV Cell Type and Efficiency ........................................................................................... 56

3.3.2 PV Module Efficiency ..................................................................................................... 57

3.3.3 PV System Efficiency and Derate Factor ........................................................................ 58

3.4 PV Module Reliability ........................................................................................................ 58

3.5 PV Module Price Trends ..................................................................................................... 59

3.6 PV Installation Cost Trends ................................................................................................ 61

3.7 PV Operations and Maintenance ........................................................................................ 69

3.7.1 PV Operations and Maintenance Not Including Inverter Replacement ........................... 69

3.7.2 PV Inverter Replacement and Warranty Trends .............................................................. 72

3.8 CSP Installation and Operations and Maintenance Cost Trends ........................................ 73

3.9 CSP Technology Characteristics and System Performance................................................ 73

3.9.1 Parabolic Trough Technology .......................................................................................... 73

3.9.2 Power Tower Technology ................................................................................................ 74

3.9.3 Dish-Engine Technology ................................................................................................. 74

3.9.4 Linear Fresnel Reflector Technology .............................................................................. 74

3.9.5 Storage ............................................................................................................................. 74

3.9.6 Heat-Transfer Fluid .......................................................................................................... 74

3.9.7 Water Use ......................................................................................................................... 75

3.9.8 Land Requirements .......................................................................................................... 75

3.10 References ......................................................................................................................... 75

4. Policy and Other Market Drivers .......................................................................................... 78

4.1. Federal Policies and Incentives, PV and CSP .................................................................... 78

4.1.1 Investment Tax Credit...................................................................................................... 78

4.1.2 Renewable Energy Grants ................................................................................................ 79

4.1.3 Manufacturing Tax Credit ................................................................................................ 80

4.1.4 MACRS and Bonus Depreciation .................................................................................... 80

4.1.5 Renewable Energy Loan Guarantee Program .................................................................. 80

4.1.6 Clean Renewable Energy Bonds ...................................................................................... 82

4.1.7 Solar on Federal Property ................................................................................................ 83

4.1.8 State Energy Program ...................................................................................................... 83

4.1.9 Energy Efficiency and Conservation Block Grant Program ............................................ 84

4.1.10 Renewable Energy Production Incentive ....................................................................... 84

4.1.11 Additional Resources ..................................................................................................... 84

4.2 State and Local Policies, Incentives, and Rules and Regulations ....................................... 85

4.2.1 Planning and Permitting ................................................................................................... 85

4.2.2 Interconnection ................................................................................................................ 86

4.2.3 Net Metering .................................................................................................................... 89

4.2.4 Direct Cash Incentive Programs ...................................................................................... 90

iii

4.2.5 Renewable Portfolio Standards and Solar Set-Asides ..................................................... 91

4.2.6 Clean Energy Funds ......................................................................................................... 94

4.2.7 Emerging Trends .............................................................................................................. 95

4.3 Private Sector and Market-Based Developments to Facilitate Solar Deployment ............. 96

4.3.1 Third-Party Power Purchase Agreement Financing ........................................................ 96

4.3.2 Customer Solar Lease Financing ..................................................................................... 98

4.3.3 Property-Assessed Clean Energy Programs ..................................................................... 99

4.3.4 Alternative Financing Structures: Partnership Flips and Leases .................................. 100

4.3.5 Increasing Utility Ownership of Solar Projects ............................................................. 101

4.4 References ......................................................................................................................... 101

5. Investments and Future Outlook ......................................................................................... 105

5.1 Private Investment in Solar Energy .................................................................................. 105

5.2 U.S. Department of Energy Investment in Solar Energy .................................................. 109

5.3 Solar Market Forecasts, PV and CSP ............................................................................... 112

5.3.1 PV Market Forecasts ...................................................................................................... 112

5.3.2 CSP Market Forecasts .................................................................................................... 116

5.4 References ......................................................................................................................... 117

Figures

Figure 1.1 Global cumulative installed PV capacity through 2008 ................................................ 1

Figure 1.2 Cumulative installed PV capacity in top seven countries ............................................. 2

Figure 1.3. Annual installed PV capacity in top seven countries ................................................... 3

Figure 1.4. Global cumulative installed PV capacity by interconnection status ............................ 4

Figure 1.5. Application market share of cumulative installed PV capacity in IEA countries

through 2008 ........................................................................................................................... 5

Figure 1.6. U.S. cumulative installed PV capacity by interconnection status ................................ 7

Figure 1.7. U.S. PV applications, 2008 market shares ................................................................... 7

Figure 1.8. Annual trend in number of U.S. grid-connected PV installations by sector ................ 8

Figure 1.9. U.S. annual grid-connected PV capacity ...................................................................... 9

Figure 1.10. Annual grid-connected PV capacity and cumulative market share in top U.S. state

markets, 2004–2008 ................................................................................................................ 9

Figure 1.11. Concentrating solar power plants of the southwest United States ........................... 12

Figure 2.1. Global annual PV cell/module production by region ................................................. 15

Figure 2.2. Top global PV cell/module producers 2008 ............................................................... 16

Figure 2.3. Global annual PV cell/module production by manufacturer 2002–2008 ................... 17

Figure 2.4. U.S. annual PV cell/module production ..................................................................... 18

Figure 2.5. U.S. annual PV cell/module production by U.S. manufacturer .................................. 19

Figure 2.6. Top U.S. PV cell/module producers 2008 .................................................................. 19

Figure 2.7. Global annual PV cell/module shipments by region .................................................. 21

Figure 2.8. Global annual PV cell/module shipments by manufacturer 2004–2008 .................... 22

Figure 2.9. Top global companies for PV cell/module shipments 2008 ....................................... 22

Figure 2.10. Top global companies for PV cell/module revenues 2008 ....................................... 23

Figure 2.11. Global annual PV cell/module shipments by PV technology 1997–2008 ............... 24

Figure 2.12. U.S. annual PV cell/module shipments 1997–2008 ................................................. 25

Figure 2.13. U.S. annual PV cell/module shipments by manufacturer 2004–2008 ...................... 26

Figure 2.14. Top U.S. companies for PV cell/module shipments 2008 ........................................ 26

Figure 2.15. Top U.S. companies for PV cell/module revenues 2008 .......................................... 27

Figure 2.16. U.S. PV cell/module shipments, exports and imports .............................................. 28

Figure 2.17. U.S. exports of PV cells and modules, 2007 destination .......................................... 28

Figure 2.18. Proportion of 2008 polysilicon production categorized by producer ....................... 32

Figure 2.19. Proportion of 2008 polysilicon production categorized by producing country ........ 32

Figure 2.20. Polysilicon supply projections through 2012 ........................................................... 33

Figure 2.21. Polysilicon price projections through 2015 .............................................................. 35

Figure 2.22. PV module sensitivity to polysilicon price and utilization ....................................... 35

Figure 3.1. LCOE for residential PV systems in several U.S. cities in 2008, with and without the

federal investment tax credit ................................................................................................. 51

Figure 3.2 Photovoltaic solar resource for the United States, Spain, and Germany ..................... 52

Figure 3.3. Direct-normal solar resource in the U.S. Southwest .................................................. 54

Figure 3.4. Direct-normal solar radiation in the U.S. Southwest, filtered by resource, land use,

and topography...................................................................................................................... 54

Figure 3.5. PV capacity factors varying by insolation and use of tracking systems ..................... 55

Figure 3.6. Best research cell efficiencies 1975–2009 ................................................................. 57

Figure 3.7. Best-in-class commercial module efficiencies, 1999–2008, compiled from module

survey data ............................................................................................................................ 58

Figure 3.8. Global, average PV module prices, all PV technologies, 1980–2008 ........................ 60

Figure 3.9. Installed cost trends over time .................................................................................... 62

Figure 3.10. Module and non-module cost trends over time ........................................................ 63

Figure 3.11. Average installed cost of residential systems completed in 2007 in Japan, Germany,

and the United States ............................................................................................................ 64

Figure 3.12. Variation in installed costs among U.S. states ......................................................... 65

Figure 3.13. PV system size trends over time ............................................................................... 66

Figure 3.14. Variation in installed cost according to PV system size ........................................... 66

Figure 3.15. Comparison of installed cost for residential retrofit vs. new construction ............... 67

Figure 3.16. Comparison of installed cost for crystalline vs. thin-film systems .......................... 67

Figure 3.17. Module, inverter, and other costs ............................................................................. 68

Figure 3.18. PV installer data on component costs ....................................................................... 69

Figure 3.19. Inverter default warranties, 2002–2008 .................................................................... 72

Figure 3.20. Generic parabolic trough CSP cost breakdown ........................................................ 73

Figure 4.1. Distribution of round 1 and 2 CREB allocation, percent of projects by technology.. 83

Figure 4.2. Distribution of round 3 CREB allocation, percent of projects by technology ........... 83

Figure 4.3. Interconnection Standards, November 2009 .............................................................. 88

Figure 4.4. Net Metering Policies, October 2009 ......................................................................... 89

Figure 4.5. States that offer direct cash incentives for solar projects, September 2009 ............... 91

Figure 4.6. State renewable portfolio standards and goals, October 2009 ................................... 92

Figure 4.7. Solar capacity to meet existing RPS solar set-aside requirements, November 2009 . 93

Figure 4.8. Estimated system benefit funds for renewables, May 2009 ....................................... 94

Figure 4.9. The residential power purchase agreement ................................................................ 97

Figure 4.10. Property-assessed clean energy programs, November 2009 .................................. 100

Figure 5.1. Global capital investments in solar energy ............................................................... 106

Figure 5.2. U.S. capital investments in solar energy .................................................................. 107

Figure 5.3. Global venture capital and private-equity investments by solar technology ............ 108

Figure 5.4. DOE SETP budget history from FY 2001 to FY 2009 ............................................ 109

Figure 5.5. 2007 and 2008 Solar America Cities ........................................................................ 111

Figure 5.6. Global total PV module production forecasts .......................................................... 113

Figure 5.7. Global thin-film PV module production forecasts ................................................... 114

Figure 5.8. Global PV module demand forecasts ....................................................................... 115

Figure 5.9. Global PV module and system price forecasts ......................................................... 116

Tables

Table 1.1. Global Installed CSP Plants ......................................................................................... 10

Table 1.2. CSP Plants Under Construction, by Country ............................................................... 11

Table 1.3. Installed CSP Plants in the United States .................................................................... 12

Table 2.1. CSP Component Manufacturers .................................................................................. 29

Table 2.2. Annual U.S. Shipments: Parabolic Dish and Trough, 1998–2007 .............................. 29

Table 2.3. Global PV Labor Intensity in 2008 (Direct and Indirect Jobs) .................................... 43

Table 3.1. Ideal CSP Land Area and Resource Potential in Seven Southwestern States ............. 54

Table 3.2. Module Price, Manufacturing Cost, and Efficiency Estimates by Technology, 2008 . 61

Table 3.3. Summary of Arizona PV System O&M Studies, Not Including O&M Related to

Inverter Replacement/Rebuilding ......................................................................................... 71

Table 4.1. DOE Loan Guarantee Programs .................................................................................. 81

Table 5.1. Global CSP Planned Projects, Capacity by Country, Through 2015 ........................ 117

Table 5.2. Global CSP Planned Projects, Market Share by Country, Through 2015 ................. 117

Table 5.3. U.S. CSP Power Purchase Agreement Pipeline, Market Share by Technology,

Through 2015 ...................................................................................................................... 117

Acknowledgments

Primary Authors: Selya Price (NREL) and Robert Margolis (NREL).

Contributing Authors: Galen Barbose (LBNL), John Bartlett (New West Technologies),

Karlynn Cory (NREL), Toby Couture, Jennifer DeCesaro (Sentech, Inc), Paul Denholm (NREL),

Easan Drury (NREL), Mark Frickel (Sentech, Inc), Charles Hemmeline (DOE SETP), Ted James

(NREL), Michael Mendelsohn (NREL), Sean Ong (NREL), Alexander Pak (MIT), Lauren Poole,

Carla Peterman (LBNL), Paul Schwabe (NREL), Arun Soni (Sentech, Inc), Bethany Speer

(NREL), Ryan Wiser (LBNL), and Jarett Zuboy.

For providing information, consultation, advice, and feedback, the authors thank: Doug

Arent (NREL), Bill Babiuch (Midwest Research Institute), Justin Baca (SEIA), Lynn Billman

(NREL), Katie Bolcar (DOE SETP), Austin Brown (DOE EERE), Christopher Cameron (SNL),

Barry Friedman (NREL), Rachel Gelman (NREL), Tom Kimbis (The Solar Foundation), Claire

Kreycik (NREL), Mark Lausten (Sentech, Inc), Jim Leyshon (NREL), Stephen Lommele

(NREL), John Lushetsky (DOE SETP), Kevin Lynn (DOE SETP), Marie Mapes (DOE SETP),

Jeff Mazer (U.S. Department of Commerce, National Institute of Standards and Technology),

Lynn McLarty (Technology & Management Services, Inc.), Jim McVeigh (Sentech, Inc), JoAnn

Milliken (DOE EERE), Hannah Muller (DOE SETP), Judy Oberg (NREL), Jay Paidipati

(Navigant Consulting), Christie Poimbeauf (Technology & Management Services, Inc.), Billy

Roberts (NREL), David Rodgers (DOE EERE), Thomas Schneider (NREL), Scott Stephens

(DOE SETP), Frank “Tex” Wilkins (DOE SETP), Peter Wong (DOE EIA), as well as Meredith

Annex, Leonardo Banchik, Victor Kane, and Jonathan Cronin-Kanevsky, who contributed to the

report during their internships at DOE SETP.

Reviewers included: Bill Babiuch (Midwest Research Institute), Lynn Billman (NREL), Charles

Hemmeline (DOE SETP), Ted James (NREL), John Lushetsky (DOE SETP), Marie Mapes

(DOE SETP), JoAnn Milliken (DOE EERE), Scott Stephens (DOE SETP), and Frank “Tex”

Wilkins (DOE SETP).

For comprehensive editing and publication support, we are grateful to: Michelle Kubik

(NREL) and Susan Moon (NREL), as well as Jarett Zuboy, and Erik Ness (NREL); and to Stacy

Buchanan (NREL) for her cover design.

For the use of and support with their data, special thanks to: Paula Mints (Photovoltaic

Service Program, Navigant Consulting), Larry Sherwood (Sherwood Associates), Nathaniel

Bullard (New Energy Finance), and Travis Bradford (Prometheus Institute).

For support of this project, the authors thank: The U.S. DOE Solar Energy Technologies

Program (SETP) and the U.S. DOE Office of Energy Efficiency and Renewable Energy (EERE).

NREL’s contributions to this report were funded by the Solar Energy Technologies Program,

Office of Energy Efficiency and Renewable Energy of the U.S. Department of Energy under

Contract No. DE-AC36-08-GO28308. The authors are solely responsible for any omissions or

errors contained herein.

vii

List of Acronyms

a-Si amorphous silicon

AC alternating current

AMT alternative minimum tax

APS Arizona Public Service

ARRA American Recovery and Reinvestment and Act of 2009 (S.1, “Stimulus Bill”)

ASES American Solar Energy Society

BIPV building-integrated photovoltaics

BLM U.S. Bureau of Land Management

BNL Brookhaven National Laboratory

CAGR compound annual growth rate

CdTe cadmium telluride

CEC California Energy Commission

CIGS copper indium gallium (di)selenide

CIS copper indium (di)selenide

CPV concentrating photovoltaics

CREB clean renewable energy bond

CSI California Solar Initiative

CSP concentrating solar power

DC direct current

DG distributed generation

DOE U.S. Department of Energy

DOI U.S. Department of the Interior

DSIRE Database of State Incentives for Renewables & Efficiency

EEGBC Energy Efficiency and Conservation Block Grants

EERE U.S. DOE’s Office of Energy Efficiency and Renewable Energy

EESA Emergency Economic Stabilization Act of 2008 (H.R. 1424, “Bailout Bill”)

EIA U.S. DOE’s Energy Information Administration

EPA U.S. Environmental Protection Agency

EPACT Energy Policy Act

EU European Union

FBR fluidized bed reactor

FERC Federal Energy Regulatory Commission

FIT feed-in tariff

FTE full-time equivalent

FY fiscal year

GW gigawatt

GWh gigawatt-hour

HTF heat-transfer fluid

IEEE Institute of Electrical and Electronics Engineers

IREC Interstate Renewable Energy Council

IRS Internal Revenue Service

ISCC integrated solar combined cycle

ISEGS Ivanpah Solar Electric Generating System

ITC investment tax credit (federal)

kW kilowatt

kWh kilowatt-hour

viii

LBNL Lawrence Berkeley National Laboratory

LCD liquid crystal display

LCOE levelized cost of energy

LSE load-serving entity

M&A mergers and acquisitions

MACRS Modified Accelerated Cost Recovery System (federal)

MENA Middle East and North Africa

MG-Si metallurgical-grade silicon

MOU memorandum of understanding

MT metric ton

MW megawatt

MWh megawatt-hour

NABCEP North American Board of Certified Energy Practitioners

NREL National Renewable Energy Laboratory

NYSERDA New York State Energy Research and Development Authority

O&M operations and maintenance

ORNL Oak Ridge National Laboratory

PACE property-assessed clean energy

PE private equity

PEIS Programmatic Environmental Impact Statement

PPA power purchase agreement

PV photovoltaics

R&D research and development

REC renewable energy certificate

REPI Renewable Energy Production Incentive (federal)

RETI Renewable Energy Transmission Initiative

ROW rest of the world

RPS renewable portfolio standard

SAI Solar America Initiative

SBC system benefits charge

SEGS Solar Electricity Generating Stations

SEIA Solar Energy Industries Association

SEP State Energy Program

SEPA Solar Electric Power Association

SETP U.S. DOE’s Solar Energy Technologies Program

SGIP Small Generator Interconnection Procedure

SNL Sandia National Laboratories

SREC Solar Renewable Energy Certificate

TEP Tucson Electric Power

TES thermal energy storage

TW terawatt

UEDS utility external disconnect switch

UMG-Si upgraded metallurgical-grade silicon

UNEP United Nations Environment Programme

VC venture capital

W watt

WGA Western Governors’ Association

WREZ Western Renewable Energy Zones

Executive Summary

The focus of this report is the U.S. solar electricity market, including photovoltaic (PV) and

concentrating solar power (CSP) technologies. The report is organized into five chapters.

Chapter 1 provides an overview of global and U.S. installation trends. Chapter 2 presents

production and shipment data, material and supply chain issues, and solar industry employment

trends. Chapter 3 presents cost, price, and performance trends. Chapter 4 discusses policy and

market drivers such as recently passed federal legislation, state and local policies, and

developments in project financing. Chapter 5 provides data on private investment trends and

near-term market forecasts.

Highlights of this report include:

• The global PV industry has seen impressive growth rates in cell/module production

during the past decade, with a 10-year compound annual growth rate (CAGR) of

46% and a 5-year CAGR of 56% through 2008. Global production reached 6.9 GW in

2008, led primarily by manufacturers in Europe, China, and Japan. China has realized

very high growth rates in recent years and was tied with Europe at 27% market share in

2008. The United States ranked fifth in 2008 at 6% market share or 0.41 GW of

production.

• Thin-film PV technologies have grown faster than crystalline silicon over the past

5 years, with a 10-year CAGR of 47% and a 5-year CAGR of 87% for thin-film

shipments through 2008. Global thin-film market share increased to 14% in 2008. The

United States was the global leader in thin-film production in 2008, with its top two

manufacturers both thin-film producers, First Solar (CdTe) and United Solar Ovonics or

Uni-Solar (a-Si). First Solar was the second-largest global PV producer in 2008.

• Global installed PV capacity increased by 6.0 GW in 2008, a 152% increase over

2.4 GW installed in 2007. The 2008 addition brought global cumulative installed PV

capacity to 13.9 GW. Leaders in 2008 capacity additions were Spain at 2.7 GW,

Germany at 1.5 GW, and the United States and Italy both at 0.34 GW. Germany

maintained its lead in cumulative installed capacity in 2008 with 5.3 GW, followed by

Spain at 3.4 GW, Japan at 2.1 GW, and the United States at 1.1 GW. The grid-connected

market accounted for 97% of 2008 capacity additions and 94% of cumulative installed

capacity in 2008.

• The United States installed 0.34 GW of PV capacity in 2008, a 63% increase over

0.21 GW in 2007. The 2008 addition brought U.S. cumulative installed PV capacity to

1.1 GW. California continued to dominate the market with nearly 180 MW installed in

2008, bringing cumulative installations to 530 MW or 67% of the U.S. market. New

Jersey followed with 23 MW installed in 2008, bringing cumulative capacity to 70 MW

or 9% of the U.S. market.

• Global average PV module prices dropped 23% from $4.75/W in 1998 to $3.65/W in

2008. Module prices rose slightly from 2002 to 2007 caused by polysilicon supply

constraints, but resumed their downward trend by decreasing from $4.07/W in 2007 to

$3.65/W in 2008. Capacity-weighted, average PV installation costs in the United States

decreased 31% from $10.8/W in 1998 to $7.5/W in 2008. The cost decline of $0.3/W

from 2007 to 2008 corresponds to a $0.42/W decline in module prices over the same

period, whereas installation cost reductions from 1998–2005 were largely attributable to

non-module costs (prices are given in real 2008$).

• Federal legislation, including the Emergency Economic Stabilization Act of 2008

(EESA, October 2008) and the American Recovery and Reinvestment Act (ARRA,

February 2009), is providing unprecedented levels of support for the U.S. solar

industry. The EESA and ARRA provide extensions and enhancements to the federal

investment tax credits (ITCs), including allowing utilities to claim the ITC, a new 30%

manufacturing ITC for solar and other clean energy technologies, and an option that

allows grants in lieu of tax credits for taxpaying corporate entities. The $787 billion

ARRA package includes additional funds for the DOE Loan Guarantee program, DOE

EERE programs, and other programs and initiatives. In addition to federal support, state

and local policies, incentives, rules and regulations, as well as financing developments

continue to encourage deployment of solar energy technologies.

• In 2008, global private-sector investment in solar energy technology topped $16

billion, including almost $4 billion invested in the United States. From 2004 to 2008,

global private sector investment increased more than 25-fold. Each of three major sources

of new investment, venture capital and private equity, debt, and public equity, grew at a

CAGR of more than 67%. Global venture capital and private equity investment in solar

grew at a 4-year CAGR of 68% from $539 million in 2004 to $4.34 billion in 2008. U.S.

venture capital and private equity investment increased from $61 million in 2004 to $2.3

billion in 2008, corresponding to a 4-year CAGR of 148%.

• Solar PV market forecasts made in early 2009 anticipate global PV production and

demand to increase fourfold between 2008 and 2012, reaching roughly 20 GW of

production and demand by 2012. Europe is expected to remain the largest market for

solar power, but the North American market is expected to grow the fastest. Module

prices are projected to decrease 34% from 2008 to 2010, and system prices are projected

to decrease 31% from 2008 to 2010.

• Globally, about 13 GW of CSP was announced or proposed through 2015, based on

forecasts made in mid-2009. Regional market shares for the 13 GW are about 51% in

the United States, 33% in Spain, 8% in the Middle East and North Africa, and 8% in

Australasia, Europe, and South Africa. Of the 6.5-GW project pipeline in the United

States, 4.3 GW have power purchase agreements (PPAs). The PPAs comprise 41%

parabolic trough, 40% power tower, and 19% dish-engine systems.

Notes

• This report includes historical price information and forecasts of future prices. Past and

future prices can be provided as "current/nominal" (actual prices paid in the year stated)

or "real" (indexed to a reference year and adjusted for inflation). In some cases, the report

states whether prices are current/nominal or real. However, some of the published

analyses from which price information is derived do not report this distinction. In

practice, prices are usually considered to be current/nominal for cases in which the

distinction is not stated explicitly.

• In some tables and figures, the sum of numerical components is not equal to the total sum

shown due to rounding. Also, note that calculations such as growth rates were computed

before numbers were rounded and reported. Standard rounding conventions were used in

the report.

• Solar water heating, space heating and cooling, and lighting technologies are not covered

in this report. DOE supports these technologies through its Building Technologies

Program.

1. Installation Trends, Photovoltaic and Concentrating

Solar Power

This chapter presents global and U.S. trends in photovoltaic (PV) and concentrating solar power

(CSP) installations. Section 1.1 summarizes global installed PV capacity, growth in PV capacity

over the past decade, and market segmentation data such as interconnection status and sector of

application. Section 1.2 does the same for the U.S. market and includes a discussion of U.S.

states with the largest PV markets. Section 1.3 presents global and U.S. installed CSP capacity.

1.1 Global Installed PV Capacity

1.1.1 Cumulative Installed PV Capacity Worldwide

Cumulative installed PV capacity worldwide is 13.9 GW, with data from multiple sources

represented in Figure 1.1 (EurObserv’ER 2009, IEA 2009, Mints et al. 2009, Sherwood/IREC

2009).

Germany is the clear leader at 5.3 GW of cumulative installed capacity, followed by

Spain, Japan, the United States, Italy, South Korea, and France. Other European Union (EU)

countries contributed about 0.37 GW of the 0.97 GW attributed to the Rest of World (ROW)

countries.

The capacity of 13.9 GW is a 75% increase over 7.9 GW of 2007 cumulative

installed capacity, for a 2008 addition of approximately 6.0 GW. The 6.0 GW represents a 152%

increase over 2.4 GW installed in 2007.

Figure 1.1 Global cumulative installed PV capacity through 2008

(EurObserv’ER 2009, IEA 2009, Mints et al. 2009, Sherwood/IREC 2009)

Data for the top countries shown in Figure 1.1 are from IEA 2009. 2008 data for EU countries not included in IEA

2009 are from EurObserv’ER 2009 and contribute to the ROW total. U.S. data are from Larry Sherwood/IREC. The

estimate for ROW installed capacity and market share is based on data from IEA 2009 and EurObserv’ER 2009 for

countries not in the top seven; ROW demand share is from Mints et al. 2009.

The IEA 2009 report estimates global cumulative installed PV capacity through 2008 to be 13.4 GW. This is close

to the 13.9 GW shown in Figure 1.1, but Figure 1.1 also includes an overall ROW estimate as described in the

previous footnote.

Germany,

5.3 GW, 38%

Spain,

3.4 GW, 24%

Japan,

2.1 GW, 15%

United States,

1.1 GW, 8%

South Korea,

0.36, 3%

Italy,

0.46 GW, 3%

France,

0.18 GW, 1%

Rest of World,

0.97 GW, 7%

The range of estimates for cumulative installed PV capacity worldwide through 2008 is between

13 and 17 GW, with at least two sources at the upper end of the range (REN21 2009, Boas et al.

2009). The higher estimates likely represent cumulative production or shipments of PV cells and

modules. In a rapidly growing industry, it makes sense that cumulative production should be

greater than cumulative shipments, which should be greater than cumulative installed capacity.

Data presented in this report reflect cumulative global production of 18.5 GW from 1997 to

2008, cumulative global shipments of 15.2 GW from 1997 to 2008, and cumulative installed

capacity of 13.9 GW.

1.1.2 Growth in Cumulative and Annual Installed PV Capacity Worldwide

As illustrated in Figure 1.2, Germany’s cumulative installed PV capacity reached 5.3 GW as of

the end of 2008. This is a 38% increase over 2007 cumulative installed capacity of 3.9 GW.

Germany’s market for PV has been supported by a feed-in tariff (FIT) since 2000, providing a

guaranteed payment for a 20-year period for PV-generated electricity feeding into Germany’s

grid. Germany’s PV market experienced its highest annual growth year in 2004, a 290% increase

from 0.15 GW in 2003 to 0.60 GW in 2004, coinciding with an amendment enhancing and

streamlining Germany’s FIT (called Erneuerbare-Energien-Gesetz [EEG]).

Figure 1.2 Cumulative installed PV capacity in top seven countries

(IEA 2008, IEA 2009, Sherwood/IREC 2009)

The revision to the EEG included an overall increase in the per-kWh payment for PV-generated electricity among

other adjustments such as the setting of degression

rates and the specification of payment rates according to PV-

system type (building versus ground mounted) and size.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Cumulative Installed Capacity (GW)

Germany

Spain

Japan

U.S.

Korea

Italy

France

Spain surpassed Japan in 2008 as the number two market as measured by cumulative installed

PV capacity of 3.4 GW, a 410% increase over Spain’s 2007 cumulative installed capacity of

0.66 GW. This tremendous increase in capacity was the result of support from Spain’s FIT. This

was the second year that Spain’s cumulative capacity grew by more than 300%. With 2006

cumulative capacity at 0.14 GW, growth to 0.66 MW in 2007 was a 360% increase.

Japan’s cumulative installed capacity reached 2.1 GW in 2008, an approximately 12% increase

over the 2007 level of 1.9 GW. Japan had the largest market for PV until Germany surpassed

Japan in 2005, coinciding with the end of Japan’s “70,000 Roofs” Program. Japan’s cumulative

installed capacity had reached 1.1 GW by the end of 2004, still greater than Germany’s 1.0 GW

at that time.

With U.S. policy support for PV via the federal investment tax credit, as well as state rebate

programs and other incentives and financing mechanisms, the U.S. PV market experienced a

43% increase in cumulative installed capacity, from 0.77 GW in 2007 to 1.1 GW in 2008.

Other leading markets include Italy, South Korea, and France, with 2008 cumulative installed PV

capacity reaching 0.46, 0.36, and 0.18 GW, respectively. This represents growth of 280% for

Italy (0.12 GW in 2007), 360% for South Korea (0.078 GW in 2007), and 140% for France

(0.075 GW in 2007).

Figure 1.3 presents annual installed PV capacity from 1997–2008 for the seven leading

countries. Spain led in 2008, adding 2.7 GW, an increase of 420% over 0.51 GW installed in

2007. Germany added 1.5 GW in 2008, an increase of 33% (1.1 GW added in 2007). The United

States and Italy were third with 0.34 GW of 2008 additions, a 63% increase for the United States

(0.21 GW added in 2007) and a 380% increase for Italy (0.070 GW added in 2007).

Figure 1.3. Annual installed PV capacity in top seven countries

(IEA 2008, IEA 2009, Sherwood/IREC 2009)

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

2.50

2.75

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Annual Installed PV Capacity (MW)

Spain

Germany

U.S.

Korea

Japan

Italy

France

Other leading markets were South Korea, Japan, and France. South Korea’s annual additions

topped Japan’s, increasing from 0.043 GW in 2007 to 0.28 GW in 2008 (540% growth). Japan

added 0.23 GW, an increase of 7% over the 2007 level of 0.21 GW. France added 0.10 GW in

2008, a 230% increase over the 0.031 GW installed in 2007.

1.1.3 Worldwide PV Installations by Interconnection Status and Application

As illustrated in Figure 1.4, grid-connected PV has represented the majority of installed capacity

additions worldwide since about 1999, increasing its cumulative market share each year: 44% in

1998, 52% in 1999, nearly 92% in 2007, and 94% in 2008. The grid-connected market

contributed 97% of 2008 capacity additions. The faster growth in the grid-connected PV market

has been supported by incentives for grid-connected PV in the top global markets. The grid-

connected market grew at 10- and 5-year CAGRs of 54% and 56%, respectively, while the off-

grid market grew at 10- and 5-year CAGRs of 14% and 15%, respectively.

Grid-connected, cumulative installed capacity represented in Figure 1.4 grew nearly 80% from

7.3 GW in 2007 to 13.1 GW in 2008. Off-grid capacity grew 24% from 0.67 GW in 2007 to

0.83 GW in 2008. Grid-connected markets are typically easier to track than off-grid, because

grid-connected data associated with incentive programs are generally available, whereas off-grid

data are often elusive.

Figure 1.4. Global cumulative installed PV capacity by interconnection status

(EurObserv’ER 2009, IEA 2008, IEA 2009)

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Cumulative Installed Capacity (GW)

Total

Off-grid

Grid-Connected

The continued success of Germany’s PV market and the recent high rates of growth in Spain and

Italy are due largely to support in the form of FITs. This production incentive has greatly

motivated the installation of rooftop, grid-connected PV and large PV power plants in Germany,

large-scale PV plants in Spain, and grid-connected distributed generation in Italy (IEA 2008).

Incentives in the United States and other top markets have also greatly favored grid-connected

additions.

Despite dominance of the grid-connected market worldwide, there is still variation in the types of

PV systems installed in different countries, reflecting various types of subsidy support, market

maturity, demand for particular applications, and various economic and cost factors. More than

half of the countries listed in Figure 1.5 have a majority of grid-connected PV, including

Germany, Spain, Japan, South Korea, Italy, and the United States (toward the right side of the

graph). Cumulative installed capacity for countries such as Sweden, Israel, Malaysia, Turkey,

Mexico, and Norway (shown to the left of the United States on the graph) have a dominant off-

grid residential market. In Canada and Australia, significant portions of the PV markets consist

of off-grid commercial PV installations. In Australia, this capacity consists mainly of off-grid

industrial and agricultural applications (IEA 2009).

Figure 1.5. Application market share of cumulative installed PV capacity

in IEA countries through 2008

(IEA 2009, Mints 2009)

20%

40%

60%

80%

100%

Norway

Mexico

Turkey

Malaysia

Canada

Israel

Australia

Sweden

Denmark

France

Austria

Netherlands

Switzerland

Great Britain

Portugal

Japan

Italy

South Korea

Spain

Germany

Market share of cumulative installed PV capacity

Grid-connected

utility

Grid-connected

distributed

Off-grid

undefined

Off-grid

non-domestic

Off-grid

domestic

1.2 U.S. Installed PV Capacity

1.2.1 Cumulative U.S. Installed PV Capacity

Cumulative installed PV capacity

in the United States topped 1 GW in 2008, increasing 43%

from 0.77 GW in 2007 to 1.1 GW by the end of 2008 (Sherwood/IREC 2009).

Enhanced government support on both the federal and state levels has been critical for expanding

the adoption of solar energy in the United States since 2005. The Energy Policy Act of 2005

(EPACT 2005) raised the federal investment tax credit (ITC) for solar from 10% to 30% for

nonresidential installations and from 0% to 30% for residential installations. EPACT 2005

capped the ITC for residential solar installations at $2,000, but this cap was removed by the

Emergency Economic Stabilization Act of 2008 (EESA), effective January 1, 2009. On the state

level, large incentive programs, such as the California Solar Initiative (CSI) and New Jersey’s

Consumer On-site Renewable Energy Program, offered rebates covering a significant proportion

of the up-front costs of PV systems. Other state and local policies, such as renewable portfolio

standards (RPSs) and improved interconnection and net metering rules, have further promoted

the growth of solar energy in recent years.

U.S. PV

installation growth has been accelerating in recent years. The United States installed 0.34 GW in

2008, an annual increase of 63% over 0.21 GW installed in 2007. Growth in annual additions

from 0.14 GW in 2006 to 0.21 GW in 2007 was 44%, and the 5-year CAGR for annual additions

from 2003 to 2008 was 37%.

1.2.2 U.S. PV Installations by Interconnection Status

In the United States, cumulative, installed off-grid PV capacity was higher than grid-connected

capacity until 2004 (Figure 1.6). The grid-connected market has since dominated and continued

to increase its market share (65% in 2007, 71% in 2008), which is due to federal and state

incentive support. Of the 1.1 GW of cumulative installed PV capacity in 2008, an estimated

0.79 GW are grid connected and 0.32 GW are off grid.

Includes grid-connected and off-grid installed PV capacity.

The U.S. installed PV capacity data presented in this report are based on data from Larry Sherwood, with the full

reference to his report, U.S. Solar Market Trends 2008,

http://irecusa.org/irec-programs/publications-reports/,

provided at the end of Chapter 1. Note that the numbers presented in this chapter are slightly different from those in

Sherwood’s report, as the data used for this report were obtained directly from him in June 2009.

