Photovoltaic Energy Program OverviewFiscal Year 1994

The PV Program’s accomplishments forFY 1994 are based on a comprehensive R&Dplan. The cornerstone of this R&D activity is

to develop pre-commercial prototypes of new PVproducts based on advancing fundamental sci-ence. The underlying purpose of this R&D activityis to increase the productivity of U.S. industry.

Current accomplishments are described in thePhotovoltaics Program Plan FY 1991 – FY 1995, whichunderscores the steady advance of PV technology.In FY 1994, the PV Program began developing itsnext program plan based on guidance from indus-try. The PV Program’s accomplishments duringFY 1994 took place in three key areas: improvingsystems, advancing basic technology, and expand-ing markets. In each of these areas, the Programleverages its funding by sharing the costs of pro-jects with its industrial and commercial partners.

IntroductionThe accomplishments of the PV Program in FY 1994 are partof a national plan, developed with U.S. industry, to maintainits preeminent position in PV technology, manufacturing, andworldwide market share.

FundamentalResearch

9%

Total funding: $73,476,000PV Program Funding in FY 1994

CollectorResearch andDevelopment

55%

Capital Equipment5%

AdvancedMaterialsResearch

31%

CD-SS24-A295511

The PV Program has a well-rounded plan for developing advanced PVtechnology and nourishing a competitive U.S. industry.

In FY 1994, the PV Program completed a study of the value of connecting PV generation to utility substations, such as this one nearFresno, California. This pilot PV project, sponsored in part by the Department of Energy, is the first such installation in the world.

Terry

O’R

ourk

e

Fiscal Year 1994 1

The PV industry made substantial progress inFY 1994 through participating in cooperativeR&D projects with the PV Program. First, U.S.

manufacturers substantially reduced the costs ofproducing modules under PVMaT, and began toapply the same cost-reduction techniques to com-plete PV systems. Second, the industry developedpre-commercial prototypes of new PV products forintegration into buildings under PV:BONUS.

The PV Program also provides the services of DOEnational laboratories to its commercial partners.A major highlight occurred at NREL in FY 1994,when the PV Program inaugurated the SERF.

PVMaTWorking with DOE, the PV industry made anumber of important advances through thePVMaT project in FY 1994. These advances

included reducing manufacturing costs, increas-ing productivity, increasing the capacity of manu-facturing lines, and improving the efficiency andoverall quality of PV modules.

Seven companies completed their third year ofPhase 2A PVMaT subcontracts with DOE. Duringthis time, these companies cut their productioncosts by a per-company average of nearly 50%.

Cooperative R&D with Industry

The PVMaT project, now in its fifth year, has fourseparate R&D phases. Funding for the project willtotal approximately $150 million, including $75millon from DOE.

p g y

Improving SystemsU.S. industry leads the world in driving down the cost of PV systems. ThePV Program supports the leadership role of industry through cooperativeresearch projects to improve PV systems and develop pre-commercialprototypes of new PV products.

1995

150 200

CD-SS24-A295509

100500

Production Cost andManufacturing Capacity

Aver

age

Mod

ule

Cos

ts ($

/Wp)

6

5

4

3

2

1

0

Total Manufacturing Capacity (MW)

1997

1994

1992

In a mere two years, the companies participating in the PV ManufacturingTechnology project substantially reduced production costs and increasedmanufacturing capacity. The costs are projected to steadily decline during thenext three years.

Engineers at the National Renewable Energy Laboratory of Golden, Colorado, test prototype PVmodules produced by companies participating in the PV Manufacturing Technology project.

War

ren

Gre

tz,N

atio

nalR

enew

able

Ener

gyLa

bora

tory

2 Photovoltaic Energy Program Overview

Phase 1: Twenty-two companies identified projectsto improve PV module manufacturing. The sub-contracts were completed in FY 1991.

Phase 2A: Seven companies are carrying out theirrecommendations from Phase 1 to make proprie-tary improvements and increase their manufactur-ing capabilities. The subcontracts are in theirthird and final year.

Phase 2B: Four companies continue with projectssimilar to those of Phase 2A. These 3-year subcon-tracts began in FY 1994.

Phase 3A: Two companies are addressing issuescommon to the PV industry as a whole, such asincreasing automation of manufacturing lines andlengthening PV system lifetimes. These subcon-tracts began in FY 1993.

Phase 4A: Several companies will undertake R&Dprojects emphasizing product-driven manufactur-ing in FY 1995. Under these subcontracts, U.S.companies will improve the manufacture andreduce the cost of PV products and systems. Thecompanies will seek to improve their processes formanufacturing modules and non-module compo-nents, packaging, manufacturing, assembling, andcombining components into complete PV systems.

In FY 1994, the PV Program issued a request forproposals and received a response from morethan 30 company teams. This response far exceedsthat of previous phases, indicating keen interest bythe PV industry in Phase 4 subcontracts.