Figure 1.6. U.S. cumulative installed PV capacity by interconnection status

(Sherwood/IREC 2009)

1.2.3 U.S. PV Installations by Application and Sector

In addition to grid-connected versus off-grid, PV installations can be categorized by application,

including building-integrated, rooftop, and ground-mounted PV, and sector, which comprises

residential, commercial, and utility markets.

As shown in Figure 1.7, rooftop PV installations were estimated at 64% or nearly two-thirds of

U.S. installations in 2008. Assuming 0.34 GW or about 340 MW installed in the United States in

2008, as stated in Section 1.2.1, rooftop installations amounted to nearly 218 MW.

Figure 1.7. U.S. PV applications, 2008 market shares

(EuPD Research 2008)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Cumulative Installed Capacity (GW)

Total

Off-Grid

Grid-Connected

64%

10%

22%

Rooftop

Roof-

integrated

Façade-

integrated

Ground-

mounted

During the past couple of years, there has been an increase in the installation of ground-mounted

PV systems. The Boulder City, Nevada, 10-MW, ground-mounted PV project, which came

online in December 2008, is the largest thin-film solar power plant in North America. The

project was developed by Sempra Generation and consists entirely of First Solar thin-film

modules (First Solar 2008). Large ground-mounted PV systems were also installed in 2007 at

Nellis Air Force Base in Nevada (14 MW) and Alamosa, Colorado (8 MW). Only one new large

installation came online in 2008, but many new ground-mounted projects began development in

2008. Examples are Pacific Gas & Electric’s 550-MW Topaz Solar Farm and the 250-MW

California Valley Solar Ranch (Bradford et al. 2008). Large PV systems are expected to increase

in market share, supported by the utilities’ need to meet renewable portfolio standard (RPS) and

solar set-aside requirements and their recently legislated ability to utilize the federal ITC

(expanded ITC passed October 2008).

Figure 1.8 shows that although the number of nonresidential installations including commercial

and utility PV is increasing, the residential sector still accounts for the vast majority of annual

installations. In 2008 nearly 17,000 grid-connected residential installations were installed,

compared to fewer than 2,000 grid-connected nonresidential PV installations.

Figure 1.8. Annual trend in number of U.S. grid-connected PV installations by sector

(Sherwood/IREC 2009)

Although grid-connected residential PV installations have greatly outnumbered grid-connected

nonresidential PV installations for almost a decade, the annual capacity added by new

nonresidential PV installations is much greater because of larger system sizes in the commercial

and utility sectors. As indicated in Figure 1.9, the additional capacity from grid-connected

nonresidential installations accounted for 73% of the 0.29 GW (or 290 MW) of grid-connected

capacity added in 2008.

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Annual Number of Installations

Nonresidential

Residential

Figure 1.9. U.S. annual grid-connected PV capacity

(Sherwood/IREC 2009)

1.2.4 U.S. States with the Largest PV Markets

The top six states in terms of cumulative, installed grid-connected PV capacity as of the end of

2008 were California (530 MW, 67% market share), New Jersey (70 MW, 9% market share),

Colorado (36 MW, 4.5% market share), Nevada (34 MW, 4% market share), Arizona (25 MW,

3% market share), and New York (22 MW, 3% market share).

Figure 1.10. Annual grid-connected PV capacity and cumulative market share

in top U.S. state markets, 2004–2008

(Sherwood/IREC 2009)

100

150

200

250

300

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Annual Installed Capacity (MW)

Nonresidential

Residential

10%

20%

30%

40%

50%

60%

70%

80%

100

120

140

160

180

California

New Jersey

Colorado

Nevada

Arizona

New York

Cumulative Market Share (%)

Annual Installed Capacity (MW)

2004

2005

2006

2007

2008

Cumulative Market Share

Figure 1.10 shows annual installed capacity for the top six states for the past 5 years. California

continued to lead the U.S. market with nearly 180 MW of new grid-connected PV in 2008,

representing 95% growth from the 92 MW installed in 2007. In New Jersey, 23 MW of new

capacity were installed in 2008, amounting to 37% annual growth from the 16 MW installed in

2007. Colorado’s annual capacity additions increased 88% from 11 MW installed in 2007 to

22 MW installed in 2008. Annual installed capacity in Nevada decreased slightly from 16 MW in

2007 to 15 MW in 2008, with most new capacity coming from the 10-MW El Dorado project.

Arizona continued to see steady growth with a 130% increase in installed capacity, from 2.8 MW

in 2007 to 6.4 MW in 2008. In New York, 7 MW were installed in 2008, an 85% increase over

the 3.8 MW installed in 2007.

1.3 Global and U.S. Installed CSP Capacity

1.3.1 Cumulative Installed CSP Worldwide

At the end of 2008, there were 430 MW of cumulative, grid-tied concentrating solar power

(CSP) capacity worldwide, with more than 95% (419 MW) of this global capacity located in the

southwestern United States. By July 2009, global capacity increased to about 550 MW with the

addition of 120 MW in Spain. This reduced the U.S. share to approximately 75%. Of the

550 MW of CSP worldwide, nearly 95% (519 MW) is parabolic trough technology, with the

remainder (31 MW) being tower technology. Table 1.1 lists installed CSP plants worldwide as of

July 2009.

Table 1.1. Global Installed CSP Plants

Plant Name Location Technology Type Year Installed Capacity (MW)

SEGS I - IX

California, United States

Trough

1985–1991

354

APS Saguaro

Arizona, United States

Trough

2005

Nevada Solar One

Nevada, United States

Trough

2007

PS10

Spain

Tower

2007

Puertollano Plant

Spain

Trough

2009

Andasol I

Spain

Trough

2009

PS20

Spain

Tower

2009

Grama et al. 2008

1.3.2 Major non-U.S. International Markets for CSP

Besides the United States, Spain and North Africa are promising markets for CSP. The first

commercial CSP plant in Spain, the 11-MW tower system known as PS10, was completed in

2007. With a 25% capacity factor, PS10 can generate 24 GWh/yr, enough to supply about 5,500

households with electricity (Grama et al. 2008). No new systems were connected to the grid in

Spain in 2008.

As of July 2009, nearly 400 MW of CSP capacity, mostly trough systems, were under

construction in Spain, supported by the government’s feed-in tariff (FIT) policy, which

guarantees payment for electricity produced by a CSP system. The FIT has two major

restrictions. First, the maximum allowable size of a plant is 50 MW. Second, there is an overall

capacity ceiling of 500 MW. During the first half of 2009, construction was completed on the

50-MW Andasol I trough plant and the 50-MW Puertollano trough plant (Solar Millennium

2009, CSP Today 2009). Andasol I became the first commercial CSP plant with energy-storage

capability designed specifically for electricity generation after sunset. This added feature enables

the plant to provide electricity for ~7.5 hours after sunset (Solar Millennium 2009).

The region of North Africa has the potential for sizable CSP growth. Feasibility studies show

that North African power plants could provide about 15% of Europe's electricity needs. One of

the requirements for this to occur is the construction of a 2,000-mile transmission cable

connecting the two continents (Merrill Lynch 2008). Discussion on the matter has occurred, but

major policy decisions have yet to be made. In Morocco, 20 MW of CSP will be hybridized with

a natural gas plant (Global Environment Facility 2009), and similar planning is under way in

Algeria and Egypt. This type of design, known as an integrated solar combined cycle (ISCC), is

gaining some traction in the region. An ISCC plant combines heat from the natural gas turbine

and the solar field, achieving capacity gains without increasing emissions. Another benefit of

such a system is that an additional turbine is not needed when the CSP portion is built. This both

speeds up the construction process and helps keep capital expenditures in check.

On the global level, nearly 600 MW of CSP were in the engineering, procurement, and

construction stages by the end of 2008 (Grama et al. 2008), as summarized in Table 1.2. The

majority were trough systems being built in Spain, with North African countries installing trough

and tower capacity as part of ISCC plants. Mexico and China were also constructing trough

systems.

Table 1.2. CSP Plants Under Construction, by Country

Country

Technology Type

Capacity (MW)

Algeria

Trough

Australia Linear Fresnel Reflector 1

China

Trough

Egypt Tower 31

Mexico

Trough

Morocco Trough 30

Spain

Trough

350

Spain Tower 37

Total

571

Grama et al. 2008

1.3.3 U.S. Installed CSP Capacity

As of the end of 2008, 419 MW of grid-tied CSP capacity had been installed in the southwestern

United States, accounting for more than 95% of global CSP capacity. The Solar Electricity

Generating Stations (SEGS) in the Mojave Desert of southern California account for 354 MW of

this capacity. SEGS comprises nine parabolic trough plants, ranging from 14 to 80 MW, located

in three main locations (Daggett, Harper Lake, and Kramer Junction); this is illustrated in

Figure 1.11. The plants were built between 1984 and 1991 and together have generated more

than 11,000 GWh (BrightSource 2008).

Figure 1.11. Concentrating solar power plants of the southwest United States

(NREL 2009)

As summarized in Table 1.3, the next CSP plant to come online in the United States after the

SEGS plants was the Arizona Public Service (APS) Saguaro 1-MW parabolic trough plant.

Installed in Red Rock, Arizona, in 2005, the system has a capacity factor of 23%, allowing for

the generation of 2 GWh per year (Grama et al. 2008). Another 64 MW of CSP capacity are

accounted for by the Nevada Solar One parabolic trough plant in Boulder City, Nevada.

Connected to the grid in 2007, this plant has a capacity factor of 23% and generates more than

130 GWh each year (Acciona Energy 2008, Grama et al. 2008). The estimated cost of electricity

generated by the Nevada Solar One plant is about $0.18/kWh (Bullard et al. 2008).

Table 1.3. Installed CSP Plants in the United States

Plant Name

Location

Technology

Type

Year

Installed

Capacity

(MW)

SEGS I

Daggett, CA

Trough

1985

SEGS II

Daggett, CA

Trough

1986

SEGS III

Kramer Junction, CA

Trough

1987

SEGS IV

Kramer Junction, CA

Trough

1987

SEGS V

Kramer Junction, CA

Trough

1988

SEGS VI

Kramer Junction, CA

Trough

1989

SEGS VII

Kramer Junction, CA

Trough

1989

SEGS VIII

Harper Lake, CA

Trough

1990

SEGS IX

Harper Lake, CA

Trough

1991

APS Saguaro

Saguaro, AZ

Trough

2005

Nevada Solar One

Boulder City, NV

Trough

2007

Total

419

Grama et al. 2008

Nevada Solar One, 64MW

APS Saguaro, 1MW

SEGS III-VII, 150MW

SEGS I & II, 44 MW

SEGS VIII & IX, 160MW

1.4 References

Acciona Energy. (2008). www.acciona-na.com/. Accessed November 2008.

Boas, R.; Farber, M.; Flynn, H.; Meyers, M.; Porter, C.; Rogol, M.; Song, J. (March 2009).

Photon International. “Looking back—sizing the 2008 solar market.” pp. 88–93.

Bradford, T.; Englander, D.; Maycock, P. (June 2008). “The Growth of Utility Scale PV.” PV

News. Prometheus Institute and Greentech Media.

BrightSource Energy, Inc. (2008). www.brightsourceenergy.com/bsii/history/. Accessed

November 2008.

Bullard, N.; Chase. J.; d’Avack, F. (May 20, 2008). “The STEG Revolution Revisited.” Research

Note. London: New Energy Finance.

CSP Today. (May 12, 2009). “Iberdrola launches its first solar thermal power plant.” Weekly

Intelligence Brief. www.csptoday.com/. Accessed May 2009.

EuPD Research. (2008). Photovoltaics in the USA: Detailed Analysis of a future solar market,

management summary. Bonn, Germany: EuPD.

EurObserv’ER. (2009). Photovoltaic Barometer. Paris, France: EurObserv’ER Consortium.

http://www.eurobserv-er.org/downloads.asp. Accessed July 2009.

First Solar Inc. (December 22, 2008). “First Solar Completes 10MW Thin Film Solar Power

Plant for Sempra Generation.” News Release.

http://investor.firstsolar.com/phoenix.zhtml?c=201491&p=irol-

newsArticle&ID=1238556&highlight=. Accessed November 25, 2009.

Global Environment Facility. (2009). GEF Project Database.

http://gefonline.org/projectDetailsSQL.cfm?projID=647. Accessed July 2009.

Grama, S.; Wayman, E.; Bradford, T.; (2008) Concentrating Solar Power—Technology, Cost,

and Markets. 2008 Industry Report. Cambridge, MA: Prometheus Institute for Sustainable

Development and Greentech Media.

International Energy Agency (IEA). (2008). Trends in photovoltaic applications: Survey report

of selected IEA countries between 1992 and 2007. Report IEA-PVPS T1-17: 2008. IEA

Photovoltaic Power Systems Programme (PVPS). http://www.iea-pvps.org/ Accessed February

2009.

International Energy Agency (IEA). (2009). Trends in photovoltaic applications. Survey report

of selected IEA countries between 1992 and 2008. Report IEA PVPS T1-18:2009. IEA

Photovoltaic Power Systems Programme (PVPS). http://www.iea-pvps.org/ Accessed November

2009.

Merrill Lynch. (2008). Solar Thermal: Not Just Smoke and Mirrors. New York, NY: Merrill

Lynch.

Mints, P.; (2009). Analysis of Worldwide Markets for Photovoltaic Products and Five-Year

Application Forecast 2008/2009. Report # NPS-Global4. Palo Alto, CA: Navigant Consulting

Photovoltaic Service Program.

Mints, P.; Pogrebnyak, L.; Burmon, A. (February 23, 2009). “Bimonthly Photovoltaic Industry

Update.” Solar Outlook. Issue SO2009-1. Palo Alto, CA: Navigant Consulting, Inc.

NREL. (2009). “Concentrating solar power plants of the southwest United States.” Golden, CO:

National Renewable Energy Laboratory (internal only).

Renewable Energy Policy Network for the 21

Century (REN21). (2009). Renewables Global

Status Report 2009 Update. Paris, France: REN21 Secretariat. www.ren21.net/publications.

Accessed July 2009.

Sherwood, L. (2009). U.S. Solar Market Trends 2008. Latham, NY: Interstate Renewable Energy

Council (IREC). http://irecusa.org/irec-programs/publications-reports/. Accessed August 2009.

Solar Millennium. (July 1, 2009). ”Andasol I is officially inaugurated.” Press Release.

http://www.solarmillennium.de/Press/Press_Releases/Andasol_1_is_officially_inaugurated,lang2

,50,1660.html. Accessed August 2009.

2. Industry Trends, Photovoltaic and Concentrating Solar Power

This chapter covers global and U.S. PV and CSP industry trends. Section 2.1 summarizes global

and U.S. PV cell/module production trends, including production levels, growth over the past

decade, and top producers. Section 2.2 presents data on global and U.S. PV cell/module

shipments and associated revenue, including shipment levels and growth over the past decade,

top companies in terms of shipments, shipment levels by type of PV technology, and U.S. import

and export data. Section 2.3 provides information on major CSP component manufacturers and

CSP component shipments. Section 2.4 discusses material and supply-chain issues for PV and

CSP, including polysilicon, rare metals, and glass supply for PV; material and water constraints

for CSP; and land and transmission constraints for utility-scale solar projects. Section 2.5 covers

global and U.S. solar industry employment trends for both PV and CSP.

2.1 PV Production Trends

2.1.1 Global PV Production

The global PV industry has seen impressive growth rates in cell/module production during the

past decade, with a 10-year CAGR of 46% and a 5-year CAGR of 56% through 2008. Annual

growth from 2007 to 2008 was 87%, higher than the 51% annual growth from 2006 to 2007.

Global production (Figure 2.1) reached 6.9 GW for the year 2008, an 87% increase over 3.7 GW

produced in 2007, which was led primarily by manufacturers in Europe, China, and Japan. The

market shares for the top regions/countries are 27% each for Europe and China, 18% for Japan,

12% for Taiwan, 6% for the United States, and 10% for the rest of the world (ROW). From 1997

to 2008, approximately 18.5 GW of PV cells were produced globally.

Figure 2.1. Global annual PV cell/module production by region

(Maycock 2002, Bradford et al. 2006, Bradford et al. 2008a, Bradford et al. 2009)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

7.0

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Annual PV cell/module production (GW)

ROW

Taiwan

China

Europe

Japan

U.S.

Europe and Japan have had strong production growth rates during the past decade, with 10-year

CAGRs of 50% and 38% (from 1998–2008), respectively, resulting in their dominance of the PV

market. Europe and Japan increased their collective market share from 24% in 1980 to 76% in

2004. Since 2004, however, their combined share has dropped to approximately 45%, which is

due to the rapid growth of emerging producers such as China and Taiwan.

China has seen the highest growth rates in recent years, with a 5-year (2003–2008) CAGR of

170%. In 2001, China contributed only about 1% of global production; it did not become a

significant global contributor until 2005 when its market share reached 7%. Taiwan has also

experienced high growth rates, with a 5-year CAGR of approximately 119%, surpassing U.S.

production levels in 2007 to become the world’s fourth-largest producer. Taiwan continued to

gain market share over the United States in 2008. Production in Taiwan increased from

approximately 0.37 GW (10% market share) in 2007 to 0.85 GW (12% market share) in 2008.

The United States produced 0.27 GW (7% market share) in 2007 and 0.41 GW (6% market

share) in 2008.

Figure 2.2 shows 2008 market shares for the top ten PV cell/module producers worldwide, and

Figure 2.3 shows production data for the top ten companies from 2002 to 2008. Japan-based

Sharp Corporation was the global leader in PV production between 2000 and 2006. In 2007,

German-based Q-cells overtook Sharp to become the world’s number-one producer at 0.39 GW

and maintained this position through 2008. Sharp’s market share for production continued to

decrease through 2008, dropping from 9% in 2007 to 6% in 2008. Q-cells also lost production

market share, dropping from 11% in 2007 to 8% in 2008.

Figure 2.2. Top global PV cell/module producers 2008

(Bradford et al 2009)

Q-Cells

0.57 GW

First Solar

0.50 GW

Suntech

0.50 GW

Sharp

0.47 GW

Motech

0.38 GW

Kyocera

0.29 GW

Baoding

Yingli

0.28 GW

JA Solar

0.28 GW

SunPower

0.24 GW

Sanyo

0.21 GW

Other

3.5 GW

49%

Figure 2.3. Global annual PV cell/module production by manufacturer 2002–2008

(Bradford et al. 2006, Bradford et al. 2008a, Bradford et al. 2009)

Q-cells, a crystalline and thin-film cell producer, has seen impressive growth rates since 2003

when it first became one of the top-ten producers. Q-cells had a 5-year (2003–2008) CAGR of

about 83%. The German company’s 2008 production was 47% greater than its 2007 production

of 0.39 GW. Although Q-cells has all of its manufacturing plants in Germany, its plans for future

expansion include development of overseas capacity in places such as Malaysia.

First Solar is a U.S. company with production facilities in the United States, Germany, and

Malaysia. First Solar became a top-ten producer for the first time in 2007 and maintained this

position through 2008. This marked the first time a predominantly thin-film manufacturer made

the top-ten list. First Solar, the largest U.S. PV cell/module manufacturer, was responsible for

36% of total U.S. production in 2008. First Solar is the world’s largest manufacturer of thin-film

(CdTe) modules with 0.15 GW of production in the United States, 0.20 GW of production in

Germany, and 0.16 GW of production in Malaysia in 2008. First Solar’s jump in the global

rankings is a result of high growth rates in recent years. Its 2008 production level grew to

0.50 GW in 2008, up 740% from the 2006 level. First Solar expects its capacity to exceed 1 GW

by the end of 2009. In addition to impressive production numbers, First Solar also boasted the

PV industry’s lowest cell manufacturing cost of $1.08/W in the third quarter of 2008 (First Solar

2008).

Suntech Power, a Chinese crystalline silicon company with production in China, was the third-

largest PV producer in 2008. The company has seen impressive growth rates in recent years,

producing 0.50 GW in 2008, up from 0.33 GW in 2007.

Sharp Corporation also has plans for growth. The Japanese company’s 2008 manufacturing

capacity was 0.86 GW, with most being crystalline cells. The company, however, plans to shift

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

2002 2003 2004 2005 2006 2007 2008

Annual PV cell/module production (GW)

Sanyo

SunPower

JA Solar

Baoding Yingli

Kyocera

Motech

Sharp

Suntech

First Solar

Q-Cells

Other

its focus from crystalline technologies to thin-film PV going forward. Sharp indicated it plans to

upgrade its current 15-MW, thin-film plant to a 160-MW plant. The company is also developing

a 1-GW, thin-film plant within a liquid crystal display (LCD) manufacturing facility. Sharp

estimates that the new factory will begin thin-film shipments in 2010.

Motech Solar, a crystalline silicon cell producer, is the largest PV producer in Taiwan and the

sixth largest globally. The company’s 2008 production of 0.38 GW is a 120% increase over its

2007 production value of 0.18 GW. Motech has indicated plans to expand its PV production

capabilities into China and the United States.

2.1.2 U.S. PV Production

As with the global trend in PV production, the United States has seen strong growth rates in the

past decade, with a 10-year CAGR of 23% and a 5-year CAGR of 32%. Production in 2008

reached 0.41 GW, an increase of 52% over 0.27 GW produced in 2007. Figure 2.4 illustrates

production levels from 1997–2008. In 2003, the U.S. PV industry experienced an 18-MW dip in

production, resulting primarily from the bankruptcy of AstroPower, the second-largest U.S.

producer at that time. AstroPower’s 30-MW production capacity and assets were acquired by GE

Energy for $15 million. After 2003, U.S. production resumed strong growth. Despite increases in

domestic production, U.S. market share in global production has fallen to 6%, which is due to

more rapid growth in other regions/countries such as Europe, China, and Taiwan. The United

States dominates global thin-film production.

Figure 2.4. U.S. annual PV cell/module production

(Maycock 2002, Bradford et al. 2006, Bradford et al. 2008a, Bradford et al. 2009)

Figures 2.5 and 2.6 summarize U.S. annual PV cell/module production by manufacturer.

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Annual PV cell/module production (GW)

Figure 2.5. U.S. annual PV cell/module production by U.S. manufacturer

(Bradford et al. 2006, Bradford et al. 2008a, Bradford et al. 2009)

Figure 2.6. Top U.S. PV cell/module producers 2008

(Bradford et al. 2009)

The United States is the global leader in thin-film PV cell/module production, and accordingly,

the top two U.S. manufacturers for 2008 were both thin-film producers. Thin-film PV

cells/modules produce electricity via extremely thin layers of semiconductor material made of

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

2001

2002

2003

2004

2005

2006

2007

2008

Annual PV cell/module production (GW)

Other

Global Solar

Schott Solar

Evergreen Solar

BP Solar

Solarworld CA (Shell Solar)

United Solar Ovonics

First Solar

First Solar,

147 MW,

36%

United Solar

Ovonics,

113 MW,

27%

Solarworld

CA (Shell

Solar),

61 MW,

15%

BP Solar,

28 MW, 7%

Evergreen

Solar,

27 MW, 6%

Schott Solar,

11 MW, 3%

Global Solar,

7 MW, 2%

Other

cadmium telluride (CdTe), amorphous silicon (a-Si), copper indium gallium diselenide (CIGS),

copper indium diselenide (CIS), or other emerging materials. First Solar is the world’s largest

manufacturer of thin-film (CdTe) modules, with 2008 global production of 0.50 GW, an increase

of 140% from 2007. U.S. production for First Solar in 2008 was 0.15 GW (shown in Figure 2.6

as 147 MW). United Solar Ovonics (or Uni-Solar), a producer of a-Si thin-film technology, was

the second-largest PV producer in the United States in 2008, with a 27% U.S. production market

share. Uni-Solar’s 2008 domestic production reached 110 MW, a 140% increase over its 2007

production level.

In 2008, SolarWorld and BP Solar were ranked third and fourth, respectively, in terms of U.S.

PV production. SolarWorld and BP Solar produced 61 and 28 MW, respectively, in 2008.

Evergreen Solar, fifth in U.S. market share, produced 27 MW in 2008, opened an 80-MW

facility in June 2008, and expects an additional 80-MW capacity at the facility to be completed

in 2009. Schott Solar was ranked sixth, producing 11 MW in 2008. Of Schott Solar’s 80 MW of

global production in 2008, most (88% or 70 MW) were manufactured in Germany. Global Solar,

an Arizona-based company, produced 7.2 MW in 2008; the company primarily manufactures

flexible, thin-film, CIGS-based cells.

2.2 Global and U.S. PV Shipments and Revenue

2.2.1 Global PV Shipments

Shipments of PV cells/modules more accurately reflect demand for PV, as not all cells/modules

end up in the market the year they are produced. Conversely, inventory (produced in an earlier

year) may get shipped out, making it possible for shipments to exceed production in a particular

year and/or for a particular manufacturer. This section presents shipment as opposed to

production data, but does not attempt to compare and reconcile these numbers.

Global shipments of PV cells/modules have seen extensive growth over the last decade, with a

10-year CAGR of 45% and 5-year CAGR of 52% through 2008 (Mints and Tomlinson 2007,

Mints and Tomlinson 2008, Mints 2009). Annual growth from 2007 to 2008 was 79%, higher

than the 55% annual growth from 2006 to 2007.

Global shipments reached 5.5 GW for the year 2008, an increase of 79% from 3.1 GW shipped

in 2007 (Figure 2.7). The market shares for the top regions/countries were 31% for Europe, 22%

for Japan, 19% for China, 11% for Taiwan, 7% for the United States, and 9% for the ROW.

From 1997–2008, a total of approximately 15.2 GW of PV cells/module were shipped globally.

Figure 2.7. Global annual PV cell/module shipments by region

(Mints and Tomlinson 2007, Mints and Tomlinson 2008, Mints 2009)

Europe and Japan have achieved significant growth in PV shipments with 10-year CAGRs of

51% and 42% (from 1998–2008), respectively. Shipments in 2008 represented increases of 71%

for Europe (1.7 GW shipped in 2008) and 36% for Japan (1.2 GW in 2008) over 2007 shipment

levels. Europe and Japan increased their joint market share from 47% in 1993 to 80% in 2005.

Since 2005, however, despite continued growth in shipments, the combined Europe and Japan

market share decreased to 53% in 2008, which is due to even more rapid growth of

manufacturers in China and Taiwan.

China has seen a 4-year CAGR of 230% since 2004. China shipped 1.0 GW in 2008, a 110%

increase from 2007. Taiwan has seen similarly impressive growth with a 4-year CAGR of 110%

since 2004. Taiwan shipped 0.61 GW in 2008, a 110% increase from 2007. In 2006, China’s

shipments surpassed those of the United States to become the third-largest contributor to global

PV shipments. In 2007, Taiwan surpassed the United States to become the fourth-largest

contributor. Both China and Taiwan continue to gain market share over the United States. While

the U.S. market share has declined from 9% in 2005 to 7% in 2008, China and Taiwan have

increased their market shares from 2% and 3% in 2005 to 19% and 11% in 2008, respectively.

Figures 2.8 and 2.9 show 2008 PV shipments for the top global manufacturers. In 2008, the

shipments of two companies, BP Solar and Mitsubishi Electric, were statistically close enough to

warrant inclusion in the top 11 manufacturers. In recent years, Sharp Corporation has been the

leading exporter of PV cells/modules. In 2008, however, Q-Cells overtook Sharp with 0.55 GW,

a 59% increase from 2007. SunTech also overtook Sharp as the second leading exporter with

0.50 GW, a 61% increase from 2007. Sharp shipped 0.46 GW in 2008, a 26% increase from

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Annual PV cell/module shipments (GW)

ROW

Taiwan

China

Japan

Europe

U.S.

2007. In 2007, Sharp experienced a 16% decrease in shipments, a drop from 0.43 GW in 2006 to

0.36 GW in 2007. From 2006 to 2008, Sharp’s growth in shipments totaled 5%.

Figure 2.8. Global annual PV cell/module shipments by manufacturer 2004–2008

(Mints and Tomlinson 2008, Mints 2009)

Figure 2.9. Top global companies for PV cell/module shipments 2008

(Mints 2009)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

2004

2005

2006

2007

2008

Annual PV cell/module shipments (GW)

Mitsubishi Electric

BP Solar

JA Solar

Sunpower

Sanyo

Motech

Kyocera

First Solar

Suntech

Q-Cells

Other

Q-Cells,

0.55 GW,

10%

Suntech,

0.50 GW,

Sharp,

0.46 GW,

First Solar,

0.44 GW,

Kyocera ,

0.29 GW,

Motech,

0.26 GW,

Sanyo,

0.22 GW,

SunPower,

0.22 GW,

JA Solar,

0.21 GW,

BP Solar,

0.15 GW,

Mitsubishi

Electric,

0.15 GW,

Other,

2.1 GW,

37%

First Solar, the world’s leading manufacturer of thin-film PV (specifically CdTe) has

experienced tremendous growth in recent years and overtook Kyocera as the fourth-largest

contributor to global shipments in 2008. First Solar shipped 0.44 GW in 2008, a 130% increase

from 2007. SunPower and JA Solar also experienced significant growth, both with about 110%

increases in shipping from 2007, with 0.22 and 0.21 GW shipped, respectively. The top three

manufacturers, Q-Cells, SunTech, and Sharp, declined in their market shares from 2007 resulting

from the rapid growth of other companies such as First Solar, SunPower, and JA Solar . Q-Cells

declined in market share from 11% in 2007 to 10% in 2008. Similarly, SunTech declined from

10% to 9% and Sharp from 12% to 8%. First Solar increased its market share from 6% in 2007 to

8% in 2008. Both SunPower and JA Solar increased their market shares from 3% in 2007 to 4%

in 2008.

2.2.2 Global PV Cell/Module Revenue

Worldwide revenue from PV cells/modules reached $20 billion in 2008, an increase of 80% from

2007 revenue of $11 billion (Mints 2009). Figure 2.10 provides revenue and associated data for

the top ten global contributors to PV shipments. Q-Cells, Suntech, and Sharp each brought in

close to $2 billion in cell and module revenue in 2008. These three companies each shipped from

nearly 0.46 to 0.55 GW in 2008.

Figure 2.10. Top global companies for PV cell/module revenues 2008

(Mints 2009, company Web sites)

Although there are many types of PV technologies, crystalline silicon cells/modules currently

dominate the world market (Figure 2.11). However, the crystalline silicon market share dropped

slightly in recent years from a 93% share in 2006 to 86% in 2008. Polycrystalline cells currently

Q-Cells,

$1.8 B, 9%

Suntech,

$1.9 B, 9%

Sharp, $1.7 B,

First Solar,

$1.2 B, 6%

Kyocera,

$1.0 B, 5%

Motech,

$0.70 B, 3%

Sanyo,

$0.87 B, 4%

Sunpower,

$0.61 B, 3%

JA Solar,

$0.80 B, 4%

BP Solar,

$0.60 B, 3%

Mitsubishi

Electric,

$0.55 B, 3%

Other,

$8.6 B,

42%

lead the global market with a 49% share of total shipments, followed by monocrystalline cells at

35% of shipments.

Globally, thin-film technologies have grown at a 10-year CAGR of 86% from 1998 to 2008. In

2003, the total market share of thin-film cells was 5% compared to the 95% market share of

crystalline cells. In 2008, thin-film technology claimed a 14% global market share (5% a-Si,

8% CdTe, and 1% other thin films). Forecasts of thin-film market share in 2012 have spanned a

broad range, from 16%–34% (see Section 5.3).

Figure 2.11. Global annual PV cell/module shipments by PV technology 1997–2008

(Mints and Tomlinson 2007, Mints and Tomlinson 2008, Mints 2009)

2.2.3 U.S. PV Shipments

During the past decade, the United States has seen steady growth in PV shipments analogous to

the global PV shipment trends, with a 10-year CAGR of 22% and a 5-year CAGR of 33%

through 2008. Figure 2.12 illustrates the annual growth of PV shipped from the United States. In

2003, shipments experienced a 15% dip from the 2002 level, which was due primarily to the

bankruptcy of AstroPower. In 2004, however, the United States experienced a rapid recovery and

increased shipping levels to 30% greater than 2002 values. From 2004 to 2006, the United States

maintained an average of 0.14 GW in shipments each year. After 2006, however, the United

States returned to a state of steady growth. In 2008, 0.39 GW were shipped, a 64% increase from

the 0.24 GW in 2007. In 2008, the United States was the fifth-greatest contributor (by

region/country) to PV shipments with a 7% market share.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Annual PV cell/module shipments (GW)

CdTe, CIGS, CIS

Amorphous Silicon

Ribbon Silicon

Monocrystalline Silicon

Polycrystalline Silicon

Figure 2.12. U.S. annual PV cell/module shipments 1997–2008

(Mints and Tomlinson 2007, Mints and Tomlinson 2008, Mints 2009)

Similar to PV production (described in Section 2.1.2), the two leading U.S. PV shippers in 2008

were thin-film manufacturers. First Solar was the leading U.S. contributor to 2008 shipments

with a 37% market share, or 145 MW, down from a market share of 46% (120 MW) in 2007 (see

Figures 2.13 and 2.14). United Solar Ovonics or Uni-Solar, an a-Si thin-film producer, had the

second-largest market share of U.S. PV shipments in 2008 with a 29% market share,

corresponding to 112 MW. Uni-Solar increased its shipments 130% from its 2007 level, resulting

in a market share increase from 20% in 2007 to 29% in 2008. Other significant contributors to

U.S. PV shipments in 2008 were: Evergreen (17%, 66 MW), BP Solar (9%, 35 MW), Schott

Solar (3%, 11 MW), and SolarWorld (3%, 10 MW). Evergreen jumped from the fifth- to the

third-largest contributor to U.S. shipments by increasing shipment levels 300% in 2008.

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Annual PV cell/module shipments (GW)

Figure 2.13. U.S. annual PV cell/module shipments by manufacturer 2004–2008

(Mints and Tomlinson 2008, Mints 2009)

Figure 2.14. Top U.S. companies for PV cell/module shipments 2008

(Mints 2009)

2.2.4 U.S. PV Cell/Module Revenue

U.S. revenue from PV cells/modules reached $1.2 billion in 2008, an increase of about 46% from

2007 revenue of $0.68 billion (Mints 2009). As shown in Figure 2.15, among U.S. companies in

2008, First Solar had the highest revenue at about $0.35 billion in PV cell/module sales, and

100

150

200

250

300

350

400

2004

2005

2006

2007

2008

Annual PV cell/module shipments (MW)

Other

Schott Solar

SolarWorld

BP Solar

Evergreen

United Solar Ovonics

First Solar

First Solar ,

145 MW, 37%

Uni-Solar ,

112 MW, 29%

BP Solar ,

35 MW, 9%

SolarWorld,

10 MW, 3%

Evergreen ,

66 MW, 17%

Schott Solar ,

11 MW, 3%

Other, 9 MW,

United Solar Ovonics (Uni-Solar) had the second highest at $0.33 billion. First Solar is also one

of three U.S.-based companies, including BP Solar and SolarWorld CA, with more than 20 years

of PV production experience. First Solar and BP also earned significant additional revenue from

non-cell or module sales, amounting to an additional $0.20 and $0.09 billion, respectively, in

2008. Four out of seven of the top U.S. PV producers are publicly owned, including First Solar,

Uni-Solar, SolarWorld, and Evergreen Solar. BP, Schott, and Global are all privately owned

solar PV production companies.

Figure 2.15. Top U.S. companies for PV cell/module revenues 2008

(Mints 2009)

2.2.5 U.S. PV Imports and Exports

Figure 2.16 presents data on U.S. PV cell/module imports and exports, including the crystalline

silicon and thin-film contributions. From 1999 to 2004, U.S. PV cell/module exports exceeded

imports significantly. This changed in 2005, when exports and imports nearly evened out,

resulting from increased U.S. imports of crystalline silicon modules and cells and somewhat

offset by increased U.S. exports of thin-film modules/cells. In 2006 and 2007, U.S. total PV

cell/module imports exceeded exports for the first time. Although U.S. thin-film exports doubled

each year from 2005 to 2007, dominating U.S. exports in 2007, U.S. crystalline silicon

module/cell imports grew to more than double the exports in 2006 and 2007. The large increase

in U.S. crystalline silicon module/cell imports was due to a rapid increase in demand for PV

modules and cells, which was particularly high in 2006 in response to the federal investment tax

credit for PV systems included in the Energy Policy Act of 2005. U.S. crystalline PV exports

have remained fairly flat, with increases in exports accounted for by the fast-growing, thin-film

industry.

First Solar,

$0.35 B, 28%

United Solar

Ovonics,

$0.33 B, 26%

Evergreen

Solar,

$0.32 B, 25%

BP Solar,

$0.14 B, 11%

Schott Solar,

$0.039 B, 3%

SolarWorld CA

(Shell Solar),

$0.043 B, 4%

Other,

$0.036 B, 3%

Figure 2.16. U.S. PV cell/module shipments, exports and imports

(EIA 2008a)

In 2007, 88% of U.S. PV exports were destined for Europe and 9% went to Asia (Figure 2.17).

The dominance of the European market is due primarily to significant government incentives in

countries such as Germany and Spain, which received about 64% and 13% of U.S. shipments,

respectively.

Figure 2.17. U.S. exports of PV cells and modules, 2007 destination

(EIA 2008a)

1999 2000 2001 2002 2003 2004 2005 2006 2007

0.00

0.05

0.10

0.15

0.20

0.25

Exports Thin Film

Imports Thin Film

Exports Crystalline Si

Imports Crystalline Si

Africa,

0.4%

Asia, 9.4%

Europe,

88%

North &

Central

America,

0.7%

Oceania &

Australia,

1.2%

South

America,

0.8%

2.3 CSP Manufacturer and Shipment Trends

Reflectors, receivers, and turbines are the three major components in CSP technologies that are

currently being installed worldwide.

2.3.1 CSP Manufacturers

Table 2.1 lists major CSP component (reflector, receiver, and turbine) manufacturers.

Historically, Flabeg has been the major manufacturer of reflectors, providing a product with 95%

or better reflectivity. Other emerging companies, such as PPG Industries and Rioglass, aim to

lower capital costs of glass and increase durability. Furthermore, other companies (e.g.,

Reflectech, 3M, and Alanod) are offering a reflective film as a lower-cost alternative to

traditional mirrors (Grama et al. 2008). In terms of receivers, Solel has historically been the

major manufacturer, but Schott Solar Systems recently has become a significant player. For

turbines, ABB, GE-Thermodyn, and Siemens are major manufacturers, and companies such as

Alstom, MAN Turbo, and ORMAT are looking to gain market share.

Table 2.1. CSP Component Manufacturers

Reflectors

Receivers

Turbines

• 3M

• Alanod

• Flabeg

• PPG Industries

• Reflectech

•

Rioglass

• Schott Solar Systems

• Solel

• ABB

• Alstom

• GE-Thermodyn

• MAN Turbo

• ORMAT

•

Siemens

Emerging Energy Research 2007, company Web sites 2008

2.3.2 CSP Shipments

Annual U.S. shipments of CSP dish and trough collectors remained relatively constant from

1999 to 2003, as shown in Table 2.2. A noticeable increase occurred in 2005, followed by a

substantial increase in 2006. The significant increase in 2006 was primarily the result of a

64-MW CSP plant in Nevada. The facility, Nevada Solar One, consists of 760 parabolic

reflectors comprising nearly 219,000 individual mirrors. It was the world’s largest plant built in

16 years (EIA 2008b). In 2007, shipments dropped dramatically, from 3,852,000 to 33,000 ft

Table 2.2. Annual U.S. Shipments: Parabolic Dish and Trough, 1998–2007

Year

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

Shipments

(1000 ft

)

21 4 5 2 2 7 — 115 3,852 33

EIA 2008b

2.4 Material and Supply Chain Issues

2.4.1 Polysilicon Supply for PV

This section presents information on polysilicon manufacturing and its importance to the PV

industry, the historical and 2008 polysilicon market, and polysilicon forecasts and trends.

About 84% of PV produced in 2008 used crystalline silicon semiconductor material derived from

polysilicon feedstock (Bartlett et al. 2009).

Polysilicon is silicon purified for use in making

semiconductors. Solar-grade polysilicon is silicon refined to be at least 99.999999% ("six nines"

or "6N") pure (Winegarner & Johnson 2006).

The polysilicon supply and demand imbalance

that became recognized widely around 2005 was caused not by a lack of silicon (silicon is

abundant and ubiquitous in Earth's crust) but by a lack of capacity for purifying silicon to the 6N

level.

Producing solar-grade polysilicon is complex and capital intensive. Quartz is heated in the

presence of a carbon source to produce liquid silicon, which is refined and allowed to solidify to

become what is known as metallurgical-grade silicon (MG-Si), with an average purity of 98.5%

(Bradford 2008). MG-Si is a relatively abundant and inexpensive commodity worldwide.

However, it must be processed further to achieve solar-grade purity using one of several

processes, of which three were most important in 2008:

• Siemens process (chemical deposition)

• Fluidized bed reactor (FBR) process (resulting in granular silicon)

• Upgraded MG-Si (UMG-Si) processes.

The Siemens process is the most widely used, followed by FBR. Siemens and FBR facilities are

capital intensive (typically $100 million or more per 1000 MT/year of capacity) and require 2–3

years to construct (Rogol et al. 2006, Winegarner & Johnson 2006). UMG-Si processes, which

enhance the purity of MG-Si, promise substantial cost and time savings over the Siemens process

and FBR, but the resulting product is of lower purity (i.e., below 6N) and must be blended with

purer polysilicon for PV applications (Bradford 2008). Maintaining polysilicon quality is critical.

Even small decreases in PV efficiency resulting from using lower-quality polysilicon can offset

the savings gained from using the lower-quality polysilicon (Rogol et al. 2006). A variety of

other polysilicon production processes (e.g., vapor liquid deposition) promise potential cost and

production-rate advantages if they can attain commercial performance goals (Bradford 2008).

Another source of solar-grade polysilicon is the electronics industry. Polysilicon used in the

electronics industry must be even purer (7N) than solar-grade polysilicon, and about 10%–20%

Mehta and Bradford 2009, Citi Investment Research 2009, Cowen & Co. 2009, Deutsche Bank 2009, Morgan

Stanley 2009, Navigant 2009, New Energy Finance 2009, Thomas Weisel 2009, as obtained from Bartlett et al.

2009.

The chemical definition of purity used here counts the numbers after the decimal point, i.e., 99.999999% is called

"six nines" or "6N" using the chemical definition. The metallurgical definition of purity, used by some in the solar

industry, counts the numbers before the decimal point. In other words, 99.999999% is called "eight nines" or "8N"

using the metallurgical definition. See the errata in Winegarner & Johnson 2006.

of the off-specification and scrap polysilicon sold to the electronics industry eventually becomes

available to the solar industry (Bradford 2008, Winegarner & Johnson 2006).

Beginning around 2004, an imbalance between polysilicon supply and demand contributed to

increasing polysilicon prices. For years, the PV industry had subsisted largely on leftover

polysilicon from the electronics industry. However, polysilicon demand for PV surpassed

demand for electronics in 2007 and has become the primary driver of growth in polysilicon

production (Bartlett et al. 2009). Polysilicon production facilities, with their high capital costs

and long construction times, were unable to respond immediately to the PV-driven spike in

polysilicon demand. This resulted in a supply/demand imbalance and a more than doubling of

the average polysilicon contract price between 2003 and 2007 (Bradford 2008, Mehta and

Bradford 2009).

The increase in polysilicon prices prompted a dramatic increase in new polysilicon producers

and significant investments in new production capacity and new technologies (including

UMG-Si). In 2008, the additional polysilicon production capacity initiated circa 2005 to satisfy

the unmet demand began production after 24–30 months of construction (Bradford et al. 2008c).

By mid-2008, the tightness in polysilicon supply began to ease, and PV cell and panel

manufacturers reported that polysilicon suppliers were more willing to sign long-term contracts

with new partners (Bradford et al. 2008c, Bradford et al. 2008b).

The median estimate from several industry analysts was 64,000 MT of polysilicon production in

2008, and the median 2008 estimate of polysilicon produced for the solar industry was

45,000 MT (10%–20% of scrap polysilicon from the electronics industry also will become

available to the solar industry) (Bartlett et al. 2009)

. This represents an estimated 54% increase

in total polysilicon production compared with 2007. At an average silicon utilization rate of

8.7 g/W, 45,000 MT is equivalent to 5.2 GW of polysilicon-based PV cell production (Bradford

2008).

Most of the polysilicon supply is sold under contract, with only a small proportion available on

the spot market; some PV manufacturers pay the higher spot market prices because they cannot

secure long-term contracts (e.g., owing to upfront cash requirements) or because additional

capacity requirements cannot be met by their contracted polysilicon supply (Bradford 2008).

Polysilicon spot prices topped $450/kg in early 2008, but they dropped to less than $150/kg by

early 2009 and less than $100/kg by mid-2009 (Wu and Chase 2009). One estimate of 2008

average polysilicon contract price was $70/kg (Mehta and Bradford 2009). Generally, spot and

contract prices are controlled by supply and demand; prices rise when demand exceeds supply

and vice versa. However, at least one analyst interpreted the high spot prices in 2008 as a false

market signal, suggesting that they actually heralded new polysilicon supply coming onto the

market at lower, contract prices (Bradford et al. 2008b). The reasoning is that marginal PV

manufacturers, who previously had little access to contract polysilicon, began securing

polysilicon contracts in 2008. As an increasing proportion of their needs were met with this

lower-priced contract polysilicon, they were able to pay higher prices for their decreasing spot-

Barclays 2009, Mehta and Bradford 2009, Cowen & Co. 2009, Morgan Stanley 2009, Oppenheimer 2009, Thomas

Weisel 2009, as obtained from Bartlett et al. 2009.

market requirements while maintaining the same blended average polysilicon cost (i.e., the

overall cost of contract polysilicon blended with spot polysilicon).

The Siemens process accounted for about 78% and FBR for about 16% of polysilicon produced

in 2008, with UMG-Si processes accounting for most of the rest (Barclays 2009, Mehta and

Bradford 2009, Cowen & Co. 2009, Morgan Stanley 2009, Bradford et al. 2008c). Figure 2.18

shows one analyst's estimate of 2008 production categorized by producer, and Figure 2.19 shows

it categorized by producing country (Bradford 2008).

Figure 2.18. Proportion of 2008 polysilicon production categorized by producer

(Bradford 2008)

Figure 2.19. Proportion of 2008 polysilicon production categorized by producing country

(Bradford 2008)

U.S. production includes a company with production in both the United States and Italy.

Hemlock (U.S.)

20%

Wacker

(Germany)

22%

Tokuyama

(Japan)

MEMC

(U.S./Italy)

12%

REC (U.S.)

11%

Mitsubishi

(Japan)

Sumitomo

(Japan)

New Entrants -

Current

Technology

(Siemens/FBR)

11%

New Entrants -

Alternative

Technology

(e.g., UMG-Si)

United

States

43%

Germany

22%

Japan

18%

South

Korea

China

Russia

Norway

Brazil

Canada

The global economic crisis that became apparent in late 2008, and which is expected to diminish

funds available for investments and the demand for PV, caused analysts to reduce their

projections for the growth of PV and polysilicon production (Bartlett et al. 2009). Nevertheless, a

number of analysts who released reports in early 2009 projected 157,000–203,000 MT of

polysilicon production by 2012, a two- to threefold increase over 2008 (Figure 2.20). Projecting

beyond 2010 is more difficult than making nearer-term projections, because decisions to build

new capacity need only be made 2 years in advance and thus would not need to be made by late

2008 or early 2009 (Bradford et al. 2008c). Solar applications were projected to use 80%–90% of

polysilicon supply by 2012 (Mehta and Bradford 2009, Cowen & Co. 2009).

Figure 2.20. Polysilicon supply projections through 2012

(Bartlett et al. 2009)

The projected increase in polysilicon supply is expected to meet or exceed PV demand for the

next several years. Several major trends will influence polysilicon supply and demand during this

period, with potential to increase supply or decrease demand beyond current projections:

• Improved silicon utilization. The PV industry has continued to improve silicon

utilization (watts of PV per gram of polysilicon) by decreasing silicon wasted

during manufacturing, producing thinner wafers, and producing cells with higher

efficiency. One analyst projects an average 14% improvement in polysilicon

utilization from 8.7 g/W in 2008 to 7.5 g/W in 2012 (Bradford 2008). Increasing

silicon utilization decreases the amount of polysilicon required to manufacture an

equivalent amount of crystalline silicon PV.

Projections through 2010 reflect estimates made by six analysts (Barclays 2009, Mehta and Bradford 2009,

Cowen & Co. 2009, Morgan Stanley 2009, Oppenheimer 2009, Thomas Weisel 2009), whereas projections for 2011

include data from four analysts (Barclays 2009, Mehta and Bradford 2009, Cowen & Co. 2009, Oppenheimer 2009),

and data for 2012 are from three analysts (Barclays 2009, Mehta and Bradford 2009, Cowen & Co. 2009), obtained

from Bartlett et al. 2009.

120

150

180

210

2006

2007

2008

2009

2010

2011

2012

Total Polysilicon Production (1,000 MT)

Low

Median

High

• Increased use of UMG-Si. As discussed above, UMG-Si processes could offer

substantial cost and time savings over the Siemens process and FBR. Significant

venture-capital UMG-Si investments in 2008 indicated continued interest in this

approach, but concerns remain about the efficiency of PV cells made with a high

percentage of UMG-Si (Bradford et al. 2008c). If technology is developed to

enable the use of UMG-Si in blends higher than the current level of 10%,

UMG-Si could become a larger source of low-cost PV feedstock. Estimates from

several industry analysts project the market share of UMG-Si rising from about

6% in 2008 to 8%–16% in 2012 (Bartlett et al. 2009).

• Market penetration of thin-film PV. Thin-film PV, which requires little or no

polysilicon feedstock, has become a major competitor to crystalline silicon PV.

Projections of market share rise from about 16% in 2008 to 16%–34% in 2012

(Bartlett et al. 2009).

Increased use of thin-film PV reduces demand for

crystalline silicon PV and polysilicon feedstock.

Other factors that could affect the supply of or demand for polysilicon include larger-than-

expected polysilicon production by companies based in China, technological breakthroughs (e.g.,

rapid penetration of concentrating PV), manufacturing disruptions (e.g., an accident in a very

large polysilicon manufacturing facility), PV supply-chain disruptions (e.g., shortages of solar-

grade graphite or glass), changes in PV-related government policies, and other "macro shocks"

(e.g., large-scale natural disasters or epidemics) (Rogol et al. 2008).

As a result of the aforementioned trends toward increased polysilicon supply and/or moderated

demand, polysilicon prices are expected to decrease over the next several years (Figure 2.21). PV

cell manufacturers may be able to minimize/stabilize the price they pay for polysilicon by

measures such as securing fixed-price contracts, producing their own polysilicon, or partnering

with polysilicon producers.

Barclays 2009, Mehta and Bradford 2009, Cowen & Co. 2009, Morgan Stanley 2009, as obtained from Bartlett et

al. 2009.

Mehta and Bradford 2009, Citi Investment Research 2009, Cowen & Co. 2009, Deutsche Bank 2009, Morgan

Stanley 2009, Navigant 2009, New Energy Finance 2009, Thomas Weisel 2009, as obtained from Bartlett et al.

2009.

Figure 2.21. Polysilicon price projections through 2015

(Mehta and Bradford 2009)

Polysilicon price and silicon utilization have a substantial effect on PV module manufacturing

cost. Figure 2.22 plots module manufacturing cost (for a European-produced, multicrystalline

silicon module manufactured in 2008) versus blended polysilicon price (a weighted average of

contract and spot price) for three levels of silicon utilization. This figure illustrates the

importance to PV module economics of lowering polysilicon costs and increasing silicon

utilization (Mehta and Bradford 2009).

Figure 2.22. PV module sensitivity to polysilicon price and utilization

(Mehta and Bradford 2009)

$50

$100

$150

$200

$250

$300

$350

2008

2009

2010

2011

2012

2013

2014

2015

Projected Polysilicon Price ($/kg)

Contract Price

Spot Price

$1.80

$2.00

$2.20

$2.40

$2.60

$2.80

$3.00

$3.20

$40

$50

$60

$70

$80

$90

$100

$110

$120

$130

$140

$150

Module manufacturing cost ($/W)

Blended polysilicon price ($/kg)

5 g/W

7 g/W

9 g/W

2.4.2 Rare Metals Supply and Demand for PV

As discussed above, thin-film PV technologies, which use 100 times less silicon than

conventional crystalline cells (e.g., for a-Si PV) or use no silicon at all (e.g., for CdTe, CIGS,

and CIS PV), are projected to garner 16%–34% of the PV market by 2012. This large-scale

production has raised concerns about the supply of rare metals such as indium, gallium, and

tellurium (Grama & Bradford 2008). Indium and tellurium have been estimated to limit total

thin-film capacity to between 100 GW for CdTe (Feltrin & Freundlich 2008) and 30 TW for

CIGS/CIS (Zweibel 2005) based on different estimates of rare metal supply.

The estimated worldwide indium reserve base

was 16,000 MT in 2007, and annual production

was 510 MT (Grama & Bradford 2008). CIGS PV requires approximately 20 MT of indium per

GW. With projected production growth to 3.1 GW by 2012, CIGS PV will require approximately

63 MT of indium (12% of current annual production) by that year (Grama & Bradford 2008).

Competing uses for indium include LCD displays, integrated circuits, and electronic devices.

The market price for indium reached $835/kg in early 2007, but fell to $750/kg in August 2008

(Grama and Bradford 2008). Indium recycling could increase significantly, and alternative

sources could be developed, if the indium price stays high. Indium has substitutes, but they

usually lead to losses in production efficiency or product characteristics. Gallium can be used in

some applications as a substitute for indium in several alloys. In glass-coating applications,

silver-zinc oxides or tin oxides can be used. Indium phosphide can be substituted by gallium

arsenide in solar cells and in many semiconductor applications. The United States has no primary

indium or gallium production capacity, and most reserves are located in other countries.

The estimated worldwide tellurium reserve base was as high as 47,000 MT in 2007, and annual

production was 475 MT. CdTe PV requires approximately 47 MT of tellurium per GW. With

projected production growth to 1.2 GW by 2012, CdTe PV will require approximately 55 MT of

tellurium (12% of current annual production) by that year. Competing uses for tellurium include

semiconductor and electronics products and metal alloys. The tellurium required for CdTe PV

has driven a fourfold price increase in the past 5 years; the market price increased from $50/kg in

2004 to more than $225/kg in August 2008. Tellurium recycling could increase significantly, and

alternative sources could be developed (e.g., bismuth telluride), if the tellurium price stays high.

Beyond 2012, tellurium production likely will need to increase to keep pace with demand from

the solar industry (Grama and Bradford 2008).

2.4.3 Glass Supply for PV

Glass is resistant to long-term weathering, is relatively inexpensive, and has good mechanical

strength, making it an ideal encapsulation material for crystalline silicon PV as well as an

encapsulation and substrate material for most thin-film PV. The demand for solar glass is

expected to see strong growth in conjunction with growing PV demand. Yole Development

forecasts that solar glass demand will grow from 50 million m

today to more than 300 million

in 2015. Mark Farber, former CEO and cofounder of Evergreen Solar, notes that PV currently

The U.S. Geological Survey defines base reserves as that part of an identified resource that meets minimum

physical and chemical criteria related to current production practices (Grama & Bradford 2008).

accounts for less than 1% of the glass market, but that it could account for up to 5% by 2012

(Podewils 2008).

The glass demand for PV (as well as for CSP) is primarily for high-quality, low-iron glass. This

demand is met by both the rolled- and float-glass markets. Rolled glass is a better choice for

solar applications because it requires 80% less energy to manufacture than float glass, it can use

resources with 30% more iron to achieve the same transparency, and it is 30% less expensive to

produce. Rolled glass currently supplies 60%–70% of the PV market, and production is growing

rapidly to meet PV demand, particularly in China. However, demand for rolled glass exceeds

supply by about 1 year. Float glass, which accounts for more than 90% of the flat glass market,

will likely supply the additional PV glass demand until the rolled glass market can catch up

(Podewils 2008).

There is an increasing trend, particularly in China, toward constructing PV-dedicated glass

factories and vertically integrating glass production in the PV supply chain by building rolled-

glass factories on site, which minimizes supply-driven price volatility and transportation costs.

China's CSG Solar Glass Co. built the first factory to produce glass exclusively for the solar

industry (Podewils 2008). Once operating at full capacity, CSG could produce enough glass for

2 GW/year of PV. CSG also is vertically integrating, from polysilicon production to finished

modules, and plans to have 450 MW of cell and module production by 2010.

2.4.4 Material and Water Constraints for CSP

Concentrating solar power facilities are constructed primarily of concrete, steel, and glass.

Although these materials are not subject to rigid supply limits, they are affected by changes in

commodity prices.

In CSP systems, steel is a commodity that cannot be substituted, and its cost rose substantially

from 2006 through the first half of 2008. These near-term price increases inflated the levelized

cost of energy (LCOE) that project developers could offer to utilities. However, the economic

downturn in the second half of 2008 brought commodities closer to pre-2006 prices (Yahoo

Finance 2009). For CSP projects to be economically viable, it will be important for steel prices to

remain at reasonable costs, as project developers do not have pricing power on this commodity

and must therefore accept the price being offered (Bullard et al 2008).

The combination of a limited number of companies producing the main CSP system components

on a commercial scale and a large construction pipeline could create a component supply

bottleneck, depending on the growth of demand. This is especially true for turbines, which

require a 24–30 month advance order by the project developer (Merrill Lynch 2008).

Conversely, several of the main materials and equipment relied on by the CSP industry could

leverage other manufacturing industries (e.g., automotive and buildings), potentially facilitating a

relatively fast production ramp-up (Andraka 2008). In addition, some companies are entering

multiple levels of the value chain, thus becoming less dependent on other companies for meeting

large supply requests (e.g., Solel and Abengoa are entering the reflector market). For these

reasons, many believe that reflector and receiver demand will not create bottlenecks in the near

future (Bullard et al. 2008, Merrill Lynch 2008).

Aside from the aforementioned solid materials, the availability of water resources might limit the

amount of CSP deployed in arid regions. As with fossil and nuclear power plants, the most

common and economical method for cooling a CSP plant is evaporative water cooling. A typical

parabolic trough plant requires 800 gallons of water per MWh, which can be reduced to about

80 gallons per MWh by implementing dry cooling. Dry cooling, however, requires higher capital

expenditure and is likely to decrease plant efficiency, especially under higher-temperature

conditions. The loss of efficiency is greatest for systems requiring lower operating temperatures.

Another cooling solution is to use hybrid wet/dry cooling, which uses less water than 100% wet

cooling, and also reduces efficiency losses associated with dry cooling. For closed-cycle CSP

heat engines such as dish-engine generators, air cooling is sufficient (U.S. DOE 2009).

2.4.5 Land and Transmission Constraints for Utility-Scale Solar

Solar project developers have been attracted to numerous locations in the Southwest, evident

from a surge of more than 220 applications received by the U.S. Bureau of Land Management

for new solar power plants (Resseguie 2008). However, transmission in many of these areas is

lacking and substantial upgrades to the western grid will be necessary for projects to move

forward. Moreover, land-use conflicts exist, as a large percentage of the area is federal land

traditionally set aside for conservation and recreational purposes. To address the transmission,

grid upgrade, and land-use issues, a number of major multi-agency agreements and initiatives

have arisen at the national, state, and regional levels. Five such agreements and initiatives are

described in this section.

The U.S. Bureau of Land Management (BLM) and DOE are collaborating on a Solar

Programmatic Environmental Impact Statement (PEIS). The PEIS will identify the impacts of,

and develop better management strategies for, utility-scale solar development on the public lands

of six states (Arizona, California, Colorado, New Mexico, Nevada, and Utah). The BLM has

announced the availability of maps that identify 24 tracts of BLM-administered land for in-depth

study for solar development and requested public comment on those study areas; the schedule for

release of the draft PEIS will be determined after the evaluation of those comments. Further

details can be found at http://solareis.anl.gov

, and solar energy study area maps are available at

http://solareis.anl.gov/eis/maps/index.cfm.

The Western Governors’ Association (WGA) and DOE launched the Western Renewable Energy

Zones (WREZ) Project in May 2008. WREZ involves working groups that are identifying high-

resource areas to include in energy development and environmentally sensitive lands to exclude

from this development. The assessment was presented to the WGA for approval in June 2009,

concluding phase one of the process. Subsequent phases will develop a modeling tool and a high-

level transmission plan coordinated through the Western Electricity Coordinating Council. More

information is available at http://www.westgov.org/.

Most recently, a multi-agency memorandum of understanding (MOU) was signed in October

2009 regarding coordination in federal agency review of electric transmission facilities on

federal land. The MOU was signed by the U.S. Department of the Interior, Department of

Agriculture, Department of Commerce, Department of Defense, Department of Energy,

Environmental Protection Agency, Council on Environmental Quality, Federal Energy

Regulatory Commission (FERC), and the Advisory Council on Historic Preservation. The

agreement supersedes an earlier MOU signed in August 2006. The October 2009 MOU will cut

approval time off the normal federal permit process and help break down the barriers to siting

new transmission lines by: 1) designating a single federal point-of-contact for all federal

authorizations; 2) facilitating coordination and unified environmental documentation among

project applicants, federal agencies, states, and tribes involved in the siting and permitting

process; 3) establishing clear timelines for agency review and coordination; and 4) establishing a

single consolidated environmental review and administrative record (U.S. DOI 2009).

In California, the Renewable Energy Transmission Initiative (RETI) is a collaboration of public

and private entities whose objective is to provide information to policymakers and stakeholders

on the transmission requirements to access cost-effective, environmentally sensitive, renewable

energy resources. Phase one of the initiative, completed at the start of 2009, identified and

ranked zones in California and nearby states that can provide and competitively deliver

renewable energy to the state. Phase two, which includes developing conceptual transmission

plans and refining previous work, is under way. Additional information is available at

www.energy.ca.gov/reti.

Another development in California was an MOU signed in August 2007 between the U.S.

Bureau of Land Management (BLM) (part of the U.S. Department of the Interior) and the

California Energy Commission (CEC) to document the roles, responsibilities, and procedures to

follow in conducting a joint environmental review of proposed solar thermal power plant

projects in the state of California. A number of these projects are proposed to be built in

California on land owned by the federal government and managed by BLM. Because projects

need both a right-of-way from BLM and certification from CEC, coordination in the preparation

of an environmental analysis for each of these projects will avoid duplication of efforts, promote

intergovernmental coordination at the federal, state, and local levels, and facilitate public review

by providing a joint document and a more efficient review process (CEC 2007a).

The long approval process for a recent transmission project in California provides an example of

a project that could have benefited from the type of coordination being carried out through the

above agreements and initiatives. After 3 years and several major revisions, the 1-GW, 123-mile

Sunrise Powerlink Transmission Project was approved by the California Public Utilities

Commission in December 2008. This project allows San Diego Gas & Electric to connect

producers in the Imperial Valley of California to end users in the San Diego area, and is expected

to be completed in 2012. The process for obtaining the permit was arduous because of the initial

lack of disclosure by the proponent, the environmental sensitivity of the land proposed to be

developed, and the number of stakeholders involved (Herndon 2009). Many PV and CSP

projects in the planning stages are dependent on the construction of this line.

2.5 Solar Industry Employment Trends

Because of the rapid expansion of the solar industry over the past several years, employers are

reporting increasing difficulty in finding qualified people to hire. As a result, labor supply and

demand within the solar industry, and the manner in which the two are tied to educational and

training opportunities, has recently become a topic of strong interest to both industry and

government. The possibility of labor shortages within the industry and a surge of interest in

green jobs as a contributor to economic recovery are two factors driving state and federal

initiatives. If Germany is any indication of how fast the U.S. solar industry can grow without

experiencing major bottlenecks in labor supply, then it is possible that the U.S. educational and

training infrastructure could meet the needs of the industry over the next several years.

Despite the poor investment climate that is clouding economic growth in the United States and

abroad, the extension and enhancement of the federal investment tax credit (ITC) as of October

2008 and additional policy enhancements and economic incentives resulting from the American

Recovery and Reinvestment Act (ARRA) in February 2009 are expected to encourage industry

expansion and job creation. In particular, the elimination of the residential investment tax credit

cap (effective January 2009) should result in more business for PV system installers. Also, the

new ability of utilities to take advantage of the ITC could result in very large installations being

constructed and could help lead to the establishment of well-defined career pathways for PV

system designers and solar electricians. For more details on recent federal policy developments

that may encourage job growth in the solar industry, see Section 4.1 of this report.

Though not the focus of this section, studies have quantified the job creation potential of

renewable energy as compared to fossil fuel technologies. A recent study by analysts at U.C.

Berkeley concluded that renewable energy technologies generate more jobs per unit of energy

than fossil fuel-based technologies. Among the renewable energy technologies, solar PV creates

the most jobs per unit of electricity output. Solar PV was estimated to create 0.87 job-

years/GWh, whereas natural gas and coal were each estimated to create 0.11 job-years/GWh.

Solar PV thus generates almost eight times as many job-years/GWh as natural gas or coal (Wei

et al. 2010).

2.5.1 Types of Jobs in the PV and CSP Industries

The following are examples of occupations associated with the manufacture and installation of

PV and CSP.

• Manufacturing positions include factory worker, sheet metal worker, glass worker,

technician (e.g., semiconductor for PV), material handler, factory supervisor,

manufacturing manager, engineer (quality assurance, manufacturing, chemical process,

mechanical, electrical, optical), material scientist.

• Installation positions include solar system installer/technician (PV), solar system designer

(PV), technical sales representative and estimator (PV), architect (PV), roofing contractor

(PV), general contractor, supervisor/foreman, heavy construction worker, welder,

pipefitter, engineer (mechanical, electrical, civil).

• Administrative and support positions include administrative assistant, purchasing agent,

accountant, health and safety officer, information technology professional, director.

Jobs in the solar industry fall into three categories: direct, indirect, or induced. Direct jobs are

those within the solar industry (e.g., manufacturing, installation, R&D); indirect jobs are those in

industries that support the solar industry (e.g., jobs in the polysilicon, glass, and steel industries);

and induced jobs are those that result from the economic activity stimulated by the solar industry

See http://rael.berkeley.edu/greenjobs for more details.

(e.g., people buying more goods and services in a region where there is a new PV manufacturing

plant or where a new PV or CSP installation is under construction).

Estimating numbers of jobs is a challenging task that is generally performed by surveying

employers and industry analysts or by using economic input-output models that account for jobs

based on the trail of goods and services leading up to, and after, production. There is significant

variation in job number and labor intensity estimates, which results from many factors,

including:

• Data collection and analysis method

• Types of jobs being considered (e.g., direct, indirect, and induced)

• Types of occupations being considered (e.g., factory worker, installer, salesperson).

• Variation in estimates of capacity being installed (for job forecasts)

• Technologies included (e.g., PV, CSP, solar water heating)

• Types of industry subsectors included (residential new and retrofit, commercial, utility,

remote or off-grid)

• Variation in metrics or units being used

• Variation in the time periods being considered, such as the lifespan of an installation or

within only a certain phase (e.g., construction, O&M)

• Whether a study is measuring gross or net job impacts (net impacts account for

displacement of jobs in other industries such as coal or natural gas).

In this and many reports, the unit of measure used for jobs is the job-year or full-time equivalent

(FTE), which represents full-time employment for one person for the duration of a year. A

common metric for labor intensity is jobs/MW (or FTEs/MW), although other metrics such as

jobs per MWh or jobs per dollar invested are also used depending on the question of interest.

2.5.2 Current and Projected Employment in the Solar Industry, Global and U.S.

In 2008, an estimated 173,000 people

were employed worldwide in the solar electric industry,

according to a 2009 New Energy Finance (NEF) study (McCrone et al. 2009). Most of these jobs

were in the PV manufacturing value chain and in the construction and installation of PV

components and projects.

Of the 173,000 jobs, the PV industry contributed about 169,000 and

the CSP industry about 4,000. The 169,000 PV jobs correspond to 5.8 GW produced and

installed globally in 2008 and 14.7 GW of cumulative installed PV capacity through 2008.

An earlier report produced by the United Nations Environment Programme (UNEP) in 2008

compiled an estimate of global PV employment based on five industry leaders: Germany, Spain,

China, Japan, and the United States. PV jobs in these countries together amounted to

approximately 170,000 jobs

These are direct and indirect jobs, in FTEs. The figure includes jobs in PV and CSP for electricity generation, not

solar hot water heating jobs. Also not included are induced jobs.

worldwide in 2007 (Renner et al. 2008). The 170,000 job number

All job and labor intensity figures quoted in this section from the McCrone et al., New Energy Finance article of

2009 refer to FTEs.

The 14.7 GW of cumulative installed PV capacity through 2008 is higher than the 13.9-GW estimate provided in

Section 1.1.1 of this report. Estimates of cumulative installed PV capacity through 2008 range from 13-17 GW.

The figure of 170,000 total jobs worldwide provided in the UNEP report includes direct and indirect jobs.

for 2007 is similar to the NEF number of 169,000 PV jobs in 2008 (though the estimates are one

year apart). However, as previously stated, it is not unusual for job estimates to vary based on

many factors, including differences in methodology and metrics, which jobs in the value chain

are being counted, and in this case, for which year and which countries the data are being

provided.

Of the 170,000 global PV jobs reported by UNEP for the year 2007, China held the most jobs,

with 70% of its 55,000 jobs going to the manufacture of PV cells, modules, and processed

silicon. Germany

Estimates of U.S. PV jobs for the year 2007 range from 6,000 to 20,000 jobs. EIA reports

employment in the PV manufacturing industry to be about 6,200 jobs in 2007, based on a survey

of PV cell/module manufacturers (EIA 2008a). The EIA number does not include all jobs along

the PV value chain and does not include indirect jobs. By comparison, for the year 2007, ASES

reported 8,700 direct jobs and 19,800 direct plus indirect jobs (ASES and MISI 2008). For the

year 2008, EIA’s survey yielded about 11,250 jobs in PV cell/module manufacturing (EIA

2009). If one used the EIA 2007 and 2008 numbers as an estimate for direct PV jobs, and used a

multiplier of 1.4 indirect jobs for every direct job, then direct plus indirect PV jobs for 2007 and

2008 would be about 14,800 and 27,000 jobs, respectively.

and Spain held approximately 35,000 and 26,000 jobs, respectively, in 2007.

Japan’s PV employment levels were estimated to be similar to Germany’s (Renner et al. 2008).

U.S. PV jobs reported by UNEP were 6,800 direct and 15,700 direct plus indirect jobs for the

year 2006, based on an estimate provided by ASES (Bezdek 2007). However, there is a large

range in the estimates of U.S. PV jobs provided by various studies.

Regarding projected employment in the United States, Navigant Consulting published an

employment impact study in September 2008 based on the assumption that the federal

investment tax credit (ITC) would be extended. A month later, the Emergency Economic

Stabilization Act (EESA) of 2008 extended and enhanced the ITC (beyond the assumptions

made in the Navigant study). The Navigant study estimated that 440,000 jobs

would be created

in the U.S. solar industry in 2016 if the ITC were extended for 8 years, which is 276,000 more

jobs than would have been created with only a reduced ITC (the residential ITC expiring and the

commercial ITC reduced to 10% at the end of 2008).

Of the 440,000 projected jobs in 2016,

110,000 were estimated to be direct, 130,000 indirect, and 200,000 induced (Navigant

Consulting 2008). The breakout by technology is 377,000 PV jobs, 38,000 CSP jobs, and 24,000

solar water heating jobs. The study methodology involved using economic multipliers to

calculate indirect and induced jobs from direct jobs; the multipliers are different for construction

and manufacturing versus O&M jobs (Navigant Consulting 2008, Grover 2007).

In Germany in 2008, total renewable energy jobs were about 278,000, with direct plus indirect solar industry jobs

at 74,400 or 27% of the total, and about 51,000 of solar jobs attributable to PV (the rest to solar thermal heating)

(Böhme et al. 2009). The increase from 35,000 jobs in 2007 to 51,000 in 2008 reflects 46% employment growth.

Job multipliers are described further in the subsequent discussion of the Navigant employment impact study.

Includes direct, indirect, and induced FTEs for PV, CSP, and solar water heating.

The 440,000 total jobs corresponds to nearly 6.5 GW of solar installations added in 2008 and 28 GW of

cumulative solar installations through 2008 in the extended ITC scenario. By comparison, the reduced ITC scenario

resulted in 9 GW of cumulative installed capacity through 2008, 19 GW less than the extended ITC scenario.

For construction and manufacturing, the ratio of indirect to direct is 1.4, and induced to direct is 2.1. For O&M,

the ratios are 0.5 for indirect to direct and 0.8 for induced to direct.

2.5.3 Labor Intensity in the PV Industry, Global and U.S.

Total employment in the solar industry is projected to increase over the next 8 years, but labor

intensity could decrease over time resulting from increased automation, economies of scale, and

greater efficiencies in the use of labor throughout the supply chain. Employment at the point of

installation is a significant component of labor intensity, thus benefiting local economies.

As with estimates of job numbers, labor intensity estimates for PV vary widely, with a range of

25 jobs/MW to more than 50 jobs/MW (including direct and indirect jobs). As discussed earlier,

the wide range is due to many factors including differences in methodology, which jobs along

the value chain are being counted, and the types of indirect jobs considered.

A recent study by New Energy Finance estimated global PV labor intensity in 2008 to be just

over 28 jobs/MW, including direct and indirect FTEs (McCrone et al. 2009).

The components

of the 28 jobs/MW along the value chain are presented in Table 2.3. PV operation/maintenance

(O&M), project construction, and rooftop installation accounted for 43% of PV jobs in 2008

(O&M by itself was 5%). Manufacture of silicon, wafers, cells, and modules accounted for 50%

of PV jobs in 2008. Adding induced jobs to the PV labor intensity of 28 jobs/MW using the

multipliers provided in section 2.5.2 (see footnotes) amounts to nearly 53 jobs/MW.

Table 2.3. Global PV Labor Intensity in 2008 (Direct and Indirect Jobs)

Job Category GW Jobs/MW Total Jobs

Operation

14.7

0.6

8,820

PV project construction and

rooftop installation

5.8

11 63,800

Silicon and wafers

5.8

3.5

20,300

Cell manufacture

5.8

29,000

Module manufacture

5.8

34,800

Inverters

5.8

1.3

7,540

Research

5.8

0.4

2,320

Development and services

5.8

0.4

2,320

Total

28.2

168,900

McCrone et al. 2009

The 28 jobs/MW worldwide labor-intensity number for 2008 is projected to decrease to about

13 jobs/MW in 2025 (McCrone et al. 2009). The labor-intensity estimate for 2025 is based on

A highly referenced labor-intensity figure for the U.S. PV industry has been 35.5 jobs/MW installed, which

originates from a study conducted by the Renewable Energy Policy Project (REPP) in 2001.

The study focused on

jobs along the value chain that are required to create a 2-kW residential PV system, and therefore does not factor in

commercial or utility-scale systems. The 35.5 job/MW figure includes mostly direct plus some indirect jobs.

Let x = direct jobs. Using the construction and manufacturing multipliers for the sake of simplicity, direct +

indirect jobs = x + 1.4x = 2.4x = 28.2. So x = 28.2/2.4 or 11.75 jobs. Total jobs = x + 1.4x + 2.1x = 4.5x =

4.5(11.75) = 52.875 or about 53 jobs. In summary: the multiplier for total to direct jobs is 4.5, the multiplier for total

jobs given direct + indirect jobs is 1.875, and the multiplier for induced jobs given direct + indirect jobs is 0.875.

The PV project construction and rooftop installation components were combined based on the data in the New

Energy Finance report for the purpose of presenting a global labor intensity for all PV system types.

39 GW of PV production and installation and 340 GW of cumulative installed PV capacity in

2025. The share of PV jobs in 2025 attributed to O&M increases to 29% (compared to 5% in

2008) based on the large cumulative installed PV capacity. PV project construction and rooftop

installation account for 33%, whereas manufacture of silicon and wafers, cell, and modules

account for 27% of PV jobs in 2025. Wafer, cell, and module manufacturing, system integration,

and residential installations are projected to have the most dramatic drops in labor intensity,

whereas commercial and utility installations will see only a slight decrease, one cause being that

many of the efficiencies in these areas have already been realized (Navigant Consulting 2008).

NREL discussions with several U.S. PV installation companies in 2008 also confirmed a pattern

of decreasing labor intensity. In addition, other insights from these conversations included the

following:

• Most installation companies do not have well-defined business groups targeting

residential, commercial, and utility markets.

• Many installation companies anticipate substantial growth in the retrofit residential

market.

• The retrofit market seems to need more skilled sales people as growth continues.

• Smaller companies are using skilled employees to do some unskilled work.

2.5.4 Employment and Labor Intensity in the U.S. and Global CSP Industry

As with PV installations, the construction phase of a CSP facility, as opposed to the operation

phase, results in the greatest economic impact. A report by NREL examining the economic

impacts of constructing a 100-MWe CSP facility in the United States, specifically Nevada,

estimated that each year of a 3-year construction period would result in slightly more than 800

direct jobs and approximately 1,600 indirect and induced jobs (Schwer and Riddel 2004).

This

equates to 8 direct jobs/MW and 24 jobs/MW including indirect and induced jobs, just for

construction. In addition, 0.45 jobs per MW are created directly during the operations and

maintenance phase.

Another U.S. example for CSP is the 400-MWe Ivanpah Solar Electric Generating System

(ISEGS) proposed for a site in California’s Mojave Desert, a power tower plant that would take

4 years to build and require approximately 500 jobs averaged over the construction period,

amounting to 1.25 jobs/MW during this time (CEC 2007b). The project also would require

100 full-time jobs for O&M, equivalent to 0.25 jobs per MW. ISEGS is a very large system, and

actually is a staged cluster of four separate CSP systems, which could explain the lower O&M

labor intensity.

By comparison, Black & Veatch estimated more than twice that number

for O&M, namely that every megawatt of CSP constructed results in 0.94 permanent O&M jobs

(Stoddard et al. 2006). It is unclear, however, whether the higher number includes indirect as

well as direct jobs.

The term “MWe” stands for “megawatt electrical,” and is used to distinguish electrical generation from thermal

generation at CSP plants.

Direct jobs are in FTEs. In total, 140 O&M jobs are created annually when including indirect and induced jobs.

Globally, CSP labor intensity in 2008 is estimated to be 23 jobs/MW (including direct and

indirect FTEs), which comprises the following components: operation, 0.8 jobs; project

construction, 12 jobs; manufacturing, 10 jobs; and development and services, 0.4 jobs (McCrone

et al. 2009).

2.5.5 Quality Assurance and Certification for Solar PV Installation

Proper installation of solar PV systems is essential for accelerating market acceptance and

maintaining consumer confidence. DOE SETP supports multi-tiered training and certification

processes for solar installation technicians by providing funding to the North American Board of

Certified Energy Practitioners (NABCEP), the only North American organization developing

and implementing consistent standards. Between years 2003 and 2008, NABCEP certified

587 PV installers

and awarded its Certificate of Basic Knowledge to 999 people.

2.5.6 DOE Response to Current Barriers in Workforce Development

The number

of PV installer certificates issued rose by 22% in 2008, and the number of Certificates of Basic

Knowledge awarded rose by more than 100%.

Research into workforce development tends to focus on estimating the number of jobs that will

result from enactment of regulatory policies such as state renewable portfolio standards (RPSs)

or financial incentives such as the ITC. Establishing baselines for the number of jobs within the

solar industry has proved to be difficult, and researchers often are hindered by the lack of

detailed data on occupations found in the solar industry. Based on a submission by DOE SETP,

however, the Office of Management and Budget (OMB) announced on January 21, 2009, that the

Standard Occupational Classification for 2010 will be revised to include the occupation of “Solar

Photovoltaic Installer” (47–2231), thus making it easier to track the number of PV technicians

and installers.

To better assess the education, training, and workforce development needs of the solar industry,

SETP activities in 2009 included:

• Commencing the development of the Solar Installer Instructor Training network, a multi-

year effort to provide funding to current educational providers to help train instructors at

other educational institutions that are starting training programs for the downstream PV

workforce.

• Exploring ways to provide support to activities that promote higher education in solar

technologies and upstream workforce development.

• Providing funding for research and analysis on knowledge gaps, including an

investigation of critical occupational profiles and estimates of employment within the

solar industry.

It is estimated that NABCEP-certified PV installers represent approximately 10% of the total number of PV

installers in the United States.

Professionals holding this certificate have demonstrated an understanding of the basic terms and operational

aspects of a PV system, but the certificate by itself does not qualify an individual to install PV systems. The

Certificate of Knowledge is held not only by persons interested in becoming installers, but also by industry

salespeople, contractors, local code officials, utility inspectors, and others.

• Through a memorandum of understanding signed in May 2009, partnering with the

Departments of Labor and Education in leveraging resources to provide green jobs

training and expertise.

• Collaborating with federal agencies and other organizations to provide more accessible

labor statistics.

2.6 References

The American Solar Energy Society (ASES) and Management Information Services, Inc. (MISI)

(2008). Defining, Estimating, and Forecasting the Renewable Energy and Energy Efficiency

Industries in the U.S. and in Colorado.

http://www.ases.org/index.php?option=com_content&view=article&id=465&Itemid=58.

Accessed December 2009.

Andraka, C.E. (July 2, 2008). Statement of Charles E. Andraka, Sandia National Laboratories.

Concentrating Solar Power. United States Senate Committee on Energy and Natural Resources

Field Hearing in Albuquerque, New Mexico.

Barclays. (2009). Solar Energy Handbook: The Second Growth Phase of Solar Era. Barclays

Capital.

Bartlett, J.E.; Margolis, R.M.; Jennings, C.E. (2009). The Effects of the Financial Crisis on

Photovoltaics: An Analysis of Changes in Market Forecasts from 2008 to 2009. National

Renewable Energy Laboratory. NREL Report no. TP-6A2-46713.

Bezdek, B. (2007). Renewable Energy and Energy Efficiency: Economic Drivers for the 21

Century. Boulder, CO: American Solar Energy Society.

http://www.ases.org/index.php?option=com_content&view=article&id=465&Itemid=58.

Accessed November 2009.

Böhme, D.; Dürrschmidt, W.; Van Mark, M., eds. (June 2009). Renewable Energy Sources in

Figures: National and international development. Berlin, Germany: Federal Ministry for the

Environment, Nature Conservation and Nuclear Safety (BMU).

http://www.erneuerbare-

energien.de/inhalt/5996/3860/. Accessed November 20, 2009.

Bradford, T. (2008). Polysilicon: Supply, Demand, and Implications for the PV Industry.

Prometheus Institute and Greentech Media.

Bradford, T.; Letendre, S.; Flynn, H.; Maycock, P. (March 2006). “22nd Annual Data

Collection.” PV News. Prometheus Institute and Greentech Media.

Bradford, T.; Englander, D.; Lewin, A.; Maycock, P. (2008a). March 2008. “24th Annual Data

Collection - Preliminary.” PV News. Prometheus Institute and Greentech Media.

Bradford, T.; Englander, D.; Lewin, A.; Maycock, P. (2008b). April 2008. PV News. “Letter

from the Editor.” Prometheus Institute and Greentech Media.

Bradford, T.; Englander, D.; Lewin, A.; Maycock, P. (2008c). May 2008. PV News. “Polysilicon

Forecasts through 2012.” Prometheus Institute and Greentech Media.

Bradford, T.; Englander, D.; Smith, F.; Maycock, P. (April 2009). “25th Annual Data Collection

Results: PV Production Explodes in 2008.” PV News. Prometheus Institute and Greentech

Media.

Bullard, N.; Chase. J.; d’Avack, F. (May 20, 2008). “The STEG Revolution Revisited.” Research

Note. London: New Energy Finance.

California Energy Commission (CEC). (2007a). August 2007. “Memorandum of Understanding

Between the U.S. Department of the Interior, Bureau of Land Management California Desert

District, and the California Energy Commission Staff Concerning Joint Environmental Review

for Solar Thermal Power Plant Projects.”

http://www.energy.ca.gov/siting/solar/BLM_CEC_MOU.PDF. Accessed November 23, 2009.

California Energy Commission (CEC). (2007b). October 12, 2007. Ivanpah Application for

Certification, Executive Summary.

http://www.energy.ca.gov/sitingcases/ivanpah/documents/applicant/AFC/Volume1/. Accessed

November 23, 2009.

Citi Investment Research. (2009). U.S. Solar Stocks: Inventory Has Yet to Peak, so It’s Still Too

Early. Citi Investment Research.

Cowen & Co. (2009). Industry Outlook: Industry Hoping for Help from Stimulus Bill. Cowen &

Co.

Deutsche Bank. (2009). Solar Photovoltaic Industry: Looking Through the Storm. Deutsche

Bank.

Emerging Energy Research. (2007). Global Concentrated Solar Power Markets and Strategies,

2007–2020. Cambridge, MA: Emerging Energy Research.

Feltrin, A.; Freundlich, A. (2008). “Material Considerations for Terawatt Level Deployment of

Photovoltaics.” Renewable Energy, 33, pp.180–185.

First Solar Inc. (2008). “Corporate Overview Q3 2008.” First Solar Web site.

http://investor.firstsolar.com/phoenix.zhtml?c=201491&p=irol-irhome. Accessed March 17,

2009.

Grama, S.; Bradford, T. (2008). Thin-Film PV 2.0: Market Outlook Through 2012. Prometheus

Institute and Greentech Media.

Grama, S.; Wayman, E.; Bradford, T. (2008) Concentrating Solar Power—Technology, Cost,

and Markets. 2008 Industry Report. Cambridge, MA: Prometheus Institute and Greentech

Media.

Grover, S. (2007). Energy, Economic, and Environmental Benefits of the Solar America

Initiative. Prepared for the National Renewable Energy Laboratory by ECONorthwest. NREL

Report SR-640-41998.

Herndon, A. (January 20, 2009). “Sunrise in the West: California’s $1.9bn Renewable Energy

Transmission Line.” Analyst Reaction. London: New Energy Finance.

Maycock, Paul D. (February 2002). “World PV Cell/Module Production (1988-2001).” PV

News.

McCrone, A.; Peyvan, M.; Zindler, E. (June 17, 2009). “Net Job Creation to 2025: Spectacular in

solar, but modest in wind.” Research Note. London: New Energy Finance.

Mehta, S.; Bradford, T.; (2009). PV Technology, Production, and Cost, 2009 Forecast: The

Anatomy of a Shakeout. Prometheus Institute and Greentech Media.

Merrill Lynch. (2008). Solar Thermal: Not Just Smoke and Mirrors. New York, NY: Merrill

Lynch.

Morgan Stanley. (2009). Solar Devices: Dislocation – Industry Reset. Morgan Stanley.

Mints, P.; Tomlinson, D. (2007). Photovoltaic Manufacturer Shipments & Competitive Analysis

2006/2007. Report # NPS-Supply2. Palo Alto, CA: Navigant Consulting Photovoltaic Service

Program.

Mints, P.; Tomlinson, D. (2008). Photovoltaic Manufacturer Shipments & Competitive Analysis

2007/2008. Report # NPS-Supply3. Palo Alto, CA: Navigant Consulting Photovoltaic Service

Program.

Mints, P. (2009). Photovoltaic Manufacturer Shipments, Capacity, & Competitive Analysis

2008/2009. Report # NPS-Supply4. Palo Alto, CA: Navigant Consulting Photovoltaic Service

Program.

Navigant Consulting (2008). Economic Impacts of Extending Federal Solar Tax Credits.

Prepared for the Solar Energy Research and Education Foundation (SEREF).

http://www.seia.org/cs/news_research_and_multimedia/research/. Accessed February 2009.

Oppenheimer. (2009). PV Market Forecast. Oppenheimer & Co.

Podewils, C. (February 2008). “On the glass path.” Photon International. pp.74–79.

Renner, M.; Sweeney, S.; Kubit, J. (2008). Green Jobs: Towards decent work in a sustainable,

low-carbon world. Washington, DC: Worldwatch Institute. Commissioned and funded by United

Nations Environment Programme (UNEP).

http://www.unep.org/labour_environment/features/greenjobs-report.asp. Accessed July 2009.

Resseguie, L. (November 2008). E-mail correspondence. U.S. Bureau of Land Management

(BLM).

Rogol, M.; Choi, P.; Conkling, J.; Fotopoulos, A.; Peltzman, K.; Roberts, S. (2006). Solar

Annual 2006: The Gun Has Gone Off. Solar Verlag GmbH/Photon Consulting.

Rogol, M.; Farber, M.; Flynn, H.; Meyers, M.; Paap, S.; Porter, C.; Rogol, J.; Song, J. (2008).

Solar Annual 2008: Four Peaks. Photon Consulting.

Schwer, R.K.; Riddel, M. (February 2004). The Potential Economic Impact of Constructing and

Operating Solar Power Generation Facilities in Nevada. Prepared for the National Renewable

Energy Laboratory by the Center for Business and Economic Research at UNLV. NREL Report

SR-550-35037.

Stoddard, L.; Abiecunas, J.; O’Connell, R. (April 2006). Economic, Energy and Environmental

Benefits of Concentrating Solar Power in California. Prepared for the National Renewable

Energy Laboratory by Black & Veatch. NREL Report SR-550-39291.

Thomas Weisel. 2009. Alternative Energy; Presentation Materials to Investors. Thomas Weisel

Partners.

U.S. DOE. (2009). Report to Congress. Concentrating Solar Power Commercial Application

Study: Reducing Water Consumption of Concentrating Solar Power Electricity Generation. U.S.

Department of Energy, www.eere.energy.gov/solar/pdfs/csp_water_study.pdf.

U.S. Department of Energy, Energy Information Administration (EIA). (2008a). Solar

Photovoltaic Cell/Module Manufacturing Activities 2003-2007.

http://www.eia.doe.gov/cneaf/solar.renewables/page/solarreport/solarpv.html. Accessed May 28,

2009.

U.S. Department of Energy, Energy Information Administration (EIA). (2008b). Solar Thermal

and Photovoltaic Collector Manufacturing Activities 2003-2007.

http://www.eia.doe.gov/cneaf/solar.renewables/page/solarreport/solar.html. Accessed May 28,

2009.

U.S. Department of Energy, Energy Information Administration (EIA). (2009). Solar

Photovoltaic Cell/Module Manufacturing Activities 2008.

http://www.eia.doe.gov/cneaf/solar.renewables/page/solarreport/solarpv.pdf. Accessed January

12, 2010.

U.S. Department of the Interior (DOI). October 28, 2009. “Memorandum of Understanding

among the U.S. Department of Agriculture, Department of Commerce, Department of Defense,

Department of Energy, Environmental Protection Agency, the Council on Environmental

Quality, the Federal Energy Regulatory Commission, the Advisory Council on Historic

Preservation, and the Department of the Interior, regarding coordination in federal agency review

of electric transmission facilities on federal land.”

http://www.doi.gov/news/09_News_Releases/102809a.html. Accessed November 25, 2009.

Wei, M; Patadia, S.; Kammen, D. (2010). “Putting Renewables and Energy Efficiency to Work:

How Many Jobs Can the Clean Energy Industry Generate in the U.S.?” Energy Policy. Volume

38. pp. 919-931. http://rael.berkeley.edu/greenjobs. Accessed January 12, 2010.

Winegarner, R.M.; Johnson, R.O. (2006). Polysilicon 2006. Sage Concepts, Inc./Solar Insights

LLC.

Wu, J.; Chase, J. (2009). “Silicon and Wafer Spot Price Index.” Analyst Reaction. London: New

Energy Finance.

Yahoo Finance. (2009). Market Vectors Steel ETF (SLX).

http://finance.yahoo.com/q/hp?s=SLX&a=11&b=31&c=2005&d=00&e=1&f=2009&g=m.

Accessed July 2009.

Zweibel, K. (2005). The Terawatt Challenge for Thin-Film PV. NREL/TP-520-38350. Golden,

CO: National Renewable Energy Laboratory.

3. Cost, Price, and Performance Trends

This chapter covers cost, price, and performance trends for PV and CSP. Sections 3.1 and 3.2

discuss levelized cost of energy, solar resource, and capacity factor for PV and CSP. Section 3.3

provides information on efficiency trends for PV cells, modules, and systems. Section 3.4

discusses PV module reliability. Sections 3.5 and 3.6 cover PV module and installed-system cost

trends. Section 3.7 discusses PV operations and maintenance (O&M) trends. Section 3.8

summarizes CSP installation and O&M cost trends, and Section 3.9 presents information on the

characteristics and performance of various CSP technologies.

3.1. Levelized Cost of Energy, PV and CSP

Levelized cost of energy (LCOE) is the ratio of an electricity-generation system's amortized

lifetime costs (installed cost plus lifetime O&M costs) to the system's lifetime electricity

generation. The calculation of LCOE is highly sensitive to installed system cost, O&M costs,

location, orientation, financing, and policy. Thus it is not surprising that estimates of LCOE vary

widely across sources.

One recent source estimates that worldwide, the range of LCOE is approximately $0.20–$0.80

per kWh for rooftop PV and $0.12–$0.18 per kWh for parabolic trough CSP, not including

government incentives (REN21 2008). The wide LCOE range for PV ($0.20–$0.80 per kWh) is

due largely to the effect on LCOE of location and the corresponding solar radiation (insolation).

Worldwide, typical PV LCOE ranges are $0.20–$0.40 per kWh for low latitudes with high

insolation of 2,500 kWh/m

/year, $0.30–$0.50 per kWh for 1,500 kWh/m

/year (typical of

Southern Europe), and $0.50–$0.80 per kWh for 1,000 kWh/m

/year (higher latitudes) (REN21

2008). See Figure 3.2 for a map of the solar resource in the United States, Germany, and Spain.

Figure 3.1 shows LCOE for residential PV systems in selected U.S. cities ranging from about

$0.20/kWh to more than $0.32/kWh (when calculated with the federal ITC) based on the quality

of the solar resource. Without the ITC, the range for these same cities is about $0.28/kWh to

$0.46/kWh.

Figure 3.1. LCOE for residential PV systems in several U.S. cities in 2008, with and without the

federal investment tax credit (NREL 2009a)

The LCOEs of commercial and utility-scale PV systems are generally lower than those of

residential PV systems located in the same region. Installed and O&M costs per watt tend to

decrease as PV system size increases owing to more advantageous economies of scale and other

factors (see Section 3.6 on PV Installation Cost Trends and Section 3.7 on PV Operations and

Maintenance). In addition, larger, optimized, better-maintained PV systems can produce

electricity more efficiently and consistently.

3.2. Solar Resource and Capacity Factor, PV and CSP

Of all the renewable resources, solar is by far the most abundant. With 162,000 terawatts

reaching Earth from the sun, just 1 hour of sunlight could theoretically provide all of society’s

energy needs for 1 year.

3.2.1. Solar Resource for PV

Photovoltaics can take advantage of direct and indirect (diffuse) insolation, whereas CSP is

designed to use only direct insolation. As a result, PV modules need not directly face and track

incident radiation as CSP systems must do. This has enabled PV systems to have broader

geographical application than CSP.

Figure 3.2 illustrates the photovoltaic solar resource in the United States, Germany, and Spain

for a flat-plate PV collector tilted south at latitude. Solar resources across the United States are

mostly good to excellent, with solar insolation levels ranging from about 1,000–2,500

Assumptions for Figure 3.1: residential market, 30-year analysis period, 30-year mortgage, 100% financed, 6% interest rate,

6% discount rate, marginal tax rate 35%, installed cost $7.5/W, $300 inverter replacement at year 10, $280 inverter replacement

at year 20, panel degradation 0.5%/year, ITC covers 30% of initial cost, no state or local incentives.

$0.00

$0.05

$0.10

$0.15

$0.20

$0.25

$0.30

$0.35

$0.40

$0.45

$0.50

Seattle,

Chicago, IL NYC, NY Washington

D.C.

Miami, FL Dallas, TX San

Francisco,

Denver, CO Los

Angeles,

Phoenix,

Levelized Cost of Energy (U.S.$/kWh)

W ith ITC

Without ITC

Increasing insolation

kWh/m

/year. The southwest U.S. is at the top of this range, while only Alaska and part of

Washington are at the low end. The range for the mainland United States is about 1,350–2,500

kWh/m

/year. The U.S. solar insolation level varies by about a factor of 2; this is considered

relatively homogeneous compared to other renewable energy resources.

As is evident from the map, the solar resource in the United States is much higher than in

Germany, and the southwest United States has better resource than southern Spain. Germany’s

solar resource has about the same range as Alaska’s, at about 1,000–1,500 kWh/m

/year, but

more of Germany’s resource is at the lower end of that range. Spain’s solar insolation ranges

from about 1,300–2,000 kWh/m

/year, which is among the best solar resource in Europe.

Figure 3.2 Photovoltaic solar resource for the United States, Spain, and Germany

(NREL 2009d)

Solar PV resource maps are typically provided for horizontal flat-plate PV collectors versus flat-

plate collectors tilted south at latitude. A horizontally oriented module produces less energy per

Annual average solar resource data are for a solar collector oriented toward the south at tilt = local latitude. The

data for Hawaii and the 48 contiguous states are derived from a model developed at SUNY/Albany using

geostationary weather satellite data for the period 1998–2005. The data for Alaska were derived by NREL in 2003

from a 40-km satellite and surface cloud cover database for the period 1985–1991. The data for Germany and Spain

were acquired from the Joint Research Centre of the European Commission and capture the yearly sum of global

irradiation on an optimally inclined surface for the period 1981–1990. State and countries are shown to scale, except

for Alaska.

unit of module power than the same module tilted south or equipped with a system that tracks the

sun. However, horizontal PV systems can be beneficial when land constraints become a

significant factor, because they have a greater energy density per unit of land area than tilted or

tracking systems. For a fixed land area, more electricity can be generated from a horizontal PV

system than from a tilted or tracking system owing to space needed to avoid self-shading and to

allow for maintenance. On flat commercial roofs, it is cheaper to install flat or nearly flat systems

(Denholm and Margolis 2008b).

The total land area suitable for PV is enormous and will not limit PV deployment. For example,

a current estimate of the total roof area suitable for PV in the United States is approximately

6 billion square meters, even after eliminating 35% to 80% of roof space to account for panel

shading (e.g., by trees) and suboptimal roof orientations. With current PV performance, this area

has the potential for more than 600 GW of capacity, which could generate more than 20% of

U.S. electricity demand. Beyond rooftops, there are many opportunities for installing PV on

underutilized real estate such as parking structures, awnings, airports, freeway margins, and

farmland set-asides. The land area required to supply all end-use electricity in the United States

using PV is about 0.6% of the country's land area (181 m

per person) or about 22% of the

“urban area” footprint (Denholm and Margolis 2008c).

3.2.2 Solar Resource for CSP

The geographic area that is most suitable for concentrating solar power is smaller than for PV

because CSP uses only direct insolation. In the United States, the best location for CSP is the

Southwest. Globally, the most suitable sites for CSP plants are arid lands within 35° north and

south of the equator. Figure 3.3 shows the direct-normal solar resource in the southwestern

United States, which includes a detailed characterization of regional climate and local land

features; red indicates the most intense solar resource, and light blue indicates the least intense.

Figure 3.4 shows locations in the southwestern United States with characteristics ideal for CSP

systems, including direct-normal insolation greater than 6.75 kWh/m

/day, a land slope of less

than 1°, and at least 10 km

of contiguous land that could accommodate large systems (Mehos

and Kearney 2007).

After implementing the appropriate insolation, slope, and contiguous land area filters,

53,900 square miles are available in the seven states considered to be CSP compatible:

California, Arizona, New Mexico, Nevada, Colorado, Utah, and Texas. Table 3.1 summarizes

the land area in these states that is ideally suited to CSP. This relatively small land area amounts

to nearly 6,900 GW of resource potential and more than 16 million GWh of generating capacity,

assuming a capacity factor between 25% and 50% (see Section 3.2.3) (Mehos and Kearney 2007,

Andraka 2008). This is quadruple the annual U.S. electricity generation of about 4 million

GWh.

EIA Net Generation by Energy Source: Total (All Sectors), rolling 12 months ending in June 2009

www.eia.doe.gov/cneaf/electricity/epm/table1_1.html.

Figure 3.3. Direct-normal solar resource in the

U.S. Southwest

(Mehos and Kearney 2007)

Figure 3.4. Direct-normal solar radiation in the

U.S. Southwest, filtered by resource, land use,

and topography

(Mehos and Kearney 2007)

Table 3.1. Ideal CSP Land Area and Resource Potential in Seven Southwestern States

State

Available Area (square miles)

Resource Potential (GW)

Arizona

19,300

2,468

California

6,900

877

Colorado

2,100

272

Nevada

5,600

715

New Mexico

15,200

1,940

Texas

1,200

149

Utah

3,600

456

Total

53,900

6,877

Mehos and Kearney 2007

3.2.3 Capacity Factor, PV and CSP

Capacity factor is the ratio of an energy-generation system's actual energy output during a given

period to the energy output that would have been generated if the system ran at full capacity for

the entire period. For example, if a system ran at its full capacity for an entire year, the capacity

factor would be 100% during that year. Because PV and CSP generate electricity only when the

sun is shining, their capacity factors are reduced because of evening, cloudy, and other low-light

periods. This can be mitigated in part by locating PV and CSP systems in areas that receive high

levels of annual sunlight. The capacity factor of PV and CSP systems is also reduced by any

necessary downtime (e.g., for maintenance), which is also the case for other generation

technologies.

For PV, electricity generation is maximized when the modules are normal (i.e., perpendicular) to

the incident sunlight. Variations in the sun's angle that are due to the season and time of day

reduce the capacity factor of fixed-orientation PV systems. This can be mitigated in part by

CSP power plants require about 5 acres of land area per MW of installed capacity. Electricity generation for a CSP plant can be

estimated by assuming an average annual solar capacity factor of 25% to 50%, depending on the amount of thermal storage.

tilting stationary PV modules to maximize annual sunlight exposure or by incorporating one- or

two-axis solar tracking systems, which rotate the modules to capture more normal sunlight

exposure than is possible with stationary modules. Figure 3.5 shows the effect of insolation and

use of tracking systems on PV capacity factors. Fixed tilt (at latitude) capacity factors are 14%–

24% for Seattle to Phoenix, whereas 1- and 2-axis tracking systems result in higher ranges.

Analysts sometimes use 18% or 19% for an average U.S. PV capacity factor.

Figure 3.5. PV capacity factors varying by insolation and use of tracking systems

(NREL 2009b)

The most recently built CSP trough, tower, and dish-engine systems have AC capacity factors in

the mid-20% range. With 6 hours of thermal storage, capacity factors increase to about 40%, and

additional increases in thermal storage will enable capacity factors and dispatchability (the

ability to increase or decrease electricity generation on demand) to increase even more.

3.3 PV Cell, Module, and System Efficiency

In addition to the solar resource and capacity factor discussed above, the amount of electricity

produced by PV systems depends primarily on the following factors:

• Cell type and efficiency

• Module efficiency

• System efficiency

• Module reliability.

These are DC capacity factors, i.e., based on the DC rating of a PV system and taking into account inverter and other system

losses. By definition, they are lower than an AC capacity factor, which is how fossil, nuclear, and CSP plants are rated and thus

are not directly comparable to more traditional AC capacity factors.

Capacity factors were estimated using data from NREL’s PVWatts, a performance calculator for on-grid PV systems:

http://rredc.nrel.gov/solar/codes_algs/PVWATTS/version1/. The capacity factors shown here reflect an overall derate factor of

0.77, with the inverter and transformer component of this derate being 0.92, the defaults used in PVWatts. The array tilt is at

latitude for the fixed tilt systems, the default in PVWatts.

24%

22%

21%

20%

18%

17%

14%

31%

27%

26%

25%

23%

22%

18%

33%

28%

26%

24%

23%

19%

10%

15%

20%

25%

30%

35%

Seattle, WA Chicago, IL Boston, MA Miami, FL Fort Worth,

Los Angeles,

Phoenix, AZ

Capacity Factor

Fixed Tilt

1-Axis Tracking

2-Axis Tracking

Increasing insolation

This section discusses the efficiency of PV cells, modules, and systems. Module reliability is

discussed in the next section.

3.3.1 PV Cell Type and Efficiency

Two categories of PV cells are used in most of today's commercial PV modules: crystalline

silicon and thin film. The crystalline silicon category, called first-generation PV, includes

monocrystalline and multicrystalline PV cells, which are the most efficient of the mainstream PV

technologies and accounted for about 84% of PV produced in 2008 (Bartlett et al. 2009). These

cells produce electricity via crystalline silicon semiconductor material derived from highly

refined polysilicon feedstock. Monocrystalline cells, made of single silicon crystals, are more

efficient than multicrystalline cells but are more expensive to manufacture.

The thin-film category, called second-generation PV, includes PV cells that produce electricity

via extremely thin layers of semiconductor material made of amorphous silicon (a-Si), copper

indium diselenide (CIS), copper indium gallium diselenide (CIGS), or cadmium telluride (CdTe).

Another PV cell technology (also second generation) is the multijunction PV cell. Multijunction

cells use multiple layers of semiconductor material (from the group III and V elements of the

periodic table of chemical elements) to absorb and convert more of the solar spectrum into

electricity than is converted by single-junction cells. Combined with light-concentrating optics

and sophisticated sun-tracking systems, these cells have demonstrated the highest sunlight-to-

electricity conversion efficiencies of any PV technologies, in excess of 40%.

Various emerging technologies, known as third-generation PV, could become viable commercial

options in the future, either by achieving very high efficiency or very low cost. Examples include

dye-sensitized and organic PV cells, which have demonstrated relatively low efficiencies to date

but offer the potential for substantial manufacturing cost reductions.

The efficiencies of all PV cell types have improved over the past several decades, as illustrated

in Figure 3.6, which shows the best research-cell efficiencies from 1975 to 2008. The highest-

efficiency research cell shown is a multijunction concentrator at 41.6% efficiency. Other

research-cell efficiencies illustrated in the figure range from 20% to almost 28% for crystalline

silicon cells, 12% to almost 20% for thin film, and about 5% and to 11% for the emerging PV

technologies organic cells and dye-sensitized cells, respectively.

Figure 3.6. Best research cell efficiencies 1975–2009

(Kazmerski 2009)

3.3.2 PV Module Efficiency

The cells described in Figure 3.6 were manufactured in small quantities under ideal laboratory

conditions and refined to attain the highest possible efficiencies. The efficiencies of mass-

produced cells are always lower than the efficiency of the best research cell. Further, the

efficiency of PV modules is lower than the efficiency of the cells from which they are made.

In 2008, the typical efficiency of crystalline silicon-based PV modules ranged from 13.5% for

multicrystalline modules to 17.5% for high-efficiency monocrystalline modules (Mehta and

Bradford 2009). For thin-film modules, typical efficiencies ranged from 6.5% for a-Si modules

to about 10% for CIGS and CdTe modules (Mehta and Bradford 2009).

Figure 3.7 shows best-in-class module efficiencies from 1999 to 2008, with the best crystalline

silicon efficiencies at 17%–19% and the best thin-film efficiencies at 7%–11% in 2008.

Figure 3.7. Best-in-class commercial module efficiencies, 1999–2008, compiled from module

survey data (Kreutzmann 2008, Photon International 1999–2008)

3.3.3 PV System Efficiency and Derate Factor

A PV system consists of multiple PV modules wired together and installed on a building or other

location. The AC output of a PV system is always less than the DC rating, which is due to

system losses.

For grid-connected applications, a PV system includes an inverter that transforms the DC

electricity produced by the PV modules into AC electricity. The average maximum efficiency of

inverters was 95.5% in 2008, up from 94.7% in 2005 (Knoll and Kreutzmann 2008). Other

factors that reduce a PV system's efficiency include dirt and other materials obscuring sun-

collecting surfaces, electrically mismatched modules in an array, wiring losses, and high cell

temperatures. For example, NREL’s PVWatts,

3.4 PV Module Reliability

a performance calculator for on-grid PV

systems, uses an overall derate factor of 0.77 as a default, with the inverter component of this

derate being 0.92.

Historic data suggest that reliability is a very important factor when considering the market

adoption of a new technology, especially during the early growth stages of an industry. PV is

currently experiencing unprecedented growth rates. To sustain these growth rates, it is imperative

that manufacturers consider the implications of product reliability.

http://rredc.nrel.gov/solar/codes_algs/PVWATTS/version1/

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Efficiency (%)

a-Si

CdTe

CIS

Monocrystalline

HIT (monocrystalline & a-Si)

Multicrystalline

Today’s PV modules usually include a 25-year warranty. Standard warranties guarantee that

output after 25 years will be at least 80% of rated output. This is in line with real world

experience and predicted performance from damp-heat testing of modules (Wohlgemuth et al.

2006).

Manufacturers in the United States, Japan, and the European Union currently implement

qualification standards and certifications that help to ensure that PV systems meet reliability

specifications. There have been efforts to bring reliability standards to Chinese manufacturers as

well, considering their rapid growth in the PV market. DOE has been a leader in engaging

Chinese manufacturers in discussions on reliability standards and codes by organizing a series of

reliability workshops and conferences in China. The global PV community realizes that if

reliability standards are not quickly implemented among the fastest-growing producers,

consumers left with high-maintenance installations could negatively impact market adoption of

PV modules both now and in the future.

3.5 PV Module Price Trends

Photovoltaic modules have experienced significant improvements and cost reductions over the

last few decades, and the market for PV modules has undergone unprecedented growth in recent

years owing to government policy support and other financial incentives encouraging the

installation of (primarily grid-connected) PV systems. Although PV module prices increased in

the past several years, prices have been falling steadily over the past few decades and began

falling again in 2008. This is illustrated in Figure 3.8, which presents average global PV module

selling prices for all PV technologies.

Although global average prices provide an index for the PV industry overall, a few caveats are in

order. First, the PV industry is dynamic and rapidly changing, with advances in cost reductions

for segments of the industry masked by looking at average prices. For example, thin-film PV

technologies are achieving manufacturing costs and selling prices significantly lower than for

crystalline silicon modules. See Table 3.2 (at the end of Section 3.5) for comparison of module

costs and prices for various PV technologies. Applications such as large ground-mounted PV

systems, for which deployment is increasing, and applications in certain countries and locations,

accrue cost advantages based on factors such as economies of scale and the benefits of a more

mature market (some of this is captured in Section 3.6 on PV installation cost trends). Finally,

historical trends may not provide an accurate picture going forward, as new developments

continue to change the PV industry landscape.

Figure 3.8 shows average global PV module selling prices at the factory gate (i.e., prices do not

include charges such as delivery and subsequent taxes), as obtained from sample market

transactions. As stated earlier, these prices are averages over all PV module technologies

including thin film, for PV power modules greater than 75 W in size and purchased in relatively

small quantities. Excluded from this category are smaller PV modules, consumer indoor PV

(e.g., for calculators and watches), and large (e.g., greater than 150 W) standard modules sold in

large quantities (e.g., 500–1,000-unit minimums) in the unrestricted international commodity

market. The large-module/large-quantity category has been increasing in market share in recent

years and constituted more than half of module sales in 2008. In 2008, the average price per watt

for the large-module/large-quantity category was 11% lower than for the power module category

shown in Figure 3.8. The current/nominal prices shown in the figure are actual prices paid in the

year stated. The 2008/real prices are adjusted for inflation (indexed to 2008 U.S. dollars).

Figure 3.8. Global, average PV module prices, all PV technologies, 1980–2008

(Mints 2006, Mints 2009, U.S. Bureau of Economic Analysis 2009)

PV module prices experienced significant drops in the mid-1980s resulting from increases in

module production and pushes for market penetration during a time of low interest in renewable

energy. Prices dropped by more than $10/W (in real 2008$) between the early- to mid-1980s.

The price of PV by 1987 was approximately $9/W. PV prices then experienced an increase from

1988 through 1990 as the supply of PV modules diminished because of a limitation in the

availability of silicon wafers. For the first time in a decade, the market was limited by supply

rather than demand. Prices then dropped significantly from 1991 to 1995 because of increases in

manufacturing capacity and a worldwide recession that slowed PV demand. Module prices

continued to fall, although at a slower rate from 1995 to 2003, which was due to global increases

in module capacities and a growing market. By 2000, module prices were below $5/W, reaching

a low of $3/W in 2003 (Mints 2009).

Prices began to increase from 2003 to 2006 as European demand, primarily from Germany and

Spain, experienced high growth rates after feed-in tariffs and other government incentives were

adopted. Also contributing to the price increases was an imbalance between polysilicon supply

and demand from around 2004 to mid-2008. Higher prices were sustained until the third quarter

of 2008 when the global recession reduced demand, polysilicon supply constraints eased, and

Current/nominal data provided by Mints, then adjusted for inflation using GDP deflator from the U.S. Bureau of

Economic Analysis.

See Table 3.2 later in this section for comparison of module costs and prices for various PV technologies.

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

Average Module Price (Current and 2008$/W)

2008/Real U.S.$/W

(Mints 2009, U.S. BEA 2009)

Current/Nominal U.S.$/W

(Mints 2009)

module supply increased. For the first time since 2003, average module prices declined to

$3.65/W, down from approximately $4/W in 2007 (in real 2008$).

Module prices vary considerably by technology (Table 3.2), influenced by variations in

manufacturing cost and sunlight-to-electricity conversion efficiency, among other factors.

Manufacturing cost is the reference point against which a manufacturer sets its profit margin; the

closer the selling price is to the manufacturing cost, the lower the profit margin and vice versa.

Higher conversion efficiency generally commands a price premium. This is because higher-

efficiency modules require less installation area per watt of electricity production and incur

lower balance-of-systems costs (i.e., wiring, racking, and other system installation costs) per watt

than lower-efficiency modules. The current estimated effect is a $0.10 increase in price per 1%

increase in efficiency; for example, all else being equal, a 20%-efficient module would cost

about $1 more per watt than a 10%-efficient module (Mehta and Bradford 2009).

Table 3.2. Module Price, Manufacturing Cost, and Efficiency Estimates by Technology, 2008

Technology

Price

(2008 $/Wp)

Manufacturing Cost

(2008 $/Wp)

Conversion

Efficiency

High-efficiency monocrystalline silicon

$3.83

$2.24

17.5%

Multicrystalline silicon

$3.43

$2.12–$3.11

13.5%

Amorphous silicon (a-Si) thin film

$3.00

$1.80

6.5%

Copper indium diselenide/copper indium

gallium diselenide (CIS/CIGS) thin film

$2.81 $1.26 10.2%

Cadmium telluride (CdTe) thin film

$2.51

$1.25

10.0%

Mehta and Bradford 2009

3.6 PV Installation Cost Trends

Lawrence Berkeley National Laboratory (LBNL) has collected project-level installed system cost

data for grid-connected, customer-sited PV installations in the Unites States from a number of

solar incentive program administrators (Wiser et al. 2009). The dataset currently includes

approximately 52,000 PV systems installed in 16 states between 1998 and 2008 and totals

565 MW, or 71% of all grid-connected PV capacity installed in the United States through 2008.

Below are trends related to the installed system cost of the PV projects in the LBNL database. In

all instances, installed costs are expressed in terms of real 2008 dollars and represent the cost to

the consumer before receipt of any grant or rebate. PV capacity is expressed in terms of the rated

module direct-current (DC) power output under Standard Test Conditions. Note that the

terminology “installed cost” in this report represents the price paid by the end-user/customer.

This should not be confused with the term “cost” as used in other contexts to refer to the cost to a

company before a product is priced for a market or end user.

The range in manufacturing cost for multicrystalline silicon includes $2.12 for global, vertically integrated PV module

manufacturers, $2.74 for European manufacturers, and $3.11 for Asian manufacturers, with much of the difference being due to

the price of polysilicon feedstock paid in 2008 by each type of producer.

This section on PV installation cost trends was provided by Wiser et al. 2009, of LBNL, before the publication of their full

report, “Tracking the Sun II. The Installed Cost of Photovoltaics in the U.S. from 1998-2008,”

http://eetd.lbl.gov/ea/EMS/re-

pubs.html, with the full reference provided in the Chapter 3 references. Note that some of the numbers presented in this section

may be slightly different from numbers in the published LBNL report, though the numbers stated in the text corresponding to the

first figure of this section, Figure 3.9, are the same.

Figure 3.9 presents the average installed cost from 1998 to 2008 for the entire sample of projects.

Capacity-weighted average costs declined from $10.8/W in 1998 to $7.5/W in 2008. This

represents an average annual reduction of $0.3/W (i.e., a drop of 3.6% per year in real

dollars). Figure 3.9 also shows that the distribution in installed costs, indicated by the standard

deviation bars, has narrowed over time, suggesting a maturing market with converging costs.

Both average installed costs and the distribution in installed costs, however, remained relatively

stable from 2005 to 2007. Costs declined from 2007 to 2008 at the historical average pace of

$0.3/W.

Figure 3.9. Installed cost trends over time

(Wiser et al. 2009)

The long-term decline in installed costs from 1998 to 2008 is attributable primarily to a drop in

non-module costs. Non-module costs can include inverters, other hardware, labor, permitting and

fees, shipping, overhead, and profit. As shown in Figure 3.10, the average non-module costs

(calculated as the difference between average total installed cost and a module price index

)

declined from approximately $5.9/W in 1998 to $3.8/W in 2008, a drop of $2.1/W. By

comparison, module prices dropped by only $1.3/W over this 11-year period. From 2005 through

2008, however, non-module costs remained relatively stable; in fact, from 2007 to 2008, non-

module costs increased slightly while module costs declined by $0.5/W.

The global, average annual price of power modules published by Navigant Consulting is used (see the previous section,

Section 3.5 on PV Module Price Trends).

$10

$12

$14

$16

1998

n=39

0.2 MW

1999

n=178

0.7 MW

2000

n=213

0.9 MW

2001

n=1239

4.8 MW

2002

n=2290

12 MW

2003

n=3334

31 MW

2004

n=5498

46 MW

2005

n=5193

60 MW

2006

n=8681

90 MW

2007

n=12107

123 MW

2008

n=13013

195 MW

Installation Year

Installed Cost (2008$/W)

Capacity-Weighted Average

Simple Average +/- Std. Dev.

Figure 3.10. Module and non-module cost trends over time

(Wiser et al. 2009)

Although experience to date confirms that significant cost reductions occurred in the United

States from 1998 through 2008, international experiences suggest that further cost reductions are

possible and may accompany increased market size. Figure 3.11 compares installed costs in

Japan, Germany, and the United States, focusing specifically on 2–5-kW residential systems

installed in 2007 (and excluding sales tax).

Among these systems, average installed costs were

substantially lower in Japan and Germany ($6.1/W and $6.8/W, respectively) than in the United

States ($8.1/W). These differences may partly reflect the much greater cumulative grid-

connected PV capacity in Japan and Germany (about 1.8 GW and 3.8 GW, respectively) at the

end of 2007 compared to just 0.5 GW in the United States. However, it is also evident that larger

market size alone does not account for all of the variation; installed costs are higher in Germany

than in Japan, despite Germany’s grid-connected PV capacity being substantially greater.

Year 2008 data are not yet available for Germany and Japan.

One potential explanation for the relatively low residential PV costs in Japan is that a large portion of the

residential PV market consists of pre-fabricated homes that include PV as a standard feature. More generally,

installed costs may differ among countries as a result of a wide variety of factors, including differences in module

prices, technical standards for grid-connected PV systems, installation labor costs, procedures for receiving

incentives and interconnection approvals, and the degree to which components are manufactured locally (thereby

reducing transportation costs).

$10

$12

1998

n=39

0.2 MW

1999

n=178

0.7 MW

2000

n=213

0.9 MW

2001

n=1239

4.8 MW

2002

n=2290

12 MW

2003

n=3334

31 MW

2004

n=5498

46 MW

2005

n=5193

60 MW

2006

n=8681

90 MW

2007

n=12107

123 MW

2008

n=13013

195 MW

Installation Year

Capacity-Weighted Average

Installed Cost (2008$/W)

Global Module Price Index

Non-Module Cost (calculated)

Total Installed Cost

Note: Non-module costs are calculated as the reported total installed costs minus the global module price index.

Figure 3.11. Average installed cost of residential systems completed in 2007

in Japan, Germany, and the United States

(Wiser et al. 2009)

The United States is not a homogenous PV market, as evidenced by Figure 3.12, which compares

the average installed cost of PV systems <10 kW completed in 2008 across 16 states. Average

costs range from a low of $7.3/W in Arizona to a high of $10.5/W in Maryland. Differences in

the average installed cost across these states may, in part, reflect the differing size and maturity

of their PV markets. Specifically, the largest PV markets in the United States (California, New

Jersey, and Colorado

) have among the lowest average costs, and a number of smaller but

relatively mature markets (Arizona, Connecticut, and Massachusetts) also have low costs. In

addition, 8 of the 16 states shown in the Figure 3.12 (Arizona, Connecticut, Massachusetts,

Minnesota, New Jersey, New York, Washington, and Vermont) exempted residential PV systems

from state sales tax in 2008, and Oregon has no state sales tax. Sales tax exemptions effectively

reduce post-sales tax installed costs by $0.2–0.4/W, depending on the otherwise applicable sales

tax rate and assuming that PV hardware costs represent approximately 65% of the total installed

cost of residential PV systems (an assumption supported by data presented later in this section).

The LBNL dataset does not include data from Colorado’s largest PV incentive program (Xcel Energy’s Solar

Reward’s Program). Thus, the sample of Colorado projects is small, even though the total number of projects

installed in the state in 2008 is relatively large.

$10

Japan Germany United States

Average Residential PV

Installed Cost without Sales

Tax (2008$/W)

1,000

2,000

3,000

4,000

5,000

Cumulative Grid-Connected

Capacity through 2007 (MW)

Avg. Installed Cost of Residential Systems in 2007 (left axis)

Cumulative Grid-Connected Capacity through 2007 (right axis)

Figure 3.12. Variation in installed costs among U.S. states

(Wiser et al. 2009)

The decline in U.S. PV installed costs over time is partly attributable to the fact that PV systems

have gotten larger, on average, and exhibit some economies of scale. As shown in Figure 3.13,

the average size of systems ≤10 kW (a proxy for residential systems) grew from 2.7 kW in 1998

to 4.6 kW in 2008.

Similarly, the average size of systems >10 kW (a proxy for non-residential

systems) rose from 25 to 88 kW over the same period, although the trend is notably more uneven

than for the smaller systems. As confirmed by Figure 3.14, installed costs generally decline as

system size increases. In particular, the average installed cost of systems installed in 2008 was

greatest for systems <2 kW, at $9.2/W, dropping to $6.2/W for systems in the 500–750-kW size

range, a decline of approximately $2.6/W, or about 28%.

Of interest, however, is that the

economies of scale are not continuous but, rather, are most apparent for systems at the lower and

upper ends of the size spectrum (i.e., for systems in the <5 kW range and the >250 kW range). In

addition, the average cost of systems >750 kW in 2008 was slightly higher than the average cost

of 500–750 kW systems ($7.1/W vs. $6.6/W, respectively), perhaps indicative of a greater

proportion of tracking systems within the larger size range.

The data provided by many PV incentive programs did not identify customer type (i.e., residential vs. non-

residential); thus ≤10 kW is used as a proxy for residential systems, recognizing that a non-trivial portion of systems

≤10 kW are, in fact, non-residential.

Note that the LBNL data do not include several larger PV systems installed in 2008 including a 2.4-MW plant in

Fontana, CA; a 3.0-MW plant in Fairless Hills, PA; and a 12.6-MW plant in Boulder City, NV. Installed costs for

the CA and NV projects ($4.3/W and $3.2/W, respectively), as reported in press releases, are considerably less than

the average installed costs shown in Figure 8 for customer-sited projects >750 kW.

$10

$12

$14

$16

$18

n=431

n=8594

n=16

n=249

n=296

n=690

n=356

n=201

n=6

n=141

n=131

n=89

n=125

n=16

n=18

n=28

Average Installed Cost (2008$/W)

State Sales Tax (if assessed)

Pre-Sales Tax Installed Cost (Avg. +/- Std. Dev.)

Systems <10 kW, Installed in 2008

Note: State Sales Tax and Pre-Sales Tax Installed Costs were calculated from 2008 sales tax rates in each state, accounting

for sales tax exemptions for PV systems. Sales tax was assumed to have been assessed only on hardware costs, which in

turn were assumed to constitute 65% of the total pre-sales-tax installed cost.

Figure 3.13. PV system size trends over time

(Wiser et al. 2009)

Figure 3.14. Variation in installed cost according to PV system size

(Wiser et al. 2009)

In addition to variation across states and system size, installed costs also vary across key market

segments and technology types. Figure 3.15 compares the average installed cost of residential

retrofit and new construction systems completed in 2008, showing separate comparisons for both

rack-mounted and building-integrated photovoltaic (BIPV), and focusing on systems of 1–3 kW,

because that is the size range typical of new construction systems. Rack-mounted systems

installed in residential new construction average $1.2/W less than comparably sized residential

retrofits. For BIPV systems, residential new construction systems average $1.5/W less than

residential retrofits. Figure 3.16 compares installed costs of systems using crystalline silicon

versus thin-film modules, among rack-mounted systems installed in 2008. The data indicate that,

in both the <10-kW and 10–100-kW size ranges, PV systems using thin-film modules were more

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Installation Year

Average Size of

Systems (kW)

100

Average Size of >10 kW

Systems (kW)

≤10 kW (left axis)

>10 kW (right axis)

$10

$12

<2 kW

n=791

1 MW

2-5 kW

n=6349

22 MW

5-10 kW

n=4247

29 MW

10-30 kW

n=1010

14 MW

30-100 kW

n=297

15 MW

100-250 kW

n=147

24 MW

250-500 kW

n=109

37 MW

500-750 kW

n=31

18 MW

>750 kW

n=32

34 MW

System Size Range (kW)

Average Installed Cost (2008$/W)

Avg. +/- Std. Dev.

Systems Installed in 2008

costly, on average, than those with crystalline technology, with a difference of $0.8/W in the

<10-kW size range and a somewhat smaller cost differential ($0.3/W) in the 10–100-kW range.

In the >100-kW size range, average installed costs for the two types of systems differed by less

than $0.1/W.

Figure 3.15. Comparison of installed cost for residential retrofit vs. new construction

(Wiser et al. 2009)

Figure 3.16. Comparison of installed cost for crystalline vs. thin-film systems

(Wiser et al. 2009)

Figure 3.10 presents implied module and non-module costs, imputed from data on total installed

cost and a module price index. Figure 3.17, in contrast, presents actual data on module, inverter,

and “other” (e.g., mounting hardware and labor) costs reported to incentive program

$8.7

$9.8

$7.9

$7.4

$8.3

$10

$12

$14

All Systems

n=1506

3.6 MW

Rack-Mounted

n=1460

3.5 MW

BIPV

n=41

0.1 MW

All Systems

n=696

1.5 MW

Rack-Mounted

n=336

0.7 MW

BIPV

n=359

0.8 MW

Retrofit New Construction

Average Installed Cost

(2008$/W)

Avg. +/- Std. Dev

Note : The number of rack-mounted systems plus BIPV systems may not sum to the total number of systems, as some

systems could not be identified as either rack-mounted or BIPV.

California CSI, NSHP, and ERP Programs:

1-3 kW Systems Installed in 2008

$7.2

$7.3

$8.1

$7.8

$9.0

$8.2

$10

$12

Crystalline

n=9102

43 MW

Thin-Film

n=678

3.2 MW

Crystalline

n=1091

24 MW

Thin-Film

n=85

2.2 MW

Crystalline

n=242

86 MW

Thin-Film

n=23

7 MW

< 10 kW 10 - 100 kW > 100 kW

Average Installed Cost (2008$/W)

Avg. +/- Std. Dev.

Rack-Mounted Systems Installed in 2008

administrators, for a limited number of projects installed in 2008.

The magnitude of each

component and its relative contribution to total costs are similar for the two system size ranges

shown (<10 kW and 10–100 kW).

Modules, on average, represent slightly more than 50% of

total costs, inverters average just under 10% of total costs, and “other” costs make up the

remaining portion.

The component cost breakdown in Figure 3.17 conforms reasonably well to PV installer data

obtained by LBNL in 2008, in which installers shared data on the typical contributions of a

variety of specific cost components, as a percentage of total installed cost (Figure 3.18).

Figure 3.18 also presents the average percentage contribution for each cost component for three

separate PV markets (residential, small commercial, and large commercial).

Installers reported

that module costs typically represent approximately 50% of total installed cost, and inverters

represent 6%–7% of total costs, which is generally consistent with the results presented

previously. The results, however, provide a greater level of granularity in understanding the

composition of the remaining costs. In particular, the data indicate that labor and other materials

contribute roughly 10% each. The remaining cost components include overhead, profit, and

regulatory compliance (e.g., permitting, interconnection, rebate application), which represent a

notably greater percentage of total installed costs for residential systems compared with non-

residential systems.

Figure 3.17. Module, inverter, and other costs

(Wiser et al. 2009)

Of the 195 MW of 2008 PV installations in the LBNL dataset, incentive program administrators provided module

and inverter cost data for only 2.6 MW (1.3%), which form the underlying data for this figure.

Insufficient data were available for systems >100 kW to warrant inclusion in this figure.

In total, six installers provided data for residential and large commercial systems, and five installers provided data

for small commercial systems.

$4.5

$4.2

$0.6

$0.7

$3.2

$3.0

(52%)

(55%)

(8%)

(9%)

(40%)

(36%)

$10

< 10 kW

(n=2170)

10 - 100 kW

(n=145)

System Size Range

Average Cost (2008$/W)

Other

Inverter

Module

Systems Installed in 2008

Figure 3.18. PV installer data on component costs

(Wiser et al. 2009)

3.7 PV Operations and Maintenance

Operations and maintenance (O&M) is a significant contributor to the lifetime cost of PV

systems, and reducing the O&M costs of system components is an important avenue to reducing

lifetime PV cost. The data, however, are difficult to track, because O&M costs are not as well

documented as other PV system cost elements (which is due in part to the long-term and periodic

nature of O&M).

3.7.1 PV Operations and Maintenance Not Including Inverter Replacement

During the past decade, Sandia National Laboratories has collected O&M data for several types

of PV systems in conjunction with Arizona Public Service (APS) and Tucson Electric Power

(TEP) (Table 3.3). Because O&M data were collected for only 5–6 years in each study, data on

scheduled inverter replacement/rebuilding were not collected; inverters are typically replaced

every 7–10 years. Therefore, the information in Table 3.3 does not include O&M costs

associated with scheduled inverter replacement/rebuilding. This issue is discussed in the next

section of this report.

As shown in Table 3.3, annual O&M costs as a percentage of installed system cost ranged from

0.12% for utility-scale generation to 5%–6% for off-grid residential hybrid systems. The O&M

energy cost was calculated to be $0.004/kWh(ac) for utility-scale generation and $0.07/kWh(ac)

for grid-connected residential systems; note that this is simply annual O&M cost divided by

annual energy output, not LCOE. For all the grid-connected systems, inverters were the major

O&M issue. Following are brief summaries of four recent studies on O&M that provide

additional context.

48%

52%

11%

10%

29%

20%

21%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Residential

(3-5 kW)

Small Commercial

(10-50 kW)

Large Commercial

(>100 kW , roof-

mounted fixed-axis)

Percent of Total Installed Cost

Overhead, Regulatory

Compliance, Other

Installation Labor

Other Materials

Inverter

Module

A study by Moore and Post (2008) of grid-connected residential systems followed the experience

of TEP's SunShare PV hardware buy-down program. From July 2002 to October 2007, O&M

data were collected for 169 roof-mounted, fixed-tilt, crystalline silicon residential systems

smaller than 5 kW(dc) and with a single inverter in the Tucson area. A total of 330 maintenance

events were recorded: 300 scheduled and 30 unscheduled. The scheduled visits were credited

with minimizing unscheduled maintenance problems. Many of the unscheduled visits involved

replacing failed inverters that were covered under the manufacturer's warranty. The mean time

between services per system was 10.1 months of operation, with a maintenance cost of $226 per

system-year of operation.

A study by Moore et al. (2005) of grid-connected commercial systems followed the experience

of PV systems installed by APS. From 1998 to 2003, O&M data were collected for nine

crystalline silicon systems 90 kW(dc) or larger with horizontal tracking. Most of the O&M issues

were related to inverters, which required adjustments for up to 6 months after system installation,

after which the inverters generally performed well. Maintenance associated with the PV modules

was minimal. Maintenance associated with the tracking components was higher initially, but

became a small factor over time.

A study by Moore and Post (2007) of utility-scale systems followed the experience of large PV

systems installed at TEP's Springerville generating plant. From 2001 to 2006, O&M data were

collected for 26 135-kW(dc) crystalline silicon systems (all 26 systems were operational

beginning in 2004). The systems were installed in a standardized manner with identical array

field design, mounting hardware, electrical interconnection, and inverter unit. About half the

O&M costs were attributed to scheduled visits and half to unscheduled visits. Many of the

156 unscheduled visits were due to unusually severe lightning storms. The mean time between

unscheduled services per system was 7.7 months of operation.

A study by Canada et al. (2005) of off-grid residential hybrid systems followed the experience of

a PV system lease program offered by APS. From 1997 to 2002, O&M data were collected for

62 standardized PV hybrid systems with nominal outputs of 2.5, 5, 7.5, or 10 kWh/day and

including PV modules, a battery bank, an inverter and battery-charge controller, and a propane

generator. Because of the geographic dispersion of the systems, travel costs accounted for 42%

of unscheduled maintenance costs. Overall, O&M (including projected battery replacement at

6-year intervals) was calculated to constitute about half of the 25-year lifecycle cost of the PV

hybrid systems, with the other half attributed to initial cost.

Table 3.3. Summary of Arizona PV System O&M Studies, Not Including O&M Related to

Inverter Replacement/Rebuilding

System Type

(Reference)

O&M Data

Collection

Period

Scheduled O&M Unscheduled O&M

Annual O&M

Cost as

Percentage of

Installed

System Cost

O&M

Energy

Cost

Grid-Connected

Residential,

Fixed Tilt

(Moore and

Post 2008)

2002–2007

Visits by category:

general

maintenance/inspection

(45%), pre-acceptance

checks required for

SunShare program

(55%)

Visits by category:

inverter (90%), PV

array (10%)

1.47%

$0.07/

kWh

(ac)

Grid-Connected

Commercial,

Horizontal

Tracking

(Moore et al.

2005)

1998–2003

Inverters were the primary maintenance issue;

most systems required inverter adjustments

during initial setup for up to 6 months after

installation, after which the inverters generally

performed well. Minimal maintenance was

associated with modules. Maintenance for

tracking components started higher during

early part of development effort, but decreased

over time.

0.35%

Not

Reported

Utility-Scale

Generation,

Fixed Tilt

(Moore and

Post 2007)

2001–2006

Mowing native

vegetation, visually

inspecting arrays and

power-handling

equipment

Costs by category:

inverter (59%), data

acquisition systems

(14%), AC

disconnects (12%),

system (6%), PV

(6%), module

junction (3%).

0.12%

$0.004/

kWh

(ac)

Off-Grid

Residential

Hybrid (Canada

et al. 2005)

1997–2002

Quarterly generator

service (oil change,

filter, adjustment, and

inspection), battery

inspection and service,

inverter inspection,

overall system

inspection;

repairs/replacements

made when problems

noted.

Costs by category:

system setup,

modification, and

removal (41.4%);

generator (27.8%);

inverter (16.5%);

batteries (4.7%);

controls (4.2%);

PV modules (2.7%);

system electrical

(2.6%).

5%–6%

Not

Reported

Annual O&M cost divided by annual energy output, not LCOE.

This cost was calculated for the 4-year period 1999–2002 because O&M costs stabilized about 2 years into the program; the

costs included battery service, but not battery replacement.

3.7.2 PV Inverter Replacement and Warranty Trends

Inverters have become a central component in the solar industry owing to the ever-growing grid-

connected PV market. Although much attention is given to increasing inverter efficiencies,

inverter reliability has a greater impact on lifetime PV system cost, which makes it an important

factor in market adoption. In the study of TEP's utility-scale PV described above,

replacing/rebuilding inverters every 10 years was projected to almost double annual O&M costs

by adding an equivalent of 0.1% of the installed system cost, bringing total annual O&M cost to

0.22% of installed system cost (Moore and Post 2007). Similarly, the O&M energy cost was

projected to increase by $0.003/kWh(ac), resulting in a total O&M energy cost of

$0.007/kWh(ac); note that this is simply annual O&M cost divided by annual energy output, not

LCOE.

Inverter cost accounts for about 6%–9% of a PV system's initial installed cost (see Section 3.6).

Current inverters have an average lifespan of about 7 to 10 years, meaning that they might have

to be replaced two to three times over the lifetime of a PV system. The warranty that a

manufacturer is willing to provide is a good indication of an inverter’s reliability.

Figure 3.19 illustrates default inverter warranty data. As inverter reliabilities increase,

manufacturers have started to offer longer warranties. Today, a majority of manufacturers are

comfortable giving default 5-year warranties as opposed to 1–3 year warranties as was the case

5 years ago. In addition, a growing number of manufacturers have begun offering customers

optional warranties with up to 10 years of coverage for an additional fee. This suggests that

inverter companies are becoming increasingly confident in the reliability of their products.

Figure 3.19. Inverter default warranties, 2002–2008

(Knoll and Kreutzmann 2008, Photon International 2002–2008)

2002

2003

2004

2005

2006

2007

2008

Percent of Manufacturers

0 years

1 - 3 years

5 years

>5 years

3.8 CSP Installation and Operations and Maintenance Cost Trends

The average cost, after federal incentives, for a CSP plant without storage is greater than

$4,000/kW in the United States (Bullard et al. 2008). For example, investment for construction

and associated costs for the 64 MW Nevada Solar One plant amounted to $260 million or about

$4,100/kW (Acciona Energy 2008). System developers strongly believe that improvements in

design will reduce this cost considerably, making it more competitive with traditional electricity

sources.

Figure 3.20 shows a typical cost breakdown for components of a parabolic trough system, with

the solar field comprising approximately 50% of the total installed cost; solar field components

include the receivers, mirrors (reflectors), structural support, drivers, and foundation. Receivers

and mirrors each contribute approximately 10% to the total. The power block, which is not

considered part of the solar field, normally has the highest cost of all the major components,

contributing roughly 20% to the total (Pitz-Paal et al. 2005).

Figure 3.20. Generic parabolic trough CSP cost breakdown

(Pitz-Paal et al. 2005)

3.9 CSP Technology Characteristics and System Performance

Four types of CSP technology are currently under development: parabolic trough technology,

power tower technology, dish-engine technology, and linear Fresnel reflector technology.

3.9.1 Parabolic Trough Technology

Trough technology uses one-axis tracking, has a concentration ratio of 80 (concentration ratio is

calculated by dividing reflector area by focal area), and achieves a maximum temperature of

about 400°C. This relatively low temperature limits potential efficiency gains and is more

susceptible to performance loss when dry cooling is used. Moreover, the relatively low operating

Contigencies

(Indirect)

17%

Storage

Power Block

22%

Others

Land

Foundation

Drivers

Mirrors

10%

Structure

12%

Receiver

11%

Instrumentation

and Controls

Connecting

Cables

Piping of

Headers

temperature makes it very difficult to provide the amount of heat storage (in a cost-effective

manner) that is required for around-the-clock dispatchability (Grama et al. 2008, Emerging

Energy Research 2007).

3.9.2 Power Tower Technology

Receiver technology uses two-axis tracking, has a concentration ratio up to 1,500, and achieves a

maximum temperature of about 650°C (Grama et al. 2008). The higher operating temperature of

tower technology reduces susceptibility of these systems to efficiency losses, especially when

dry cooling is used. The reflectors, called heliostats, comprise 40% of capital costs (Emerging

Energy Research 2007).

3.9.3 Dish-Engine Technology

Dish-engine technology uses two-axis tracking, has a concentration ratio up to 1,500, and

achieves a maximum temperature of about 700°C (Emerging Energy Research 2007). This

technology set the world record for solar thermal conversion efficiency, achieving 31.25%

(Andraka 2008).

3.9.4 Linear Fresnel Reflector Technology

Linear Fresnel reflector technology uses one-axis tracking, has a concentration ratio of 80, and

achieves a maximum temperature of about 400°C. The reduced efficiency (15% to 25%)

compared to troughs is expected to be offset by lower capital costs (Grama et al. 2008, Emerging

Energy Research 2007).

3.9.5 Storage

Thermal energy storage (TES) has the potential to extend CSP production time up to 16 hours

per day, increasing capacity factor to more than 50% and allowing for greater dispatchability.

Although capital expenditure increases when storage is added, the LCOE will most likely

decrease because of the increased capacity factor and greater utilization of the power block

(NREL 2009c). Moreover, storage increases the technology’s marketability, as utilities can

dispatch the electricity to meet non-peak demand. For example, a molten salt mixture of 60%

sodium nitrate and 40% potassium nitrate is used as the storage medium for the 50-MW

Andasol I plant located in Spain, enabling more than 7 hours of additional electricity production

after direct-normal insolation is no longer available. Various mixtures of molten salt are being

investigated to optimize the storage capacity, and research is being conducted on other mediums

such as phase-change materials. Synthetic mineral oil, which has been the historical heat transfer

fluid used in CSP systems, is also being viewed as a potential storage medium for future systems.

3.9.6 Heat-Transfer Fluid

Improvements in the heat-transfer fluid (HTF) are necessary to bring down the LCOE for CSP.

This can be accomplished by lowering the melting points and increasing the vapor pressure of

these substances. Dow Chemical’s and Solutia’s synthetic mineral oils have been used widely as

the HTF in trough systems. The problem with these synthetic oils is that they break down at

higher temperatures, preventing the power block from operating at higher, more efficient

temperatures. Molten salts can withstand higher temperatures than the currently available

synthetic oils and are being considered as an HTF. The downside is that they freeze at a higher

temperature than the synthetic oils, which means a drop in temperature during the night may

solidify the substance. This, in turn, can damage the equipment when the salt expands and puts

pressure on the receivers. Corrosion of the receivers is another potential concern when salts are

introduced. Nonetheless, research is being conducted to use this substance as both an HTF and

TES medium. If this can be accomplished, costly heat exchangers would not be needed, thus

helping to reduce the LCOE.

3.9.7 Water Use

As stated previously in Section 2.4.4, a water-cooled parabolic trough plant typically requires

approximately 800 gallons per MWh. Power towers operate at a higher temperature and have

lower water cooling needs, ranging from 500–750 gallons per MWh. Dish-engine systems do not

require water cooling. As CSP plants are usually constructed in dry regions, water-scarcity issues

and competing uses are of concern (except for dish-engine systems). An alternative to water

cooling is dry or air cooling, which eliminates about 90% of water consumption (U.S. DOE

2009). However, air cooling requires higher upfront capital costs and can result in a 5% decrease

in electricity generation, depending on location temperature. This plant-efficiency reduction

amounts to a 2%–9% increase in LCOE . An alternative is to implement hybrid cooling, which

decreases water use while minimizing the generation losses experienced with dry cooling.

3.9.8 Land Requirements

The amount of acreage needed for a CSP facility depends partly on the type of technology

deployed. More importantly, though, land use is dependent on thermal storage hours and a

location’s solar insolation. Common practice is to state land requirements in terms of acres per

MW. The range normally provided is 4–8 acres in a location with solar insolation similar to that

found in the U.S. desert Southwest (SNL 2009). The low end of the range is possible when

greater self-shading of reflectors is allowed, although this results in reduced electricity output.

The high end represents the additional land needed for energy storage, with energy storage

resulting in higher capacity factor. Because of such variation, when considering land needs, it

can be more useful to provide a number in terms of acres per MWh. When this is done, a

comparison among CSP technology types can more easily be made. The general trend at this

stage of technology development is that power towers require approximately 20% more land per

MWh than troughs. Commercial dish-engine facilities have not yet been built, so a comparison

of land requirements for this technology with other CSP technologies has not yet been made.

3.10 References

Acciona Energy. (2008). www.acciona-na.com. Accessed November 2008.

Andraka, C. E. (July 2, 2008). Statement of Charles E. Andraka, Sandia National Laboratories.

Concentrating Solar Power. United States Senate Committee on Energy and Natural Resources

Field Hearing in Albuquerque, New Mexico.

Bartlett, J.E.; Margolis, R.M.; Jennings, C.E. (September 2009). The Effects of the Financial

Crisis on Photovoltaics: An Analysis of Changes in Market Forecasts from 2008 to 2009.

National Renewable Energy Laboratory. Technical Report NREL/TP-6A2-46713.

http://www.nrel.gov/docs/fy10osti/46713.pdf. Accessed September 2009.

Bullard, N.; Chase. J.; d’Avack, F. (May 20, 2008). “The STEG Revolution Revisited.” Research

Note. London: New Energy Finance.

Canada, S.; Moore, L.; Post, H.; Strachan, J. (2005). "Operation and Maintenance Field

Experience for Off-grid Residential Photovoltaic Systems." Prog. Photovolt: Res. Appl.; 13:67–

74.

Denholm, P.; Margolis, R. (2008a). Supply Curves for Rooftop Solar PV Generated Electricity

for the U.S. NREL Report TP-670-44073. Golden, CO: National Renewable Energy Laboratory.

Denholm, P.; Margolis, R. (2008b). “Impacts of Array Configuration on Land-Use Requirements

for Large-Scale Photovoltaic Deployment in the United States.” NREL Conference Paper

CP-670-42971. Golden, CO: National Renewable Energy Laboratory.

Denholm, P.; Margolis, R. (2008c). “Land-Use Requirements and the Per-Capita Solar Footprint

for Photovoltaic Generation in the United States.” Energy Policy; 36:3531–3543.

Emerging Energy Research, LLC. (2007). Global Concentrated Solar Power Markets and

Strategies, 2007–2020. Cambridge, MA: Emerging Energy Research.

Grama, S.; Wayman, E.; Bradford, T. (2008). Concentrating Solar Power—Technology, Cost,

and Markets. 2008 Industry Report. Cambridge, MA: Prometheus Institute for Sustainable

Development and Greentech Media.

Kazmerski, L. (November 2009). “Best Research-Cell Efficiencies.” Golden, CO: National

Renewable Energy Laboratory.

Knoll, B.; Kreutzmann, A. (April 2008). “Market survey on on-grid inverters 2008.” Photon

International, pp.104–111.

Kreutzmann, A. (February 2008). “Solar modules in 2008.” Photon International, pp.126–135.

Mehos, M.; Kearney, D. (2007). “Potential Carbon Emissions Reductions from Concentrating

Solar Power by 2030.” Kutscher, C. ed. Tackling Climate Change in the U.S.: Potential Carbon

Emissions Reductions from Energy Efficiency and Renewable Energy by 2030. NREL Report

CH-550-41270. Boulder, CO: American Solar Energy Society, pp. 79–89.

www.ases.org/climatechange/

. Accessed June 2009.

Mehta, S.; Bradford, T.; (2009). PV Technology, Production, and Cost, 2009 Forecast: The

Anatomy of a Shakeout. Prometheus Institute and Greentech Media.

Mints, P.; Tomlinson, D. (2006). Photovoltaic Manufacturer Shipments 2005/2006. Report #

NPS-Supply1. Palo Alto, CA: Navigant Consulting Photovoltaic Service Program.

Mints, P.; Tomlinson, D. (2008). Photovoltaic Manufacturer Shipments & Competitive Analysis

2007/2008. Report # NPS-Supply3. Palo Alto, CA: Navigant Consulting Photovoltaic Service

Program.

Mints, P. (2009). Photovoltaic Manufacturer Shipments, Capacity, & Competitive Analysis

2008/2009. Report # NPS-Supply4. Palo Alto, CA: Navigant Consulting Photovoltaic Service

Program.

Moore, L.; Post, H.; Hayden, H.; Canada, S.; Narang, D. (2005). "Photovoltaic Power Plant

Experience at Arizona Public Service: A 5-year Assessment." Prog. Photovolt: Res. Appl.;

online.

Moore, L.M.; Post, H.N. (2007). "Five Years of Operating Experience at a Large, Utility-Scale

Photovoltaic Generating Plant." Prog. Photovolt: Res. Appl.; online.

Moore, L.M.; Post, H.N. (2008). "Five Years of Residential Photovoltaic System Experience at

Tucson Electric Power." Strategic Planning for Energy and the Environment; 28:2, pp. 58–73.

NREL. (2009a). LCOE analysis for U.S. cities. Golden, CO: National Renewable Energy

Laboratory (internal only).

NREL. (2009b). PVWatts: A Performance Calculator for Grid-Connected PV Systems.

Version 1. Golden, CO: National Renewable Energy Laboratory.

http://rredc.nrel.gov/solar/codes_algs/PVWATTS/version1/. Accessed August 2009.

NREL. (2009c). Solar Advisor Model (SAM). Version 3.0. Modeling runs conducted for troughs

and towers in July 2009. Golden, CO: National Renewable Energy Laboratory (internal only).

NREL. (2009d). Photovoltaic Solar Resource: United States – Spain – Germany. Produced by

Billy Roberts of the NREL Geographic Information System (GIS) team.

Photon International. (1999–2008). Annual market surveys for solar modules.

Photon International. (2002–2008). Annual market surveys for on-grid inverters.

Pitz-Paal, R.; Dersch, J.; Milow, B., eds. (February 16, 2005). European Concentrated Solar

Thermal Road-Mapping (ECOSTAR). Roadmap Document. European Union: European

Commission. http://www.solarpaces.org/Library/library.htm. Accessed November 20, 2009.

Renewable Energy Policy Network for the 21

Century (REN21). (2008). Renewables 2007

Global Status Report 2007. Paris: REN21 Secretariat. www.ren21.net/publications

. Accessed

February 2009.

Sandia National Laboratories (SNL). (September 2009). E-mail correspondence with CSP

technical staff. Albuquerque, NM: Sandia National Laboratories.

U.S. Bureau of Economic Analysis (BEA). (2009). National Income and Product Accounts

(NIPA): GDP Implicit Price Deflator Data. http://www.econstats.com/nipa/NIPA1_1_1_4_.htm.

Published August 20, 2009.

U.S. Department of Energy (DOE). (2009). Report to Congress. Concentrating Solar Power

Commercial Application Study: Reducing Water Consumption of Concentrating Solar Power

Electricity Generation.

www.eere.energy.gov/solar/pdfs/csp_water_study.pdf. Accessed October

2009.

Wiser, R.; Barbose, G.; Peterman, C.; Darghouth, N. (2009). Tracking the Sun II. The Installed

Cost of Photovoltaics in the U.S. from 1998–2008. Berkeley, CA: Lawrence Berkeley National

Laboratory.

http://eetd.lbl.gov/ea/EMS/re-pubs.html. Accessed October 2009.

Wohlgemuth, J.H.; Cunningham, D.W.; Nguyen A.M.; Miller J. (2006). Long Term Reliability of

PV Modules. Frederick, MD: BP Solar.

4. Policy and Other Market Drivers

This chapter covers key elements of U.S. federal, state, and local policies pertaining to solar

energy technologies, as well as market-based developments that affect U.S. solar market

evolution. Section 4.1 discusses federal policies, incentives, and programs including tax credits,

depreciation benefits, grants, the DOE loan guarantee program, clean renewable energy bonds,

and other federal programs and incentives. Section 4.2 discusses state and local policies and

incentives, and rules and regulations including permitting, interconnection, net metering, direct

cash incentive programs, renewable portfolio standards and solar set-asides, and clean energy

funds. Section 4.3 provides information on major financing mechanisms and programs: third-

party power purchase agreement financing, customer solar lease financing, property-assessed

clean energy programs, and other emerging financing structures.

4.1. Federal Policies and Incentives, PV and CSP

Federal policies and incentives play an important role in the commercialization and adoption of

solar technologies; they have enabled rapid expansion of solar markets in countries such as

Germany, Spain, and Japan. Legislation enacted in the United States in 2008 and early 2009

provides unprecedented levels of federal support for U.S. renewable energy projects, including

solar energy projects.

The Emergency Economic Stabilization Act of 2008 (EESA or “bailout bill”) became law on

October 3, 2008. It contains tax incentives designed to encourage individuals and businesses to

invest in renewable energy, including 8-year extensions of the business and residential solar

investment tax credits (ITCs).

The American Recovery and Reinvestment Act (ARRA or “stimulus bill”) was signed into law

on February 17, 2009, with an estimated $787 billion overall in tax incentives and spending

programs. Many ARRA provisions support solar energy.

This section discusses the major U.S. federal policies and incentives directed toward solar

energy, with an emphasis on provisions in the EESA and ARRA. For additional information,

including how to apply for the benefits of the policies, see the list of Web sites at the end of this

section.

4.1.1 Investment Tax Credit

Sections 48 (for businesses) and 25D (for residences) of the Internal Revenue Code detail the

federal ITC for certain types of energy projects, including “equipment which uses solar energy to

generate electricity." Like other tax credits, the ITC reduces the tax burden of individuals and

commercial entities that make investments in solar energy technology. On an industry level, a

long-term ITC provides consistent financial support for growth such as building manufacturing

plants, developing an installer workforce, and investing in large-scale solar electric plants that

require extended planning and construction time.

For commercial projects, the ITC is realized in the year in which the solar project begins

commercial operations, but vests linearly over a 5-year period (i.e., one-fifth of the 30% credit

vests each year over a 5-year period). Thus, if the project owner sells the project before the end

of the fifth year since the start of commercial operations, the unvested portion of the credit will

be recaptured by the IRS. This period is sometimes referred to as the 5-year “clawback” period.

The EESA extended the ITC through December 31, 2016, including extensions of the

commercial and residential solar ITCs, providing a credit of up to 30% of the total capital costs

of a project (including equipment and labor) (SEIA 2008).

4.1.2 Renewable Energy Grants

The EESA removed the cap on the

ITC for residential PV systems (previously $2,000), effective for property placed in service after

December 31, 2008. The bill also allows individual taxpayers to use the credit to offset

alternative minimum tax liability. Another change to the ITC was to allow regulated utilities to

claim the tax credit, providing significant support for increased utility investment in solar energy

projects. The ARRA enhanced the ITC further by allowing individuals and businesses to qualify

for the full amount of the solar tax credit, even if projects receive subsidized energy financing.

Previously, the ITC would not apply to the portion of the investment funded via subsidized

financing such as below-market loans. Also, the ARRA removed the $2,000 cap on the ITC for

residential solar water heating systems.

Section 1603 of the ARRA authorizes the Department of the Treasury to issue renewable energy

project developers cash grants in lieu of the ITC. The grants program was created in response to

the lack of available financing and limited appetite for tax credits resulting from the financial

crisis and economic downturn. The program is designed like the ITC and offers an equivalent

30% benefit based on the cost of the solar property that is placed in service.

Grants are available for qualifying property that is placed in service during 2009 and 2010. Solar

projects that commence construction by December 31, 2010, and are placed in service prior to

2017 also qualify. Developers must apply for the grant by September 30, 2011, and only

taxpaying corporate entities are eligible. Grant applications will be processed within 60 days

from the date it is received or the system is placed in service, whichever is later.

There are several variables that determine whether a developer might opt to apply for a cash

grant. These factors may include state and local incentives and mandates, project scale and

required lead time for development, and the ability to monetize tax credits. The Treasury

Department began accepting applications for grants on July 31, 2009, and the first payments

were announced on September 1. By the end of November 2009, 65 solar projects had received

funds, with allocations totaling more than $18 million.

More information is available online: http://www.treasury.gov/recovery/1603.shtml

Historically, through 2005, the size of the commercial solar credit was equal to 10% of the project’s “tax credit basis,” the

portion of system costs to which the ITC applies. The Energy Policy Act of 2005 temporarily increased the solar credit to 30% of

a project’s tax credit basis, for projects placed in service between January 1, 2006, and January 1, 2008. In late December 2006,

the Tax Relief and Healthcare Act of 2006 extended the in-service deadline to December 31, 2008, and in October 2008, the

EESA extended it once again for a full 8 years, through December 31, 2016. Unless extended again or otherwise altered over the

next 8 years, the Section 48 commercial solar credit will revert back to 10% for projects placed in service on January 1, 2017.

4.1.3 Manufacturing Tax Credit

The ARRA created a tax credit for new investments in advanced energy manufacturing that,

similar to the ITC, is equal to 30% of the investment. Eligible technologies include renewable

energy, energy conservation, electric grids supporting intermittent sources of renewable energy,

carbon capture and storage, biofuel refining or blending, and hybrid-electric vehicles and

components. The cap for new manufacturing investment credits is $2.3 billion, supporting

$7.7 billion of manufacturing capital investment.

In determining which projects receive the tax

credits, the Treasury Department, in coordination with the Department of Energy, will consider

the following factors: commercial viability, job creation, greenhouse gas impact, technological

innovation and cost reduction, and time to completion. More information is available online:

http://www.energy.gov/recovery/48C.htm

4.1.4 MACRS and Bonus Depreciation

Section 168 of the tax code allows investors to depreciate certain investments in solar power

(and other types of) projects using a 5-year accelerated depreciation schedule known as the

Modified Accelerated Cost Recovery System (MACRS). Under this provision, “equipment

which uses solar energy to generate electricity” qualifies for 5-year, double declining-balance

depreciation. In most cases, 100% of a solar project’s cost will qualify for this accelerated

schedule, but the 30% ITC will reduce the project’s depreciable basis by 15%. Assuming a 40%

combined effective state and federal tax bracket and a 10% nominal discount rate, on a present-

value basis, this 5-year MACRS depreciation schedule provides a tax benefit equal to about 26%

of system costs (Bolinger 2009).

In addition to the standard MACRS, the EESA includes a bonus depreciation schedule for solar

projects installed in 2008. Qualifying projects can receive 50% depreciation in the first year, with

the remaining 50% depreciated over the 5-year MACRS schedule. The ARRA extends the 50%

year-one bonus depreciation incentive for qualified renewable energy investments made through

2009. The ARRA also extends through 2009 a recent increase in the size of the write-off

available (up to 100% of a $250,000 investment, a declining percentage after $250,000, and

phasing out at $800,000).

Taken together, the 30% ITC and accelerated depreciation

provide a combined tax benefit equal to about 56% of the installed cost of a commercial solar

system (Bolinger 2009).

4.1.5 Renewable Energy Loan Guarantee Program

The DOE loan guarantee program established by Title XVII of the Energy Policy Act of 2005

was expanded by the ARRA to include a new Section 1705 loan guarantee program, in addition

to the existing Section 1703 program. The ARRA permitted the guarantee of about $40 billion of

loans by the Section 1705 program, in addition to the $51 billion authorized for Section 1703.

Table 4.1 summarizes the loan guarantee programs, including differences pertaining to project

eligibility and benefits.

Of the $2.3 billion in total awards to 183 projects, more than $1.0 billion went to 60 solar projects. See:

http://www.whitehouse.gov/the-press-office/president-obama-awards-23-billion-new-clean-tech-manufacturing-jobs

Only 12% of this benefit is attributable to the acceleration of the depreciation schedule; the remaining 14% would

be realized even if the project were instead depreciated using a less-advantageous, 20-year straight-line schedule.

Table 4.1. DOE Loan Guarantee Programs

Year

FY 2007

FY 2008

FY 2009 Omnibus

FY 2009 ARRA

Amount

$4.0 billion

$38.5 billion

$8.5 billion

$40 billion (estimated)

Authorization

EPACT 2005, Title XVII, Section 1703

EPACT Section 1705,

added by ARRA

Uses

New or significantly improved technologies

Commercial and novel

technologies

Credit Subsidy

Borrower pays

$4.0 billion appropriated

Term

Available until used

Projects must be started by

September 30, 2011

Carve-outs

No carve-out

stipulated by

Congress

• $10.0 billion for

energy efficiency,

renewable energy, and

advanced transmission

and distribution

technologies

• $18.5 billion for

advanced nuclear

power facilities

• $2.0 billion for “front

end” nuclear fuel cycle

facilities

• $6.0 billion for coal-

based power

generation, industrial

gasification, and

carbon capture and

sequestration

• $2.0 billion for

advanced coal

gasification

The FY 2009

Omnibus Budget

provides an

additional

$8.5 billion in loan

authority for

energy efficiency,

renewable energy,

and advanced

transmission and

distribution

projects

No carve-outs were

stipulated, but three project

categories were listed:

• Renewable energy

installations and

manufacturing facilities

for renewable energy

components

• Electric power

transmission systems

• Advanced biofuel

projects

U.S. DOE 2009c

Projects eligible for the Section 1703 program include those that “avoid, reduce or sequester air

pollutants or anthropogenic emissions of greenhouse gases; and employ new or significantly

improved technologies as compared to commercial technologies,” including energy efficiency,

renewable energy, and advanced transmission and distribution as well as advanced nuclear

power, advanced coal-based power, and carbon capture and sequestration technologies.

Section 1705 is limited to renewable energy installations and manufacturing facilities for

renewable energy components, electric power transmission systems, and advanced biofuel

projects and is targeted toward projects at the commercialization stage (though new or earlier

stage technologies are still eligible). The Section 1705 program requires projects to commence

construction by September 30, 2011, encouraging near-term deployment. Another difference is

that the Section 1705 program provides for DOE to pay the cost of credit subsidies, required up-

front payments equal to about 10% of a loan guarantee's value, up to a total of $4 billion.

There are currently two open solicitations, released by DOE in July 2009. The first solicitation is

for "new or significantly improved" energy efficiency, renewable energy, and advanced

transmission and distribution technologies. It combined the authorities of the Section 1703 and

The FY 2009 ARRA appropriation for the credit subsidy was originally $6 billion; however, $2 billion were

transferred to the Car Allowance Rebate System (also known as the "Cash for Clunkers" program).

1705 programs; projects eligible for 1703 but not 1705 may still secure a loan guarantee, but

may not receive 1705 appropriations to cover the credit subsidy cost. The $2.5 billion in

1705 appropriations allocated to this solicitation include $2 billion for renewable energy and

transmission projects and $0.5 billion for advanced biofuels.

The second solicitation is for large transmission infrastructure projects using commercial

technologies. The authority for the second solicitation is only under Section 1705, so all projects

under this solicitation must commence construction by September 30, 2011. The second

solicitation was allocated $750 million from Section 1705 appropriations to cover the credit

subsidy fees. Combined, these two solicitations are estimated to result in about $30 billion in

loan guarantees.

The first of several loan guarantees was conditionally awarded on March 20, 2009, under the

1703 solicitation and closed on September 4, 2009, in the amount of $535 million, for Solyndra,

Inc., based in Freemont, California. The funds will support the construction of a manufacturing

plant for its proprietary cylindrical solar PV panels, which are expected to provide systems with

low installed cost and high solar electricity output (U.S. DOE 2009d).

For more information on DOE loan guarantee program solicitations, see

http://www.lgprogram.energy.gov/keydocs.html

4.1.6 Clean Renewable Energy Bonds

Clean renewable energy bonds (CREBs) were established by EPACT 2005 to provide renewable

project financing for non-taxable entities (governmental entities, electric cooperatives, and public

power providers) that cannot directly use the ITC for solar facilities, production tax credits for

other types of renewable energy facilities, or accelerated tax-depreciation benefits. CREBs are

“tax credit bonds,” which means that the bond purchaser receives a federal income tax credit in

lieu of interest payments. From the borrower’s perspective, CREBs are essentially the equivalent

of a zero-interest loan, aside from the various transaction costs of bond issuance described

below, which reportedly can be considerable (Cory et al. 2008).

The CREB program received an initial allocation of $800 million in 2005 (round 1), which was

then increased to $1.2 billion by legislation in 2006, providing a second allocation of about

$400 million (round 2). The round 1 CREB allocation of $800 million was awarded in 2006. The

round 2 CREB allocation of $477 million was awarded in 2008 (the extra $77 million was due to

surrendered volume from the first allocation). Of the $1.2 billion CREB allocations, state and

local governments were limited to $750 million or 62.5% of allocations, with the rest intended

for municipal and cooperative electric companies (DSIRE 2010a).

The Energy Improvement and Extension Act of 2008 authorized $800 million and the ARRA

authorized an additional $1.6 billion for new CREBs, for a total allocation of $2.4 billion

(round 3). In April 2009, the IRS opened a solicitation for the $2.4 billion allocation, which

closed on August 4, 2009. In October 2009, $2.2 billion of CREBs applications were given

issuing authority by the IRS for a period of 3 years. Round 3 CREBS funding was to be allocated

as follows: one-third for qualifying projects of state/local/tribal governments, one-third for

public power providers, and one-third for electric cooperatives (DSIRE 2010a).

Figure 4.1. Distribution of round 1 and 2 CREB

allocation, percent of projects by technology

(IRS 2006, IRS 2007)

Figure 4.2. Distribution of round 3 CREB

allocation, percent of projects by technology

(IRS 2009)

Figures 4.1 and 4.2 capture the distribution of CREB allocations by technology. By number of

projects, solar accounted for 62% of round 1 and 2 (year 2006-2008) allocations and 89% of

round 3 (2009) allocations (IRS 2006, IRS 2007, IRS 2009).

By funding amount, solar

accounted for 21% or $84 million of round 2 allocations and 38% or $839 million of round 3

allocations (IRS 2007, IRS 2009).

4.1.7 Solar on Federal Property

The EESA and ARRA together appropriated $5.5 billion to the Federal Buildings Fund for green

building improvements including energy efficiency measures and the use of renewable energy

sources such as solar. Of the total appropriation, $4.5 billion will be available to federal facilities

managed by the General Services Administration (Cory et al. 2009, Prometheus 2009).

4.1.8 State Energy Program

Funding from DOE’s State Energy Program (SEP) goes to state energy offices in all states and

U.S. territories. Under ARRA, SEP will distribute $3.1 billion to the states (U.S. DOE 2009e).

Activities eligible for SEP funding include energy audits, building retrofits, education and

training efforts, and new financing mechanisms to promote renewable energy investments. Many

of the state governments will use part of these funds to support solar programs and installations.

For every dollar of federal investment in SEP, the states estimate saving $7.22 in energy costs

and also leveraging $10 of non-federal investment in energy projects, typically by cosponsoring

energy projects with local and private partners (U.S. DOE 2009e). For more information on the

DOE State Energy Program, see: http://apps1.eere.energy.gov/state_energy_program/.

The percentages for round 2 are based on data representing $405 million in initial funding for round 2.

Detailed funding amounts for round 1 are not available. The percentages for round 2 are based on data

representing $405 million in initial funding for round 2.

Solar

62%

Wind

23%

Landfill

Gas

8.8%

Biomass

2.8%

Hydro-

power

2.6%

Other

< 1%

Solar

89%

Wind

5.9%

Biomass

1.9%

Hydro-

power

3.2%

Other

< 1%

4.1.9 Energy Efficiency and Conservation Block Grant Program

The Energy Efficiency and Conservation Block Grants (EECBG) Program, authorized in the

Energy Independence and Security Act of 2007, was funded for the first time by the ARRA

(DOE 2009a). Through formula

and competitive grants, the U.S. DOE will distribute

$3.2 billion to U.S. cities, counties, states, territories, and Indian tribes to develop, implement,

and manage energy efficiency and conservation projects and programs (U.S. DOE 2009b). Local

and state governments may utilize funds for solar installations on government buildings and

engage in energy strategy development, which may include solar energy technology, along with

energy efficiency and conservation. For more information on the DOE EECBG, see:

www.eecbg.energy.gov.

4.1.10 Renewable Energy Production Incentive

The Renewable Energy Production Incentive (REPI), a federal incentive for renewable energy

systems, was implemented as part of the Energy Policy Act (EPACT) of 1992 and amended in

EPACT 2005 to expand eligible facilities and authorize federal appropriations through 2026.

Qualified facilities are categorized as Tier 1 (solar, wind, ocean, geothermal, closed-loop

biomass) and Tier 2 (open-loop biomass such as landfill gas and livestock methane; municipal

solid waste is excluded). The program originated to support facilities owned by public utilities

and other nonprofit load-serving entities (LSEs), providing per kWh-based payments for the first

10 years of operation. The actual payment amount is dependent on availability of federal budget

appropriations each year (U.S. DOE 2008). Historically, the REPI program has been

insufficiently funded and has not made full payments for the electricity generated by qualifying

facilities since 1995.

4.1.11 Additional Resources

With continued uncertainty over the funding appropriation from year to

year, the REPI program is likely to serve only as a supplementary incentive program.

For additional information, including how to apply for the benefits of the policies, see the

following Web sites:

• DOE Solar Energy Technologies Program Financial Opportunities

(www.eere.energy.gov/solar/financial_opportunities.html)

•

DOE Energy Efficiency & Renewable Energy (EERE) Financial Opportunities

(www.eere.energy.gov/financing/)

• DOE EERE Recovery Act Web site (www.eere.energy.gov/recovery/)

• U.S. Department of the Treasury, American Recovery and Reinvestment Act

(www.treasury.gov/recovery)

More information on the formula methodology for EECBG can be found at:

http://www.eecbg.energy.gov/downloads/EECBG_Federal_Register_Notice_04_15_09.pdf.

For production year 2007, DOE paid out only $2.7 million of the $11.25 million in eligible funds to Tier 1

projects and $1.8 million of the nearly $8 million in eligible funds to Tier 2 projects (U.S. DOE 2008). Only a small

fraction of these funds have gone to solar projects; the average payment was about $900 for each of the 25 solar

projects in 2007 (Cory et al. 2008).

• U.S. Internal Revenue Service (IRS), Energy Provisions of the American Recovery and

Reinvestment Act of 2009 (www.irs.gov/newsroom/article/0,,id=206871,00.html)

• Solar Energy Industries Association (SEIA), Government Affairs & Advocacy

(www.seia.org/cs/government_affairs_and_advocacy)

• Database of State Incentives for Renewables & Efficiency (www.dsireusa.org)

4.2 State and Local Policies, Incentives, and Rules and Regulations

As outlined in the previous section, there are a number of federal-level financial incentives

available to support the increased deployment of solar energy technologies. State and local

policies work in parallel with these federal-level initiatives, but go beyond financial incentives to

include renewable energy mandates and other mechanisms to further stimulate adoption of solar

energy technologies. State legislators and utility commissioners hold primary responsibility for

setting a state’s overarching energy policy and regulatory framework. How solar technologies

are treated in this process will significantly affect how, or even if, a solar market develops in a

state. Local governments also influence solar policies. In areas where the local government has

jurisdiction over a utility, the government can directly influence solar rebate programs and

renewable generation requirements. In areas served by investor-owned utilities or cooperatives,

local governments can still play an important role in solar market development by streamlining

permitting processes or developing innovative financing mechanisms.

State and local policies in support of increased solar deployment are more prevalent than federal

policies and have a well-established history of both successes and failures. As such, states and

regions with stronger and longer-term policies and incentives, coupled with a favorable

electricity market (e.g., higher than average electricity prices) and an adequate solar resource,

have established pockets of wide-scale solar installations. In addition, because states are often

innovation hubs, there is a continuous flow of new policies and approaches to driving solar

development that bears watching.

4.2.1 Planning and Permitting

Planning is an effort by governments to ensure community land and resources are used in a

beneficial manner. Permits are allowances issued by governments to ensure that activities

undertaken within their jurisdictions meet established guidelines. Planning and permitting are

important steps in the installation of solar technologies. Done properly, they ensure that a solar

project meets necessary safety, operational, environmental, and community compatibility

standards while not unduly hindering the project's completion. However, planning and permitting

processes not well designed for solar applications can increase the cost and time requirements of

a project substantially or even create enough delay and difficulty that the project is not

completed.

Section 2.4.5 describes some of the planning and permitting issues related to utility-scale solar

installations. This section focuses on planning and permitting by local and state governments for

smaller-scale, distributed PV installations. Installing a grid-connected PV system requires an

electrical permit from the local government and, in some cases, a building permit followed by

inspection of the installation (DOE 2009f).

Numerous planning and permitting barriers can hinder distributed PV system installation. The

following have been identified as most significant (Pitt 2008):

• Complex and unclear local permitting requirements

• Inspectors and permitting authorities inexperienced with renewable energy systems

• Permitting requirements that vary significantly across jurisdictions

• Permit fees that are high enough to become a significant, additional project cost

• Unfair and often illegal enforcement of restrictive housing covenants.

Recommendations to local and state governments for overcoming planning and permitting

barriers include the following (Pitt 2008, U.S. DOE 2009f):

• Understand the entire permitting and inspection process for PV systems and the dynamics

between all entities involved.

• Establish a clear path for communication between code enforcement offices and the local

utility provider to expedite the interconnection and inspection processes.

• Remove barriers to PV systems from building and zoning codes.

• Simplify PV permit application forms and review processes.

• Allow over-the-counter building permits for standard roof-mounted systems that do not

exceed the roof support capabilities of a structure meeting minimum building code

requirements.

• Adopt flat permit fees or fee waivers for PV systems.

• Ease permitting processes by establishing statewide interconnection standards (see

Section 4.2.2 below) and educating building and electrical inspectors about proper

installation procedures for distributed renewable energy systems.

• Adopt state-level legislation mandating consistent and appropriate permitting

requirements for distributed renewable energy systems.

A number of U.S. cities have led the way in modifying their planning and permitting policies to

encourage solar energy development (U.S. DOE 2009f). For example, San Jose, California,

grants electrical permits for PV systems over the counter and requires building permits only for

rooftop installations that meet certain criteria. Portland, Oregon, allows residential PV installers

to submit permit applications online and trains some permitting staff to be experts on solar

installations. Madison, Wisconsin, amended city laws to comply with state statutes that make it

illegal to forbid PV systems in historic districts. The Solar America Board for Codes and

Standards released a model expedited permitting process in October 2009 (Brooks 2009).

Continued efforts such as these will be necessary to expedite implementation of PV systems in

communities nationwide.

4.2.2 Interconnection

Interconnection standards specify the technical, legal, and procedural requirements by which

customers and utilities must abide when a customer wishes to connect a PV system to the grid

(or electricity distribution system). State governments can authorize or require their state public

utility commissions to develop comprehensive interconnection standards. Some state

interconnection standards apply to all types of utilities (investor-owned utilities, municipal

utilities, and electric cooperatives); other states have chosen to specify standards only for

investor-owned utilities. Although most utilities fall under the jurisdiction of state public utility

commissions, cities with municipal utilities can have significant influence over interconnection

standards in their territory.

The aspects of interconnection standards that are most often debated are procedural, not

technical, in nature. In setting technical interconnection standards, most regulatory bodies

reference compliance with the Institute of Electrical and Electronics Engineers’ (IEEE’s) “1547

Standard for Interconnecting Distributed Resources with Electric Power Systems,” which was

adopted in 2003. The most debated procedural aspects of interconnection standards are:

requirements for small inverter-based PV systems to have a utility external disconnect switch

(UEDS), limitations placed on PV system size, technical screens for interconnection, and

requirements for additional insurance (NNEC 2009).

States continue to demonstrate a mix of approaches to these key aspects. Eight states and many

major utility companies have recognized that safety devices and features already built into all

code-compliant PV systems make the UEDS redundant in small systems (less than 10 kW) and

have eliminated its requirement (Sheehan 2008).

Many states leave the UEDS requirement to

utility discretion. Interconnection standards with regard to PV system-size limitations also vary

widely among states, ranging from 10 kW to no cap on system size. As of November 2009,

seven states

and Puerto Rico specify no limit on system size (DSIRE 2009). The size of the PV

system and complexity of the interconnection typically dictate the rigor and extent of the

technical screens required before interconnection.

States also differ in their approaches to the issue of insurance requirements. States and some

utilities require owners of solar PV systems that are interconnecting to the grid to purchase

additional liability insurance to mitigate the risks of potential personal injury (e.g., to utility workers)

and property damage (NNEC 2009).

Thirteen states plus Washington, D.C., and Puerto Rico

require varying levels of insurance based on system size (IREC 2009).

Other states, including

Arizona, Indiana, Iowa, New Mexico, and Washington, leave the insurance requirement to utility

discretion (IREC 2009). Twelve states

do not require additional insurance and seven states

Figure 4.3 shows that 37 states plus Washington, D.C., and Puerto Rico have adopted an

interconnection policy. All of the states with robust solar markets have interconnection policies

in place. Robust solar markets do not exist in any of the states without interconnection policies.

not specify insurance as part of their interconnection standards (IREC 2009).

The eight states that have waived the UEDS requirement for small systems are Arkansas, Delaware, Florida,

Nevada, New Jersey, New Hampshire, North Carolina, and Utah. Utilities that have waived the UEDS requirement

for small systems include Pacific Gas and Electric (PG&E) and Sacramento Municipal Utility District (SMUD) in

California and National Grid U.S.A. in the Northeast United States.

California, Hawaii, Indiana, New Hampshire, North Carolina, Michigan, and Vermont do not have limits on the

capacity of interconnected solar PV systems (DSIRE 2009).

States with insurance requirements based on the size of the interconnected system are Colorado, Connecticut,

Delaware, Florida, Illinois, Massachusetts, Michigan, Minnesota, Missouri, Oregon, South Dakota, Virginia, and

Wisconsin.

States that do not require additional insurance for interconnected solar PV systems are California, Georgia,

Hawaii, Kentucky, Nevada, New Hampshire, New Jersey, New York, North Carolina, Pennsylvania, South

Carolina, and Vermont.

States that have not specified insurance requirements as part of their interconnection standards are Arkansas,

Louisiana, Maryland, Montana, Ohio, Texas, and Wyoming.

Figure 4.3. Interconnection Standards, November 2009

(DSIRE 2010b)

In the absence of a national interconnection standard, state regulators often consider the four

leading interconnection models when developing such policies.

These models include the

Federal Energy Regulatory Commission’s Small Generator Interconnection Procedure (SGIP)

and Small Generator Interconnection Agreement, California Rule 21, the Mid-Atlantic Demand

Resource Initiative’s Model Interconnection Procedures, and the Interstate Renewable Energy

Council’s (IREC’s) Model Interconnection Standards for Small Generator Facilities.

All four

procedures have comprehensive coverage of interconnection standards including specifications

for interconnecting systems up to 10 MW, pro forma interconnection agreements, fast-track

procedures for systems up to 2 MW, and a review process for interconnecting larger systems

(typically greater than 10kW).

California Rule 21, which was approved in December 2000 and

updated slightly based on utility tariff filings, is used for the interconnection of all solar and

distributed generation (DG) systems in utility service territories in California, which constitute a

majority of the solar installations in the United States (Keyes and Fox 2008).

Numbers for each state indicate system capacity limits in kW. Some state limits vary by utility, customer type

(e.g., residential/non-residential), technology, and/or system application. “No limit” means that there is no stated

maximum size for individual systems. Generally, state interconnection standards apply only to investor-owned

utilities. For more detail on the interconnection standards for each state, see

http://www.dsireusa.org/incentives/index.cfm?SearchType=Interconnection&&EE=0&RE=1.

U.S. Representative Jay Inslee (D-Washington) introduced the Clean Energy Buy-Back Act in March 2008, which

included federal interconnection standards. The bill did not pass.

For more details on the four interconnection models, see: www.solarabcs.org/interconnection/ABCS-

07_studyreport.pdf.

IREC and SGIP have simplified interconnection procedures for systems of 10 kW or less; California Rule 21

prequalifies systems less than 11 kW for three of the eight necessary technical screens (Keyes and Fox 2008).

State policy

Standard only applies to net-metered systems

WA: 20,000

OR: 10,000

CA: no limit

MT: 50*

NV: 20,000

UT: 25/2,000*

NM: 80,000

WY: 25*

HI: no limit

CO: 10,000

MN: 10,000

LA: 25/300*

AR: 25/300*

MI: no limit

WI: 15,000

MO: 100*

IN: no limit

IL: 10,000

FL: 2,000*

KY: 30*

OH: 20,000

NC: no limit

VT: no limit

NH: 100*

MA: no limit

37 states +

DC & PR

have adopted an

interconnection

policy

CT: 20,000

PA: 5,000*

NJ: 2,000*

DC: 10,000

MD: 10,000

NY: 2,000

VA: 20,000

SC: 20/100

GA: 10/100*

PR: no limit

TX: 10,000

NE: 25*

KS: 25/200*

SD: 10,000

State policy

Standard only applies to net-metered systems

WA: 20,000

OR: 10,000

CA: no limit

MT: 50*

NV: 20,000

UT: 25/2,000*

NM: 80,000

WY: 25*

HI: no limit

CO: 10,000

MN: 10,000

LA: 25/300*

AR: 25/300*

MI: no limit

WI: 15,000

MO: 100*

IN: no limit

IL: 10,000

FL: 2,000*

KY: 30*

OH: 20,000

NC: no limit

VT: no limit

NH: 100*

MA: no limit

37 states +

DC & PR

have adopted an

interconnection

policy

CT: 20,000

PA: 5,000*

NJ: 2,000*

DC: 10,000

MD: 10,000

NY: 2,000

VA: 20,000

SC: 20/100

GA: 10/100*

PR: no limit

TX: 10,000

NE: 25*

KS: 25/200*

SD: 10,000

4.2.3 Net Metering

Net metering is a policy that allows PV system owners to offset electricity purchases from the

utility with every kilowatt-hour of solar electricity a PV system produces. As with

interconnection standards, state governments can authorize or require their state public utilities

commissions to develop comprehensive net metering rules, and cities with municipal utilities can

have significant influence over net metering rules in their territory.

Net metering is one of the most important policy drivers for distributed PV systems because it

enables system owners to recover some of their investment through electricity bill savings

(Coughlin and Cory 2009). Under the simplest implementation of net metering, a utility

customer’s billing meter runs backward as solar electricity is generated and exported to the

electricity grid and runs forward as electricity is consumed from the grid. At the end of a billing

period, a utility customer receives a bill for the net electricity, which is the amount of electricity

consumed less the amount of electricity produced and exported by the customer’s PV system.

Figure 4.4 illustrates the variety of net-metering, system-size limitations across the United States.

Currently, 42 states and Washington, D.C., have net metering policies in place. Net metering

policies differ in several ways, including the eligibility of different technology types, customer

classes, system sizes, the use of aggregate caps for DG contribution back to the grid, the

treatment of customer net-excess generation, the types of affected utilities, and the issue of REC

ownership (IREC and NCSC 2007). Detailed state-specific information regarding net-metering

availability and regulation is available through the Web site for the Database of State Incentives

for Renewables & Efficiency (DSIRE) (www.dsireusa.org/

Figure 4.4. Net Metering Policies, October 2009

(DSIRE 2009)

Numbers for each state indicate system capacity limits in kW. Some state limits vary by utility, customer type

(e.g., residential/non-residential), technology, and/or system application. “No limit” means that there is no stated

maximum size for individual systems. For more detail on the net metering standards for each state, see

http://www.dsireusa.org/incentives/index.cfm?SearchType=Net&&EE=0&RE=1.

RI: 1,650/2,250/3,500*

MA: 60/1,000/2,000*

PA: 50/3,000/5,000*

State policy

Voluntary utility program(s) only

State policy applies to certain utility types only (e.g., investor-owned utilities)

WA: 100

MT: 50*

NV: 1,000*

UT: 25/2,000*

AZ: no limit*

ND: 100*

NM: 80,000*

WY: 25*

HI: 100

KIUC: 50

CO: no limit

co-ops & munis: 10/25

OK: 100*

MN: 40

LA: 25/300

AR: 25/300

MI: 150*

WI: 20*

MO: 100

IA: 500*

IN: 10*

IL: 40*

FL: 2,000*

KY: 30*

OH: no limit*

GA: 10/100

WV: 25

NC: 1,000*

VT: 250

VA: 20/500*

NH: 100

CT: 2,000*

NY: 25/500/2,000*

NJ: 2,000*

DE: 25/500/2,000*

MD: 2,000

DC: 1,000

42 states + DC

have adopted a

net metering policy

NE: 25

KS: 25/200*

ME: 660, co-ops & munis: 100

PR: 25/1,000

OR: 25/2,000*

CA: 1,000*

4.2.4 Direct Cash Incentive Programs

Direct cash incentives give solar energy system owners cash back for a qualified solar

installation. Qualified solar installations vary by state and may include solar electricity-

producing, water heating, and space heating and cooling technologies. Direct cash incentives

include rebates, grants, and production- or performance-based incentives that complement other

financial incentives such as tax credits.

The manner and timing in which direct cash incentives are paid varies by location and program

design. Rebate and grant amounts are often based on system size or system cost, and the funding

is typically awarded at the time of installation. Performance or production-based incentives are

distributed to project owners over several years based on the amount of energy the system

produces. Expected performance rebates are based on solar system capacity as well as system

rating, location, tilt and orientation, and shading. Expected performance rebates may be

distributed in a lump sum, but are calculated based on the expected energy output of the system.

Payments based on performance or expected performance instead of capital investments are

gaining prominence among program administrators because they encourage optimized system

design and installation. To avoid a boom-and-bust cycle that can disrupt solar energy markets,

careful consideration should be given to incentive levels, program caps, and long-term funding

mechanisms for direct cash incentive programs.

California, the leading U.S. state in terms of installed PV capacity, provides an example of a

direct cash incentive program. In January 2006, the California Public Utilities Commission

launched the California Solar Initiative (CSI), a direct cash incentive program providing more

than $3 billion for solar energy projects with the objective of installing 3,000 MW of solar

capacity by 2016. CSI includes a transition to performance-based and expected performance-

based incentives (as opposed to up-front payments based only on system size), with the aim of

maximizing system performance through effective system design and installation. CSI incentive

levels will automatically be reduced over the duration of the program in ten steps based on the

aggregate capacity of solar installed in each utility service area. The California Public Utility

Commission designated funding sources for the CSI program for 10 years (2006–2016).

Figure 4.5 shows the states in which direct cash incentives are available.

Figure 4.5. States that offer direct cash incentives for solar projects, September 2009

(DSIRE 2009)

States and utilities usually administer direct cash incentive programs, but some local

governments also offer these incentives to consumers. Currently, approximately 24 states and

more than 100 utilities offer direct incentives for solar installations. Direct cash incentives are

often funded through a public or systems benefits fund, clean energy funds, a revolving loan

fund, or the general fund. The incentives typically cover 20% to 50% of project costs and range

from a few hundred to millions of dollars (U.S. DOE 2009f).

4.2.5 Renewable Portfolio Standards and Solar Set-Asides

A renewable portfolio standard (RPS) is a policy that requires utilities or load-serving entities

(LSEs) to provide its customers with a certain amount of electricity generated from renewable

resources. While an RPS is typically a mandate, it can also be a non-binding goal; it is almost

always stated as a percentage of the total electricity provided to be reached by a predetermined

future date (Bird and Lockey 2008). As indicated in Figure 4.6, 29 states plus the District of

Columbia have renewable portfolio standards in place, and an additional 6 states have non-

binding renewable energy production goals.

State Direct Cash Incentives for PV

State Direct Cash Incentives for Solar Water Heating

State Direct Cash Incentives for both PV and Solar Water Heating

24 States +

DC and the

Virgin Islands

offer Direct Cash

Incentives for

Solar Projects

Utility Direct Cash Incentive(s) for PV and/or Solar Water Heating

U.S. Virgin Islands

Figure 4.6. State renewable portfolio standards and goals, October 2009

(DSIRE 2009)

In effect, two products are produced from renewable energy generation: the environmental

attributes sold in the form of renewable energy credits (RECs) and the actual electricity produced

by the renewable generator. A REC typically represents the attributes of 1 MWh of electricity

generated from renewable energy. An unbundled REC represents the environmental benefits

without the actual energy, and bundled RECs include both the environmental benefits and the

actual energy produced by a renewable source. Many states allow RECs to be bought unbundled

from the associated electricity and used to fulfill RPS obligations (Holt and Wiser 2007).

A growing number of states are incorporating a “set-aside” or “carve-out” within the RPS,

stipulating that a portion of the required renewable energy percentage or overall retail sales be

derived from solar or DG resources

Note that when solar PV is located on residential, business, or government property and not part of a large

centralized power system, it is considered to be distributed generation.

(DSIRE 2009). Figure 4.7 shows 15 states along with

Washington, D.C., that have these set-asides or carve-outs for solar electricity generation and

solar water heating. In addition, Massachusetts legislation has established a DG set-aside, but

specific targets have not yet been established. Only three states and Washington, D.C., allow

solar water heating to count toward the solar set-aside requirements. Figure 4.7 illustrates that

solar electricity demand based on existing RPS set-aside requirements alone would amount to the

installation of more than 8 GW of cumulative installed solar capacity by 2025 (Wiser and

Barbose 2008, Wiser and Barbose 2009).

State renewable portfolio standard

State renewable portfolio goal

Solar water heating eligible

†

Extra credit for solar or customer-sited renewables

Includes non-renewable alternative resources

WA: 15% by 2020*

☼ NV: 25% by 2025*

☼ AZ: 15% by 2025

☼ NM: 20% by 2020 (IOUs)

10% by 2020 (co-ops)

HI: 40% by 2030

☼

Minimum solar or customer-sited requirement

TX: 5,880 MW by 2015

UT: 20% by 2025*

☼ CO: 20% by 2020 (IOUs)

10% by 2020 (co-ops & large

munis)*

MT: 15% by 2015

ND: 10% by 2015

SD: 10% by 2015

IA: 105 MW

MN: 25% by 2025

(Xcel: 30% by 2020)

☼ MO: 15% by 2021

WI: Varies by utility;

10% by 2015 goal

MI: 10% + 1,100 MW by 2015*

☼ OH: 25% by 2025†

ME: 30% by 2000

New RE: 10% by 2017

☼ NH: 23.8% by 2025

☼ MA: 15% by 2020

+ 1% annual increase

(Class I Renewables)

RI: 16% by 2020

CT: 23% by 2020

☼ NY: 24% by 2013

☼ NJ: 22.5% by 2021

☼ PA: 18% by 2020†

☼ MD: 20% by 2022

☼ DE: 20% by 2019*

☼ DC: 20% by 2020

VA: 15% by 2025*

☼ NC: 12.5% by 2021 (IOUs)

10% by 2018 (co-ops & munis)

VT: (1) RE meets any increase in

retail sales by 2012;

(2) 20% RE & CHP by 2017

29 states + DC

have an RPS

6 states have goals

KS: 20% by 2020

☼ OR: 25% by 2025 (large utilities)*

5% - 10% by 2025 (smaller utilities)

☼ IL: 25% by 2025

WV: 25% by 2025*†

CA: 33% by 2020

Figure 4.7. Solar capacity to meet existing RPS solar set-aside requirements, November 2009

(Wiser and Barbose 2009)

As with the overall RPS requirements, to reach the goal of the solar set-aside, utilities or LSEs

can either own the solar generation capacity or purchase bundled or unbundled solar RECs

(SRECs) (Cory et al. 2008). One major difference between RECs and SRECs is their cost;

SRECs typically generate more revenue than RECs, providing an additional financial incentive

to install solar power systems. To create a value for RECs and SRECs, however, the RPS must

include a penalty or alternative compliance mechanism that has a distinctly higher penalty for

those not complying with the RPS or solar set-aside (Cory and Coughlin 2009).

SRECs are one of the key elements that can make a third-party PV ownership project financially

feasible, resulting in as much as 40% to 80% of revenues for third-party PV projects. Two states

that have witnessed high revenues from SRECs are Colorado and New Jersey (where SREC

prices have ranged from $160/MWh to $265/MWh). In New Jersey, SREC prices are expected to

rise from $300/MWh to $711/MWh in 2009 as the up-front rebates are removed (Cory et al.

2008).

New Jersey announced its Solar Renewable Energy Credit Registration Program in August 2009,

designed to incentivize solar development in the state. One SREC is issued for each MWh of

electricity generated from a solar electric system. The SRECs represent all the clean energy

benefits from the solar generation and are sold or traded separately from the power, providing

solar system owners with a source of revenue to help offset the costs of installation. The New

Jersey SREC Program is expected to almost entirely replace the state’s rebates, which fueled

solar growth in the early years of the state’s solar program.

200

400

600

800

1,000

1,200

1,400

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

Annual Solar Additions (MW)

Cumulative Solar Capacity (MW)

Annual Capacity

(right axis)

Cumulative

Capacity

(left axis)

4.2.6 Clean Energy Funds

In the mid- to late-1990s, a bevy of states introduced retail competition and with it concerns over

continued funding for energy efficiency, renewable energy, and low-income energy assistance

programs. To address those concerns, system (or public) benefit funds were established and

supported through an additional charge to end-users’ electricity bills, either as a per-kWh charge

or a flat fee. The fund is then dispersed in various forms, such as grants, loans, and rebates, to

support investments in renewable energy, energy efficiency, and related improvements for low-

income housing. Eligibility to receive support from a system benefit fund is contingent on the

recipient being an electricity ratepayer (Cory et al. 2008).

As Figure 4.8 indicates, the systems benefits charges can amount to large funds in the aggregate.

For example, California is expecting to raise more than $4 billion from 1998 to 2016, and New

Jersey expects to raise $647 million from 2001 to 2012. Across the nation, an estimated

$7.3 billion in 16 states and Washington, D.C., will have been collected by 2017.

In subsequent years, a second generation of “clean energy funds” was created and supported

through a variety of methods, including RPS alternative compliance payments, general fund

obligations, and oil and gas severance tax payments. These clean energy funds typically support

energy efficiency and renewable energy programs and represent a renewed interest by states in

the deployment of clean energy technologies. These second-generation funds were established in

Alaska, Colorado, Iowa, and Maryland.

Figure 4.8. Estimated system benefit funds for renewables, May 2009

(DSIRE 2009)

The way in which these funds are dispersed varies from state to state. In California, for example,

the system benefit fund supports the CSI outlined in Section 4.2.3.

State PBF supported by voluntary contributions

* Fund does not have a specified

expiration date

** The Oregon Energy Trust is scheduled

to expire in 2025

RI: $2.2M in 2009

$38M from 1997-2017*

MA: $25M in FY2009

$524M from 1998-2017*

NJ: $78.3M in FY2009

$647M from 2001-2012

DE: $3.4M in 2009

$48M from 1999-2017*

CT: $28M in FY2009

$444M from 2000-2017*

VT: $5.2M in FY2009

$33M from 2004-2011

PA: $950,000 in 2009

$63M from 1999-2010

IL: $3.3M in FY2009

$97M from 1998-2015

NY: $15.7M in FY2009

$114M from 1999-2011

WI: $7.9M in 2009

$90M from 2001-2017*

MN: $19.5M in 2009

$327M from 1999-2017*

MT: $750,000 in 2009

$14M from 1999-2017*

OH: $3.2M in 2009

$63M from 2001-2010

MI: $6.7M in FY2009

$27M from 2001-2017*

ME: 2009 funding TBD

$580,300 from 2002-2009

DC: $2M in FY2009

$8.8M from 2004-2012

OR: $13.8M in 2009

$191M from 2001-2017**

CA: $363.7M in 2009

$4,566M from 1998-2016

State PBF

16 states + DC

have public benefits

funds ($7.3 billion

collected by 2017).

ME has a voluntary

public benefits fund

It is important to note that the state funds have been instrumental in driving the installation of

grid-connected PV systems; more than 75% of the grid-connected PV systems installed in the

United States in 2007 were located in states with a clean energy fund. Moreover, the state funds

have invested more than 75% of their available funding in PV projects (Clean Energy States

Alliance 2009).

4.2.7 Emerging Trends

Several policy/financing mechanisms are emerging that have the potential to incite further solar

market expansion through the establishment of widespread local and utility programs. The three

topics discussed in this section are feed-in tariffs, property tax financing, and rate structures.

A feed-in tariff (FIT) is a requirement for utilities to purchase electricity from eligible renewable

systems at a guaranteed price over a fixed period. Alternatively, a FIT can consist of a fixed or

variable premium above the market price. A FIT increases the rate of PV deployment by

providing a more stable revenue stream from PV systems and improving the rate of return on PV

investments. The payment level generally is designed to ensure that the systems are able to

recover costs and provide a modest profit. In addition to offering guaranteed prices, FITs

typically guarantee grid access, allowing both small and large projects to connect to the grid

according to uniform interconnection standards. FITs have been used extensively in Europe and

are starting to be implemented in the United States (Couture and Cory 2009).

Municipalities and counties across the country are launching innovative public/private financing

programs that allow property owners to spread the cost of renewable energy systems over the

long term. For example, Berkeley, California, and Boulder, Colorado, have passed initiatives to

allow homeowners and businesses the opportunity to finance PV systems through adjustments to

their property taxes, thus taking advantage of the government entities’ tax-free financing

capabilities to support expansion of these resources at the local level. Programs utilizing this

financing approach are commonly referred to as property-assessed clean energy (PACE)

programs and will be discussed in more detail in Section 4.3.3, along with other innovative

financing mechanisms.

As customer-sited generation and advanced metering technologies become more prevalent, there

is an increased interest in developing alternative rate structures that reflect the resulting changes

in electricity use. The majority of existing rate structures do not capture the actual value of time-

varying increases or decreases in demand for electricity, and therefore are unlikely to capture the

value of energy produced by customer-sited generation including solar PV. This is because solar

PV peak generation often correlates well with peak electricity demand. With the availability of

more advanced metering, it is possible to create rate structures that better reflect the variances in

the value of electricity as demand fluctuates. Appropriate rate structures for PV could enable

better capture of the value of excess customer generation exported to the grid. While time-of-use

rates and other emerging rate structures are still relatively uncommon, it is anticipated that they

will become increasingly more prevalent and serve as a driver for solar market expansion.

4.3 Private Sector and Market-Based Developments to Facilitate Solar

Deployment

The financing of solar and other renewable energy technologies experienced significant changes

in 2008 and early 2009. The passage of the ARRA in February 2009 greatly expanded the

availability and usability of various tax credits, depreciation opportunities, loan guarantees, and

other mechanisms designed to incent private and public investment in renewable energy and

energy efficiency projects. The ARRA expanded and extended an array of incentives made

available with passage of the EESA (and other earlier laws).

In addition to the previously described support mechanisms, private sector and other solar market

stakeholders, including states, counties, and municipalities, have developed mechanisms to

support renewable energy financing by residents, businesses, and institutional and government

consumers of energy.

Three prominent financing mechanisms/programs for solar PV and CSP will be discussed in this

section: the third-party power purchase agreement (PPA), the solar lease, and property-assessed

clean energy (PACE) programs.

4.3.1 Third-Party Power Purchase Agreement Financing

All sectors can use the third-party ownership PPA, including homeowners, businesses, utilities,

and state and local governments.

In a third-party ownership PPA model, one party hosts a PV system on his or her property and a

solar developer purchases, installs, owns, operates, and maintains the system. In the residential

sector, it is the homeowner that hosts and does not purchase or own the PV system, and instead

buys the electricity produced by the PV system under a long-term PPA (see Figure 4.9). In

exchange for signing the PPA, the homeowner avoids paying for the PV system up front and

usually is not responsible for the operation and maintenance (O&M) of the system. The PPA

provider receives the monthly cash flows in the form of power sales and the fully monetized

federal tax benefits, including the investment tax credit and accelerated depreciation. An

example of a PPA provider for the residential (and commercial) market is SolarCity, which

offers a variety of lease structures, including a zero-down option; however, the higher the down

payment, the lower the monthly lease payment (Coughlin and Cory 2009).

Figure 4.9. The residential power purchase agreement

(NREL 2009b)

The commercial PV market has witnessed a rapid proliferation of the use of the PPA-financing

model of ownership. Greentech Media estimated that roughly 80% of the commercial market in

2008 will have used a PPA for PV installations, an increase from previous levels of 50% in 2007

and 10% in 2006 (Guice and King 2008). The benefits of the PPA method for financing PV

deployment in the commercial sector are similar to those for customers in the residential sectors.

Thus, a PPA provides the opportunity for a commercial owner to host rather than own a PV

system. Instead of securing capital up front and being responsible for O&M, the business owner

signs a long-term PPA to purchase the electricity generated by the system. The PPA is typically

priced at or below the prevailing utility retail rate in the first year (with perhaps some escalation

over the life of the contract). The owner of the business avoids most, if not all, of the up-front

purchase and installation costs, as well as O&M responsibilities.

Utilities often rely on third parties to design, finance, manage construction of, and operate and

maintain solar facilities. Development of these facilities requires the long-term procurement of

the power output. Accordingly, utilities sign PPAs with the developers, allowing the developer to

obtain lower-cost financing, passing on the savings through relatively lower power prices. PPAs

can come in many forms and durations, but generally payments are made for both the plant

capacity (maximum capable output) and energy production, and PPAs cover a 15- to 20-year

period starting with a facility’s commercial operation. However, now that utilities can use the

ITC directly, their use of third-party PPAs may be affected (see Section 4.3.5).

Utilities benefit from PPAs as they are designed to leverage the technical expertise and

experience of the solar developer. PPAs also allocate risks of cost overruns, plant availability,

etc., to a pre-specified party, typically the plant developer. In return for accepting most

development and operating risks, the developer receives price certainty and marketability of its

product.

PPAs allow faster development of solar resources as utilities can leverage the capability of many

developers simultaneously. For example, Pacific Gas & Electric has approximately 3 GW of

large-scale CSP and PV solar facilities under development via its PPA counterparties, including

such entities as BrightSource Energy, First Solar, Sun Power, Solel, and Solaren Corp.

(Mendelsohn and Kreycik, forthcoming).

State and local governments are also responding to the challenges of funding PV development on

their buildings and land by using innovative finance structures such as the third-party-ownership

PPA model. As with the residential and commercial sectors, the benefits of transferring the up-

front costs and O&M responsibilities to the owner/developer, maintaining steady electricity

prices, and using federal tax benefits inherent to the PPA ownership model have made it an

attractive option for state and local properties.

Note that third-party PPA financing may face regulatory or legal challenges in some states,

especially where the issue of utility commission regulation of third-party owned systems has not

been specifically addressed (Kollins, forthcoming).

4.3.2 Customer Solar Lease Financing

The customer solar lease is similar to the residential or commercial PPA in that a property owner

hosts, but does not own, a solar PV system. To take advantage of federal tax incentives, a third-

party lessor finances and owns the solar PV installation. However, distinct to a solar lease, the

property owner (as lessee) pays to use the equipment instead of purchasing the generated power.

Thus, the customer’s lease payment remains constant even if the system’s output fluctuates. If

the system does not meet the customer’s entire energy needs, the customer purchases additional

electricity from his/her utility. Any excess electricity generated by the system can be net

metered, earning the customer cents/kWh credits on his/her electric utility bill.

Similar to a third-party PPA, the solar lease transfers the high up-front costs to the system

owner/developer, who can take advantage of valuable federal tax incentives. Some of the cost

savings might be passed down to the customer in the form of lower payments. In states with

complementary incentives, lease payments can be less than or equal to monthly utility savings.

Also, like the third-party PPA, the lease may shift maintenance responsibilities to the developer.

There are challenges associated with the solar lease. For example, the leasing company may not

have as strong an incentive to maintain the system as it would under a third-party PPA contract,

because the customer’s payments are fixed regardless of the system’s output. However, some

companies will monitor the system’s output and will provide maintenance promptly, or will

include a performance guarantee that ensures a minimum kWh output (Kollins et al.,

forthcoming). Also, as with the third-party PPA, the solar lease may face regulatory challenges

in some states (Kollins et al., forthcoming). In addition, the traditional solar lease may not be

available to non-taxable entities such as state and local governments because of uncertainty

For lease payments to nearly equal utility bill savings, the leasing company must have additional incentives, such

as up-front rebates or performance based payments (Coughlin and Cory 2009).

about renewing contracts on a year-to-year basis. However, state and local governments may be

able to use a tax-exempt lease where payments to the lessor are tax exempt (Bolinger 2009).

4.3.3 Property-Assessed Clean Energy Programs

In addition to private-sector financing mechanisms, local governments have also designed

programs to fund energy efficiency and renewable energy development on private property, with

a particular focus on funding PV installations.

Several municipalities are assisting residential and commercial ownership of renewable energy

systems through financing via property tax assessments.

Property-assessed clean energy (PACE) assessments

Piloted by the Berkeley FIRST

program and replicated elsewhere, the property tax assessment model finances the cost of

renewable energy and energy efficiency improvements through the creation of special tax

districts (Coughlin and Cory 2009). Interested property owners may opt into the program and

pay for an additional line item on their property tax bill.

transform the high upfront costs into the

equivalent of a moderate monthly payment

and allow the property owner to transfer the

assessment and capital improvement to new property owners in the event of a sale (Koenig and

Speer, forthcoming).

Figure 4.10 indicates active PACE programs in the United States, as well as states with current

or pending PACE-enabling legislation.

Under a PACE program, a municipality provides the financing to pay for

the up-front system costs for a renewable energy system through an additional property tax

assessment. The municipality funds the assessments by public bond offerings, micro bonds,

general funds, or municipal waste funds (Koenig and Speer, forthcoming). The property owner

repays the cost of the system, plus interest and administrative fees, through additional

assessments placed on his or her property tax bill, which are collected over a time period that

reflects the useful life of the improvements. The property tax assessment model eliminates nearly

all of the up-front costs of installation to the property owner with the exception of program

administrative fees. Further, the renewable energy system and accompanying special property

tax is fixed to the property and not the property owner. Thus, homeowners pay for benefits from

the renewable energy system only while they own the property (Coughlin and Cory 2009;

Koenig and Speer, forthcoming).

To date, only the Sonoma County (CA) program is open to commercial property owners. However, Boulder

County (CO) will be opening up the ClimateSmart Loan program to business owners in 2010 with a new round of

funding.

Assessments are similar to loans in that they allow a property owner to pay off debt in installments over a long

period of time. However, PACE assessments are not legally considered to be loans.

Note that payments are typically made semi-annually. However, the semi-annual payment could be considered by

the property owner as six moderate monthly installments.

While property owners may be able to transfer the assessment to a new buyer, a buyer could require that all liens

on the property (including the PACE assessment) be settled before the property is transferred.

100

Figure 4.10. Property-assessed clean energy programs, November 2009

(DSIRE 2010b)

4.3.4 Alternative Financing Structures: Partnership Flips and Leases

PPAs, solar leases, and SRECs are used between a customer and a developer to help finance

solar project development. There are also several other financing mechanisms that are used

between a project developer and a separate tax investor. These financing alternatives are

designed to facilitate full and efficient use of federal and state tax benefits by transferring tax

subsidies to tax-burdened investors. Examples of these financing structures include partnership

flips and leases.

In a partnership flip, ownership of a solar project is shared between a developer and a tax equity

investor, who contributes project investment capital in exchange for federal and state tax benefits

and some revenue. Once the tax equity investor reaches a specified rate of return, the project’s

economic returns are redistributed, or “flipped,” between the developer and tax equity investor,

with the developer typically receiving the majority of electricity sales revenue (Martin 2009).

Solar developers and investors have also financed solar projects with various forms of equipment

leases such as a sale-leaseback and an inverted pass-through. In a sale-leaseback, a developer

sells a solar system and accompanying tax benefits to a tax equity investor, who in turn leases

back the use and possession of the solar property to the same developer (Martin 2009). In an

inverted pass-through lease, the roles of the developer and tax equity investor are effectively

reversed (inverted). The developer makes an election to pass through the investment tax credit to

the tax equity investor along with revenue from the system’s electricity sales. The developer

PACE enabling legislation

Existing authority without new legislation

Active PACE Program

1. Babylon, NY

2. Berkeley, CA

3. Boulder County, CO

4. Palm Desert, CA

5. Sonoma County, CA

101

receives fixed lease payments from the tax equity investor, as well as the tax benefit of

accelerated depreciation (Jones and Lowman 2009).

4.3.5 Increasing Utility Ownership of Solar Projects

Recent federal legislation has also greatly increased the incentive for utilities to directly own

solar projects themselves and not require a separate tax investor. In particular, the EESA

contained three provisions that promote direct utility ownership of solar projects. First, the

8-year investment tax credit extension provides long-term certainty regarding the availability of

the credit. Second, utilities are permitted to take the investment tax credit directly, which was

previously unavailable.

Third, the investment tax credit can also be applied to a renewable

energy system owner’s alternative minimum tax―formerly a significant barrier to entry

(Schwabe et al. 2009).

The ARRA also provided an extension of the 50% bonus depreciation in year one in addition to

MACRS, the five-year accelerated depreciation. However, public utilities may not be able to use

MACRS unless their regulators allow them to include the solar property under a normalization

method of accounting (SEPA 2009). Also, utility ownership enables utilities to count electricity

generated toward state RPS requirements.

4.4 References

Bird, L; Lockey, E. (2008). Interaction of Compliance and Voluntary Renewable Energy

Markets. National Renewable Energy Laboratory. NREL Report No. TP-670-42096.

Bolinger, M. (2009). Financing Non-Residential Photovoltaic Projects: Options and

Implications. LBNL Report No. LBNL-1410E. Berkeley, CA: Lawrence Berkeley National

Laboratory.

Brooks, B. (2009). Expedited Permit Process for PV Systems. A Standardized Process for the

Review of Small-Scale PV Systems. Solar America Board for Codes and Standards Report.

http://www.solarabcs.org/permitting/

. Accessed November 2009.

Clean Energy States Alliance. (2009). State Clean Energy Fund Support for Renewable Energy

Projects. Key Findings from the CESA National Database.

www.cleanenergystates.org.

Accessed October 20, 2009.

Cory, K.; Coggeshall, C. Coughlin, J., Kreycik, C. (2009). Solar Photovoltaic Financing:

Deployment by Federal Government Agencies. NREL/TP-6A2-46397. Golden, CO: National

Renewable Energy Laboratory.

Cory, K.; Coughlin, J.; Coggeshall, C. (2008). Solar Photovoltaic Financing: Deployment on

Public Property by State and Local Governments. NREL/TP-670-43115. Golden, CO: National

Renewable Energy Laboratory.

Coughlin, J.; Cory, K. (2009). Solar Photovoltaic Financing: Residential Sector Deployment.

National Renewable Energy Laboratory. NREL/TP-6A2-44853. Golden, CO: National

Renewable Energy Laboratory.

Some questions remain about how a utility can take advantage of the ITC under Normalization Accounting rules

(where IOUs have to share tax benefits with customers over time). This is explained further in a forthcoming article,

“Navigating Regulated Investment in Solar Generation: Rewards, Challenges, and Options,” by Alvarez, P., and

Hodges, B., in Electric Perspectives, the magazine of the Edison Electric Institute.

102

Couture, T.; Cory, K. (2009). State Clean Energy Policies Analysis (SCEPA) Project: An

Analysis of Renewable Energy Feed-in Tariffs in the United States. National Renewable Energy

Laboratory. NREL Report No. TP-6A2-45551.

DSIRE. (2009). “Database of State Incentives for Renewables and Efficiency.” Summary Maps.

www.dsireusa.org/summarymaps/index.cfm?ee=1&RE=1. Accessed October 2009.

DSIRE. (2010a). “Database of State Incentives for Renewables and Efficiency.” Clean

Renewable Energy Bonds.

http://www.dsireusa.org/incentives/incentive.cfm?Incentive_Code=US45F&re=1&ee=1.

Accessed January 2010.

DSIRE. (2010b). “Database of State Incentives for Renewables and Efficiency.” Summary

Maps. www.dsireusa.org/summarymaps/index.cfm?ee=1&RE=1. Accessed January 2010.

Guice, J.; King, J.D.H. (February 14, 2008). Solar Power Services: How PPAs are Changing the

PV Value Chain. Prometheus Institute and Greentech Media.

Holt, E.; Wiser, R.H. (2007). The Treatment of Renewable Energy Certificates, Emissions

Allowances, and Green Power Programs in State Renewables Portfolio Standards. Berkeley,

CA: Lawrence Berkeley National Laboratory.

Interstate Renewable Energy Council (IREC); North Carolina Solar Center (NCSC). (2007).

Connecting to the Grid 5th Edition, A Guide to Distributed Generation Interconnection Issues.

www.irecusa.org/fileadmin/user_upload/ConnectDocs/IC_Guide.pdf. Accessed March 16, 2009.

Interstate Renewable Energy Council (IREC). (2008). Updates and Trends.

www.irecusa.org/fileadmin/user_upload/NationalOutreachDocs/IREC_Solar_Market_Trends_Re

vision_11_19_08-1.pdf . Accessed November 16, 2009.

Interstate Renewable Energy Council (IREC). (2009). “State-by-State Interconnection Table.”

http://irecusa.org/wp-content/uploads/2009/11/November_2009_IC_Table.doc. Accessed

December 22, 2009.

Jones, L.; Lowman, D. (2009). “Treasury Grant Program.” Webinar. Hunton and Williams, LLP.

www.hunton.com/media/webinars/recordings/20090714/20090714.html?utm_source=MagnetMa

il&utm_medium=email&utm_term=g[email protected]&utm_content=Treasury%20Grant%20

Program%20Web%20Conference%20Recording%20Available&utm_campaign=martech_0499_

treasury_grant_prgrm_followup_7.15.09. Accessed September 22, 2009.

Kassakhian, Ken. (2009). Clean renewable energy bonds analysis. Golden, CO: National

Renewable Energy Laboratory (internal only).

Koenig, R.; Speer, B. Property Assessed Clean Energy (PACE) Financing: Challenges and

benefits of attracting larger capital investors. Golden, CO: National Renewable Energy

Laboratory (forthcoming NREL technical report).

Kollins, K.; Speer, B.; Cory, K. Solar Photovoltaic Financing: Regulatory and Legal Issues with

Third-Party PPA System Owners. Golden, CO: National Renewable Energy Laboratory

(forthcoming NREL technical report).

Martin, K. (2009). Utility Solar Tax Manual: A Utility Guide to the Federal Investment Tax

Credit. Solar Electric Power Association.

103

Mendelsohn, M.; Kreycik, C. Concentrating Solar Power and Utility Scale Photovoltaics in the

U.S.: Policy Drivers and Financing Incentives. Golden, CO: National Renewable Energy

Laboratory (forthcoming NREL technical report).

Network for New Energy Choices (NNEC) (2009). Freeing the Grid: 2009 Edition. New York, NY:

Network for New Energy Choices.

http://www.newenergychoices.org/uploads/FreeingTheGrid2009.pdf

. Accessed January 18, 2010.

NREL. (2009a). Solar Leasing for Residential Photovoltaic Systems. Fact Sheet. NREL Report

TP-670-43572. Golden, CO: National Renewable Energy Laboratory.

NREL. (2009b). “The residential power purchase agreement.” Golden, CO: National Renewable

Energy Laboratory (internal only).

Pitt, D. (2008). Taking the Red Tape Out of Green Power: How to Overcome Permitting

Obstacles to Small-Scale Distributed Renewable Energy. Network for New Energy Choices.

www.newenergychoices.org/uploads/redTape-rep.pdf. Accessed November 30, 2009.

Prometheus Institute (2009). “Renewable Energy Industry Note.”

www.prometheus.org/system/files/Solar+for+Federal+Buildings.pdf. Accessed December 31,

2009.

Schwabe, P.; Cory, K.; Newcomb, J. (2009). Renewable Energy Project Financing: Impacts of

the Financial Crisis and Federal Legislation. NREL Report No. TP-6A2-44930. Golden, CO:

National Renewable Energy Laboratory.

SEIA. (2008). “The Solar Investment Tax Credit Frequently Asked Questions.” Solar Energy

Industries Association.

http://seia.org/galleries/pdf/ITC_Frequently_Asked_Questions_10_9_08.pdf. Accessed

November 2008.

SEPA. (2009). Utility Solar Tax Manual. Report #02-09. Solar Electric Power Association.

http://www.solarelectricpower.org/resources/reports.aspx. Accessed January 2010.

Sheehan, M. (2008). Utility External Disconnect Switch Practical, Legal, and Technical Reasons

to Eliminate the Requirement. Solar America Board for Codes and Standards.

http://www.solarabcs.org/utilitydisconnect/. Accessed November 20, 2009.

U.S. DOE. (2008). Renewable Energy Production Incentive. U.S. Department of Energy. Office

of Energy Efficiency and Renewable Energy.

http://apps.eere.energy.gov/repi/. Accessed

October 30, 2008.

U.S. DOE. (2009a). Energy Efficiency and Conservation Block Grant Web site. U.S. Department

of Energy. Office of Energy Efficiency and Renewable Energy.

www.eecbg.energy.gov.

Accessed August 2009.

U.S. DOE. (2009b). Energy Efficiency and Conservation Block Grant Program Fact Sheet. U.S.

Department of Energy. Office of Energy Efficiency and Renewable Energy.

www.eecbg.energy.gov/downloads/WIP_EECBG_Fact_Sheet_FINAL_Web.pdf. Accessed

September 2009.

U.S. DOE (2009c). “DOE Loan Guarantee Programs.” U.S. Department of Energy. Solar

Technologies Program and Loan Guarantee Program (internal only).

104

U.S. DOE. (2009d). (March 20, 2009). “Obama Administration Offers $535 Million Loan

Guarantee to Solyndra, Inc.” Press Release. U.S. Department of Energy. Loan Guarantee

Program. www.lgprogram.energy.gov/press/032009.pdf. Accessed September 2009.

U.S. DOE. (2009e). State Energy Program Fact Sheet. U.S. Department of Energy. State Energy

Program. http://apps.eere.energy.gov/state_energy_program/pdfs/sep_factsheet.pdf. Accessed

September 2009.

U.S. DOE. (2009f). Solar Powering Your Community: A Guide for Local Governments. U.S.

Department of Energy. Solar Energy Technologies Program.

www.solaramericacities.energy.gov/resources. Accessed October 2009.

U.S. Department of Treasury, Internal Revenue Service (IRS). (2006). “Clean Renewable Energy

Bond Volume Cap Allocation Information.”

http://www.irs.gov/newsroom/article/0,,id=164423,00.html. Accessed January 2010.

U.S. Department of Treasury, Internal Revenue Service (IRS). (2007). “2007 Disclosure of

Allocations.” http://www.irs.gov/pub/irs-tege/creb_2007_disclosure.pdf. Accessed January 2010.

U.S. Department of Treasury, Internal Revenue Service (IRS). (2009). “New Clean Renewable

Energy Bonds - 2009 Allocations.” http://www.irs.gov/pub/irs-

tege/ncrebs_2009_allocations_v1.1.pdf. Accessed January 2010.

Wiser, R.; Barbose, G. (2008). Renewables Portfolio Standards in the United States. A Status

Report with Data Through 2007. Berkeley, CA: Lawrence Berkeley National Laboratory. LBNL

Report No. LBNL-154E.

Wiser, R.; Barbose, G. (November 2009). “Solar capacity to meet existing RPS solar set-aside

requirements.” Berkeley, CA: Lawrence Berkeley National Laboratory (internal only).

105

5. Investments and Future Outlook

This chapter provides information on trends in private investment in solar energy (Section 5.1), a

summary of DOE investment in solar energy and its role in the solar industry (Section 5.2), and a

review of near-term forecasts for PV and CSP (Section 5.3).

5.1 Private Investment in Solar Energy

This section discusses private investment in solar energy including venture capital (VC), private

equity (PE), debt, public equity, and mergers and acquisitions (M&A) (New Energy Finance

2009).

Figure 5.1 shows the tremendous rise of global investment in solar energy, particularly from

2004 to 2008. Total annual investment from 2000 through 2003 was $66 million, $144 million,

$110 million, and $417 million, sequentially, for a 3-year CAGR of 85%. This investment set the

foundation for the rapid expansion of the industry in 2004, as generous incentive programs in

Germany and Japan brought solar energy into the mainstream in both countries. Total investment

in 2005 of $2.5 billion marked a 256% increase over the $702 million of investment in 2004.

This was followed by increases of 183% to $7.08 billion in 2006, 75% to $12.38 billion in 2007,

and 31% to $16.20 billion in 2008.

Private investment in solar energy grew rapidly from 2000 to 2008, with explosive

growth occurring in the latter half of this period. From 2004 to 2008, global private-sector

investment in solar energy increased by a factor of 25 plus. Moreover, the growth in investment

has been widespread, occurring across sources, technologies, and regions. Each of the three

major sources of new investment examined here, venture capital and private equity, debt, and

public equity, grew at a compound annual growth rate (CAGR) of more than 67% from 2004 to

2008. In addition, funding to solar companies increased dramatically for different technologies,

including crystalline silicon PV, thin-film PV, and CPV, and in each of the three main regions,

the United States, European Union, and Asia.

An NREL report summarizing private investment information through 2007 was published by NREL in 2008, but

the analysis included in this section was updated by John Bartlett in 2009. The earlier report is: Jennings, C.E.;

Margolis, R.M.; Bartlett, J.E. (2008). Historical Analysis of Investment in Solar Energy Technologies (2000-2007).

National Renewable Energy Laboratory.

www.nrel.gov/docs/fy09osti/43602.pdf.

106

Figure 5.1. Global capital investments in solar energy

(New Energy Finance 2009)

The role of debt in the solar industry, which totaled $104 million in 2005, $479 million in 2006,

$1.63 billion in 2007, and $5.48 billion in 2008, has greatly increased as banks and other lenders

have become involved in financing the operations and expansions of solar companies. Greater

debt financing is a positive trend, suggesting that the perceived market and technology risks have

decreased. Furthermore, increased debt financing allows industry participants to lower their cost

of capital significantly.

Public equity offerings of solar companies were extremely limited in 2004, but in 2005

$1.74 billion of new equity was raised, followed by $4.83 billion and $7.92 billion in 2006 and

2007, respectively. In addition, the number of solar public offerings grew rapidly, with 45, 64,

and 88 offerings, sequentially, from 2005 to 2007. In 2008, the total value and number of public

solar equity offerings fell to $6.11 billion and 39, respectively, as the financial crisis deepened.

Nonetheless, the 2008 numbers still represent an enormous market expansion compared to 2004

levels.

Disclosed M&A deals raised new equity of $49.5 million, $13.8 million, $58.6 million, and

$269.5 million from 2005 to 2008, sequentially. Mergers and acquisitions deal volume surpassed

$5.3 billion in 2008, with large transactions such as Robert Bosch’s acquisition of ErSol Solar

Energy for $1.79 billion and Schneider Electric’s acquisition of Xantrex Technology for

$494 million. In M&A transactions, however, equity mostly is transferred between market

participants, and thus M&A generates comparatively little new investment for the solar sector.

Excluding government R&D and project finance investment.

$2,000

$4,000

$6,000

$8,000

$10,000

$12,000

$14,000

$16,000

$18,000

2000 2001 2002 2003 2004 2005 2006 2007 2008

Year

$ in Millions

VC & PE Investments

Solar Debt

Solar Public Equity Activity

Solar M&A Activity

Source: New Energy Finance

99% CAGR 2000-2008

74% CAGR 2000-2004

127% CAGR 2004-2008

*Excluding government R&D and project finance investments

107

Global venture capital and private-equity investment in solar totaled $539 million in 2004 and

$604 million in 2005; it then jumped to $1.76 billion in 2006, $2.78 billion in 2007, and

$4.34 billion in 2008, for a 4-year CAGR of 68% and a 3-year CAGR of 93% through 2008.

Figure 5.2 shows investments in solar energy in the United States. Following a pattern similar to

that of worldwide investment, the chart presents investment during the period 2000–2004

progressing from $77 million to $130 million, a CAGR of 14%. During the period from 2004 to

2008, investment grew at a 4-year CAGR of 133%, expanding from $130 million to about

$3.9 billion. Venture capital and private-equity investment grew fastest, from $61 million in

2004 to $2.3 billion in 2008, corresponding to a 4-year CAGR of 148%.

Figure 5.2. U.S. capital investments in solar energy

(New Energy Finance 2009)

A major theme in the recent history of solar investment is that regional differences in subsidy

programs, policies, and regulations have produced significant differences in investment patterns.

Figure 5.3 shows the value of private investment (both venture capital and private equity) in

solar by year, region, and technology on the left axis and the number of transactions by year and

region on the right axis. The regional differences in investment in solar technologies are striking.

Excluding government R&D and project finance investments.

$500

$1,000

$1,500

$2,000

$2,500

$3,000

$3,500

$4,000

$4,500

2000 2001 2002 2003 2004 2005 2006 2007 2008

Year

$ in Millions

Solar M&A Activity

VC & PE Investments

Solar Debt

Solar Public Equity Activity

* Excluding government R&D and project finance investments

Total Investment Increase

63% CAGR 2000-2008

14% CAGR 2000-2004

133% CAGR 2004-2008

Source: New Energy Finance

108

Figure 5.3. Global venture capital and private-equity investments by solar technology

(New Energy Finance 2009)

Private investments in Asia, almost nonexistent until 2005, have remained focused mainly on the

production of crystalline silicon PV. Private investments in Europe started somewhat earlier and

have focused on crystalline silicon PV and polysilicon production, with additional interest in

project developers and thin-film technologies in recent years. In contrast, U.S. private

investments have been broadly diversified, with investments in nearly all areas of the solar

industry and increasing interest in concentrating photovoltaics (CPV), next-generation PV,

concentrating solar power (CSP), and project developers. Most importantly, of the $1.52 billion

of global private investment that thin-film PV received in 2008, $1.11 billion went to U.S.-based

companies.

In terms of regional differences in private-equity versus venture-capital investments, private-

equity investment has been predominant in Europe. A great majority of private-equity

investment in the solar industry has been to finance capacity expansions (often by means of

constructing new factories), thus indicating that companies based in the European Union have

been building a majority of these factories. In contrast, venture-capital investment has been

predominant in the United States. Venture-capital investment is an indicator of new technologies

or business models. Whereas generous subsidy programs in the European Union have spurred

companies there to expand capacity rapidly, the market in the United States has not been

sufficiently attractive to enable significant growth of incumbent products. Therefore more U.S.

investment has been directed to innovative technologies with longer-term prospects.

$500

$1,000

$1,500

$2,000

$2,500

$3,000

2000

2001

2002

2003

2004

2005

2006

2007

2008

2000

2001

2002

2003

2004

2005

2006

2007

2008

2000

2001

2002

2003

2004

2005

2006

2007

2008

Asia EU* US**

Region / Year

Millions of $

Number of Transactions

Thin Film PV

Solar Heating & Cooling

Project Developer

Polysilicon

Other

Next Generation PV

Multijunction

Manufacturing Equipment

Inverters

CSP

Crystalline Silicon PV

CPV

Number of Transactions

Source: New Energy Finance

*EU includes Israel, Morocco, & South Africa

**U.S. includes Australia and Canada

109

5.2 U.S. Department of Energy Investment in Solar Energy

The U.S. Department of Energy (DOE) Solar Energy Technologies Program (SETP)

plays a

key role in accelerating the development of the U.S. solar industry and the advancement of solar

technologies. SETP investments have helped foster a diverse set of technology pathways.

SETP’s long-term objective is to achieve high market penetration of solar energy technologies

with an interim goal of achieving cost parity with conventional forms of electricity by 2015 (U.S.

DOE 2009).

SETP efforts are implemented through four subprograms: Photovoltaics (PV),

Concentrating Solar Power (CSP), Systems Integration, and Market Transformation. The PV and

CSP subprograms focus on lowering the levelized cost of solar energy through research and

development. Systems Integration focuses on technologies, tools, and strategies to optimize the

integration of solar energy into the grid. Market Transformation addresses non-R&D barriers to

achieving high market penetration of solar energy technologies.

Figure 5.4. DOE SETP budget history from FY 2001 to FY 2009

DOE SETP FY 2007 and FY 2008 research and development budgets were $157 million and

$164 million, respectively, with the FY 2007 budget representing an increase of approximately

$75 million compared to FY 2006 (see Figure 5.4). Most of the budget increase for FY 2007, and

continuing into FY 2008, was conveyed under the umbrella of the Solar America Initiative

(SAI). In addition to increased funding for PV research and development (R&D), funding in

FY 2007 and FY 2008 provided additional resources for CSP R&D, Systems Integration, and

Market Transformation. In FY 2008, funding allocated to the four subprograms was $112 million

for PV, $24 million for CSP, $12 million for Systems Integration, and $16 million for Market

Transformation. In FY 2009, the SETP budget increased again, to $175 million, comprised of

$125 million for PV, $24 million for CSP, $12 million for Systems Integration, and $14 million

for Market Transformation. In addition, the American Recovery and Reinvestment Act of 2009

DOE SETP Web site: http://www.eere.energy.gov/solar/

DOE SETP FY 2008 Annual Report: http://www.nrel.gov/docs/fy09osti/43987.pdf.

110

provided nearly$118 million

The majority of SETP funding is directed at cost-shared research, development, demonstration,

and deployment efforts with national laboratories, states, industry, and university partners. For

current, upcoming, and past funding opportunities in all research areas, see:

in additional funds to SETP, bringing the total FY 2009 budget to

$293 million. Recovery Act funding in FY 2009, as allocated to the subprograms, was $52

million for PV, $26 million for CSP, $26 million for Systems Integration, and $15 million for

Market Transformation.

http://www.eere.energy.gov/solar/financial_opportunities.html.

The DOE SETP Photovoltaics subprogram

The Concentrating Solar Power subprogram

invests in technologies across the development

pipeline that demonstrate progress toward minimizing the effective life-cycle cost of solar

energy. The PV subprogram’s activities are organized into three focus areas: new devices and

processes, prototype design and pilot production, and systems development and manufacturing.

Highlights in FY 2008 were: 1) Awarded more than $65 million for 62 industry projects

spanning early-stage to market development, addressing the challenges of scaling up novel,

low-cost manufacturing for crystalline silicon, thin film, and concentrating PV technologies; and

2) Achieved world-record efficiencies through applied research at the national laboratories,

including a 20.0%-efficient CIGS thin-film PV device and a 40.8%-efficient inverted

metamorphic multijunction solar cell.

The Systems Integration subprogram

has been ramping up R&D and deployment efforts

in recent years, leveraging industry partners and the national laboratories. The CSP subprogram

aims to increase U.S. deployment of CSP, achieve intermediate power market competitiveness

by 2015, and develop advanced technologies to reduce system and storage costs, enabling base-

load power market competitiveness by 2020. R&D activities focus on linear concentrator

systems such as parabolic troughs and linear Fresnel reflectors, dish-engine systems such as

dish/Stirling engine systems, thermal storage systems and advanced heat transfer fluids,

advanced concepts R&D, and CSP market transformation. Highlights in FY 2008 were:

1) Established 15 partnerships with universities and CSP companies to support innovations in

advanced high-temperature, heat-transfer fluids and thermal storage systems; and 2) Partnered

with the Bureau of Land Management to initiate a Programmatic Environmental Impact

Statement and conducted other joint activities necessary for the development of federal land in

the Southwest for utility-scale solar projects (also see Section 2.4.5, Land and Transmission

Constraints for Utility-Scale Solar).

The exact Recovery Act funding level was $117.6 million.

focuses on breaking down the regulatory, technical, and

economic barriers to integrating solar electricity into the electric grid by developing technologies

and strategies in partnership with utilities and the solar industry. Systems Integration R&D

includes Solar System Technology Development, Advanced Systems Integration, System

Testing and Demonstrations, Renewable Energy System Analysis, Solar Resource Assessment,

and Codes, Standards and Regulatory Implementation. Highlights in FY 2008 were: 1) Awarded

funds to 12 industry teams through the Solar Energy Grid Integration Systems project to develop

DOE SETP PV subprogram Web site: http://www.eere.energy.gov/solar/photovoltaics_program.html

DOE SETP CSP subprogram Web site: http://www.eere.energy.gov/solar/csp_program.html

DOE SETP Systems Integration subprogram Web site:

http://www.eere.energy.gov/solar/systems_integration_program.html.

111

new inverters and controllers with interfaces to energy-management systems; and 2) Established

monitoring of large-scale PV performance at high-penetration sites in California, Colorado, and

Hawaii to better understand how high levels of PV impact the grid and reduce installation costs.

The Market Transformation subprogram

promotes the commercialization of solar technologies

by addressing non-R&D barriers to solar energy adoption. Activities include codes and standards

development, outreach to state and utility decision makers, workforce development, solar

installation technical assistance, and the Solar America Cities program. DOE partners with

several organizations including the Solar America Board for Codes and Standards, the Solar

Electric Power Association, the Interstate Renewable Energy Council, the National Association

of Regulatory Utility Commissioners, the National Conference of State Legislatures, and the

Clean Energy Group. Highlights in FY 2008 were: 1) Strengthened the responsiveness,

effectiveness, and accessibility of PV codes and standards through the Solar America Board for

Codes and Standards, including release of three studies on interconnection procedures for utility

regulators, solar access laws, and external disconnect switches; and 2) Expanded the Solar

America Cities program from 13 to 25 partnerships, to further accelerate deployment of solar

energy technologies by providing financial and technical assistance to cities committed to

making solar a mainstream energy source (Figure 5.5).

Figure 5.5. 2007 and 2008 Solar America Cities

DOE SETP Market Transformation subprogram:

http://www.eere.energy.gov/solar/market_transformation_program.html

112

5.3 Solar Market Forecasts, PV and CSP

5.3.1 PV Market Forecasts

The recent expansion of the PV market has prompted numerous analysts from financial

institutions and research and consulting firms to provide analysis and forecasts for the PV sector.

This section analyzes these projections, both to identify the expected path of the industry and to

recognize the substantial variance in market forecasts. Key trends and uncertainties for the PV

market in the next several years also are discussed.

The global economic crisis that became apparent in late 2008, and which is expected to diminish

funds available for investments and reduce the demand for PV, caused some analysts to revise

their forecasts substantially in reports released in early 2009 compared to forecasts released in

mid-to-late 2008. This section focuses on the projections made in early 2009. For a detailed

discussion about the effects of the economic crisis on PV forecasts, see the full report from

which the data and discussion in this section have been extracted (Bartlett et al. 2009).

Figure 5.6 illustrates the forecasted size and composition of PV production through 2012. For

total production, the median estimate increases from 5.6 GW in 2008 to 21 GW in 2012, a 4-year

CAGR of 40%. For both the c-Si and thin-film PV segments, the growth is expected to be

greatest through 2010, with growth slowing somewhat in 2011 before accelerating again in 2012.

In addition to the growth of the median estimate, the range of estimates is significant. In 2012,

the high estimates for c-Si and thin film are roughly three times as great as the low estimates.

These uncertainties are due to different opinions about the demand for PV modules, the ability to

expand production sufficiently for each part of the PV supply chain, and the technological and

cost improvements of c-Si.

The Effects of the Financial Crisis on Photovoltaics: An Analysis of Changes in Market Forecasts from 2008 to

2009:

www.nrel.gov/docs/fy10osti/46713.pdf

113

Figure 5.6. Global total PV module production forecasts

(Mehta and Bradford 2009, Citi Investment Research 2009, Cowen & Co. 2009,

Deutsche Bank 2009, Lazard 2009, Morgan Stanley 2009, Mints and Tomlinson 2008,

Mints 2009, Chase et al. 2009, Thomas Weisel 2009)

Regarding thin-film versus c-Si production, thin-film production is expected to grow faster than

c-Si production during the next several years, with a forecasted 2008–2012 CAGR for thin-film

PV of 52% versus 33% for c-Si PV. However, c-Si is still expected to be the dominant

technology for the next several years, accounting for about 77% of total PV production in 2012.

There is, however, considerable disagreement among analysts about the future PV market share

of c-Si versus thin film; the 2012 c-Si market share estimates range from 66% to 84%.

To better describe the thin-film sector, Figure 5.7 presents the projected rise in thin-film PV

module production by technology through 2010. The range of supply estimates (i.e., the

considerable uncertainty) for thin-film PV is reasonable given that thin film faces technology and

scale-up risks in addition to overall PV market uncertainty. Despite the uncertainty, tremendous

growth is implied by the median estimates, with 2008–2010 CAGRs of 70% for CdTe, 85% for

a-Si, and 96% for CIGS production.

For CdTe, First Solar accounts for virtually all the 2008 CdTe production of roughly 0.50 GW,

and the company plans to have 1.1 GW of manufacturing capacity by the end of 2010. The

uncertainty in the 2009 and 2010 production estimates might result from divergent opinions

regarding the prospects of new CdTe entrants, such as PrimeStar Solar and Abound Solar, as

well as uncertainty regarding the pace of First Solar’s expansion.

For a-Si, there are established producers such as Energy Conversion Devices, Sharp, and

Kaneka, as well as numerous new entrants, many of which have planned to enter the market

through the purchase of turnkey systems from Applied Materials or Oerlikon. However, given

the capital expenditures necessary for the purchase of turnkey production lines, expansion of a-Si

2006 2007 2008 2009 2010 2011 2012

Ye ar

PV Production (GW-dc)

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

c-Si Market Share (%)

c-Si

Thin Film

Total

c-Si market share

Low

Median

High

HIST ORICAL

PROJ EC T ED

114

production from new entrants has been curtailed by the tight credit environment. In addition, as

PV module prices have fallen faster than system prices over the past year, non-module costs have

risen as a proportion of total system costs. Because non-module costs per watt rise as module

efficiency declines, a-Si (which has the lowest efficiency of any of the principal PV

technologies) has become less attractive.

CIGS module production is starting from a very low base, a median 2008 estimate of just

0.10 GW, but is expected to grow substantially in the near term. The median projection for 2010

is 0.40 GW, with a low estimate of about 0.10 GW and a high estimate of about 1.0 GW. The

enormous range reflects the substantial scale-up and technology risks as companies such as

Miasole, Nanosolar, and Solyndra expand commercial production.

Figure 5.7. Global thin-film PV module production forecasts

(Mehta and Bradford 2009, Citi Investment Research 2009, Deutsche Bank 2009,

Lux Research 2009)

Figure 5.8 shows the demand projections for solar PV modules by location. Global demand is

expected to grow from 5.8 GW in 2008 to 19 GW in 2012, a 4-year CAGR of approximately

34%. Europe, with Germany continuing to be the dominant market in the continent, is expected

to remain the largest region for solar power. However, the North American market is expected to

grow the fastest. Four-year CAGRs are 14% for Europe, 87% for North America, 39% for Japan

and South Korea, and 50% for the rest of the world (ROW). As with the production projections,

there is tremendous range in the demand estimates resulting from the uncertainty about policy

incentives, electricity prices, cost reductions of PV systems, and the price elasticity of PV

demand.

The United States is expected to account for a large majority of North American demand through

2012. However, analysts also expect Canada to contribute meaningfully, particularly as a result

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

2007 2008 2009 2010

Ye ar

PV Production (GW-dc)

CIGS

a-Si

CdTe

Total

Low

Median

High

HIST ORICAL

PROJEC T ED

115

of Ontario’s enhanced feed-in tariff. Mexico and other North American countries are not

expected to contribute substantially to PV demand in the near term.

Figure 5.8. Global PV module demand forecasts

(Barclays 2009, Citi Investment Research 2009, Cowen & Co. 2009, Lazard 2009,

Morgan Stanley 2009, Oppenheimer 2009, Thomas Weisel 2009, UBS 2009)

Figure 5.9 shows forecasted global module and system prices through 2010. Module prices are

expected to decrease from $3.72/W in 2008 to $2.45/W in 2010, a 2-year CAGR of –18.8%.

System prices are expected to decrease by a slightly smaller proportion, from $6.08/W in 2008 to

$4.21/W in 2010, a 2-year CAGR of –16.8%. By comparison, the average U.S. PV system cost

was $7.50/W in 2008. Although estimated non-module prices cannot be calculated directly from

Figure 5.9 (some analysts forecast only system prices and others forecast only module prices), it

is likely that the module prices are expected to decline faster than non-module prices through

2010. This is not surprising given the previous supply shortages that have kept PV module prices

from declining over the last few years despite cost improvements. Thus, there is significant room

for price declines within the PV-module cost structure.

2006 2007 2008 2009 2010 2011 2012

Ye ar

PV Demand (GW-dc)

Rest of world

Japan & S Korea

North America

Europe

Total

Low

Median

High

HIST ORICAL

PROJ EC T ED

116

Figure 5.9. Global PV module and system price forecasts

(Cowen & Co. 2009, Deutsche Bank 2009, Deutsche Bank 2009a, Lazard 2009,

Lux Research 2009, Morgan Stanley 2009, UBS 2009)

5.3.2 CSP Market Forecasts

CSP differs markedly from PV with respect to history, installation size, permitting and

construction duration, and technological readiness. Whereas PV has had a history of consistent

annual installations, 350 MW of CSP were built in the 1980s with no subsequent installations in

the United States until 2005. Installation sizes on the order of tens of megawatts, and up to

4-year permitting and construction durations, contribute to the difference in deployment patterns

between CSP and PV. Finally, only one CSP technology (parabolic troughs) has been

demonstrated long term on a fully commercial scale. As of mid-2009, two power towers were

grid-tied in Spain, but this technology type is still relatively new to the commercial market. In

addition, dish-engine systems and linear Fresnel reflectors have not yet been deployed at a near-

commercial scale. For these reasons, the scope of CSP forecasting is more limited.

Tables 5.1 and 5.2 list the amount and location of planned CSP installations through 2015; these

include systems that have been commissioned, financed, under construction, announced, or

proposed (Bullard and d’Avack 2009). Of the more than 12 GW in the CSP pipeline, more than

50% is in the United States, 33% is in Spain, about 8% is in the Middle East and North Africa

(MENA) region, and the remaining 8% is dispersed across Australasia, Europe, and the Republic

of South Africa. It should be noted that the projects in the global pipeline are by no means

guaranteed. Several major factors could prevent many of these projects from being completed,

resulting in a pipeline that will likely be reshaped on a continual basis.

Table 5.3 shows the U.S. CSP power purchase agreement (PPA) pipeline market shares by

technology, based on about 4.3 GW of PPAs (two-thirds of the 6.5 GW in the U.S. pipeline

shown in Table 5.1). In contrast to CSP installed and under construction (see Section 1.3), which

is dominated by parabolic trough technology, troughs and towers are equally represented in the

U.S. PPA pipeline.

$0.00

$1.00

$2.00

$3.00

$4.00

$5.00

$6.00

$7.00

$8.00

2007 2008 2009 2010

Ye ar

Average Selling Price ($/W-dc)

Module

System

Low

Median

High

HIST ORICAL

PROJEC T ED

117

Table 5.1. Global CSP Planned Projects,

Capacity by Country, Through 2015

Country

Capacity

(GW)

United States

6.5

Spain

4.2

India

0.46

Jordan

0.40

Israel

0.40

Italy

0.11

China

0.10

U.A.E

0.10

South Africa

0.10

Australia

0.083

Greece

0.062

Mexico

0.052

Oman

0.050

Egypt

0.030

Algeria

0.025

Morocco

0.020

France

0.012

Chile

0.010

Total

12.7

Bullard and d’Avack 2009

Table 5.2. Global CSP Planned Projects,

Market Share by Country, Through 2015

Country

Market Share

United States

51%

Spain

33%

MENA

ROW

Bullard and d’Avack 2009

Table 5.3. U.S. CSP Power Purchase

Agreement Pipeline, Market Share by

Technology, Through 2015

Technology

Capacity

(GW)

Market

Parabolic trough

1.8

41%

Tower

1.7

40%

Dish-engine

0.83

19%

Total

4.3

Bullard and d’Avack 2009

5.4 References

Barclays Capital. (2009). Solar Energy Handbook: The Second Growth Phase of Solar Era.

Barclays Capital.

Bartlett, J.E.; Margolis, R.M.; Jennings, C.E. (2009). The Effects of the Financial Crisis on

Photovoltaics: An Analysis of Changes in Market Forecasts from 2008 to 2009. National

Renewable Energy Laboratory. Technical Report NREL/TP-6A2-46713.

http://www.nrel.gov/docs/fy10osti/46713.pdf. Accessed September 2009.

Bullard, N.; d’Avack, F. (June 23, 2009). STEG Revolution III: Still Steaming. Research Note.

London: New Energy Finance.

Chase, J.; Wu, J.; d’Avack, F.; Bullard, N.; Simonek, M.. (March 31, 2009). PV Market Outlook:

It’s Darkest Before the Dawn. Research Note. London: New Energy Finance.

Citi Investment Research. (2009). U.S. Solar Stocks: Inventory Has Yet to Peak, so It’s Still Too

Early. Citi Investment Research.

Cowen & Co. (2009). Industry Outlook: Industry Hoping for Help from Stimulus Bill. Cowen &

Co.

Deutsche Bank. (2009). Solar Photovoltaic Industry: Looking Through the Storm. Deutsche

Bank.

Deutsche Bank. (2009a). Personal communication.

118

Lazard. (2009). A Framework for Assessing Solar Stocks in Turbulent Times. Lazard Capital

Markets.

Lux Research. (2009). Finding the Solar Market’s Nadir. Lux Research.

Mehta, S.; Bradford, T.; (January 2009). PV Technology, Production, and Cost, 2009 Forecast:

The Anatomy of a Shakeout. Greentech Media, Inc., and the Prometheus Institute.

Morgan Stanley. (2009). Solar Devices: Dislocation – Industry Reset. Morgan Stanley.

Mints, P.; Tomlinson, D. (2008). Photovoltaic Manufacturer Shipments & Competitive Analysis

2007/2008. Report # NPS-Supply3. Palo Alto, CA: Navigant Consulting Photovoltaic Service

Program.

Mints, P. (2009). Photovoltaic Manufacturer Shipments, Capacity, & Competitive Analysis

2008/2009. Report # NPS-Supply4. Palo Alto, CA: Navigant Consulting Photovoltaic Service

Program.

New Energy Finance (2009). New Energy Finance Desktop 3.0. www.newenergymatters.com.

Accessed May 20, 2009.

Oppenheimer & Co. (2009). PV Market Forecast. Oppenheimer & Co.

Thomas Weisel Partners. (2009). Alternative Energy; Presentation Materials to Investors.

Thomas Weisel Partners.

U.S. DOE. (2009). DOE Solar Energy Technologies Program. FY 2008 Annual Report.

http://www.nrel.gov/docs/fy09osti/43987.pdf. Accessed August 2009.

UBS. (2009). Darwin’s Theory – Survival of the Fittest. UBS Investment Research.

Solar Energy Web Sites

U.S. Department of Energy, Solar Energy Technologies Program

www.solar.energy.gov

U.S. Department of Energy, Solar America Cities Web site

www.solaramericacities.energy.gov

National Renewable Energy Laboratory (NREL) Solar Energy Technologies Program

www.nrel.gov/solar

National Renewable Energy Laboratory (NREL) Energy Analysis

www.nrel.gov/analysis

Sandia National Laboratories (SNL) Solar Technologies Department

www.sandia.gov/solar

Lawrence Berkeley National Laboratory (LBNL) Electricity Markets and Policy

eetd.lbl.gov/EA/EMP/emp-pubs.html

Database of State Incentives for Renewables and Eciency (DSIRE)

www.dsireusa.org

Interstate Renewable Energy Council (IREC)

irecusa.org

Solar Energy Industries Association (SEIA)

www.seia.org

Solar Electric Power Association (SEPA)

www.solarelectricpower.org

American Solar Energy Society (ASES)

www.ases.org

Key Report Contacts

For more information on this report, please contact:

Selya Price, National Renewable Energy Laboratory

202-488-2205, Selya.Price@nrel.gov

Robert Margolis, National Renewable Energy Laboratory

202-488-2222, Robert.Margolis@nrel.gov

On the Cover

(Main photo) At more than 8 MW of nameplate capacity, the photovoltaic (PV) power plant near Alamosa, Colorado, is one of the largest

in the United States. The Alamosa plant came online in December 2007 through a partnership between SunEdison and Xcel Energy.

– Courtesy of Zinn Photography and SunEdison/PIX 15563

(Insert) The Nevada Solar One concentrating solar power (CSP) plant near Boulder City, Nevada, is a 64 MW parabolic trough installation

completed in June 2007 by developer Acciona Energy.

– Courtesy of Acciona Energy Corporation/PIX 15996

For more information contact:

EERE Information Center

1-877-EERE-INF (1-877-337-3463)

www.eere.energy.gov/informationcenter

Printed with a renewable-source ink on paper containing at

least 50% wastepaper, including 10% post consumer waste.

Prepared by the National Renewable Energy Laboratory (NREL)

NREL is a national laboratory of the U.S. Department of Energy

Office of Energy Efficiency and Renewable Energy

Operated by the Alliance for Sustainable Energy, LLC

DOE/GO-102010-2867

January 2010

Energy Efficiency &

Renewable Energy