Industry AccomplishmentsUnder PVMaT

Each of the companies participating in PVMaTis approaching the manufacture of cost-effectivePV modules and systems in a different way. InFY 1994, the companies participating in PVMaTsubcontracts made the following advances:

Phase 2A

AstroPower, Inc., a small business of Newark,Delaware, produces flat-plate PV modules madefrom its patented polycrystalline Silicon-Film™.Under PVMaT, AstroPower is developing both amachine to generate sheets of Silicon-Film™ andprocesses for fabricating its cells and modules ona large-area production line.

By FY 1994, AstroPower’s accomplishmentsincluded fabricating a 15-cm x 45-cm, Silicon-Film™ cell, the largest silicon solar cell ever

produced. AstroPower also increased its wafer pro-duction capacity to 4.4 megawatts (MW) per yearand reduced the cost of producing the wafers by53%. All together, it reduced module fabricationcosts by 42%.

Energy Conversion Devices (ECD), Inc., a small busi-ness of Troy, Michigan, manufactures modulesmade from thin-film a-Si. Under PVMaT, ECD isdeveloping technology for continuously manufac-turing rolls of films made from multijunction a-Si

An AstroPower, Inc., employee showcases one of the company’s new sheetsof Silicon-Film.

Ast

roP

ower

,Inc

.

Energy Conversion Devices has been manufacturing rolls of films made frommultijunction amorphous silicon alloys since 1993.

Ene

rgy

Con

vers

ion

Dev

ices

Fiscal Year 1994 3

alloys. By FY 1994, ECD had demonstrated full-scale production runs of rolls of a-Si film with atotal length of 762 m (2506 ft). The film hadexcellent uniformity and subcell yields that metECD’s quality standards of 99.7%. Uniformity isa critical property of thin films. The companyalso produced a triple-junction module, 0.37 m2

(4 ft2) in size, with a stabilized efficiency of 8.0%.Overall, ECD reduced material costs by 56%.

ENTECH, Inc., a small business of Dallas-FortWorth, Texas, produces concentrator PV mod-ules that focus sunlight onto lines of solar cellsmade from single-crystal silicon. Under PVMaT,ENTECH is increasing the durability, quality, andperformance of its concentrator modules; in-creasing capacity and automation of its manufac-turing line; and reducing costs.

By FY 1994, ENTECH had developed severalimproved production processes to reduce mate-rial and labor costs. Working with 3M Company,of St. Paul, Minnesota, ENTECH developed animproved prismatic cell cover, reducing the mate-rial and labor costs of that processing step by 90%.

ENTECH also identified a team capable of produc-ing all parts of its concentrator modules at a pro-duction capacity of more than 10 MW per year. Aspart of this team, three companies produce PVcells compatible with ENTECH’s module, with effi-ciencies greater than 19%.

Mobil Solar Energy Corporation, of Billerica,Massachusetts, produces c-Si cells and modules.Mobil Solar was recently purchased by ASEAmericas, Inc.

By FY 1994, Mobil Solar had produced thin (250-µm) silicon wafers from unique, octagon-shapedtubes, resulting in cells with an efficiencyof 13.8%. In addition, the company developed aworkstation on its production line for cuttingwafers with a laser beam to increase throughput

on the line by a factor of two. Overall,Mobil Solar reduced its module pro-duction costs by 15%.

Siemens Solar Industries, of Camarillo,California, produces PV modulesmade from single-crystal siliconcells. Under PVMaT, the company isimproving the quality of its crystallineingots, increasing the efficiency ofmaterials use through improvedsawing of Si wafers, automating signi-ficant portions of its module manu-facturing lines, and reducing theamount of hazardous waste generatedduring production.

By FY 1994, Siemens had improvedthe design of its crystal growers, result-ing in a 30% reduction in the cost ofproducing crystalline ingots and a

These concentrator PV modules produced by ENTECH, Inc., are rated at 430 W each.

EN

TEC

H,I

nc.

This new workstation developed by Mobil Solar Energy Corporation (nowASE Americas, Inc.) cuts single-crystal silicon into wafers with a laser beam.

Mob

ilS

olar

Ener

gyC

orpo

ratio

n

4 Photovoltaic Energy Program Overview

savings of $300,000 per year. Furthermore, thecompany concentrated the waste streams, result-ing in a 10% reduction in waste volume and a 20%reduction in waste-handling costs.

Solarex Corporation, of Rockville, Maryland, manu-factures large-area, multijunction a-Si alloy mod-ules. Multijunction modules are constructed inlayers, each one tailored to absorb a different por-tion of the solar spectrum. Under PVMaT, Solarexis improving the quality of components and opti-mizing manufacture of its modules.

By FY 1994, Solarex had developed a monolithic(made in a single piece), multijunction, a-Si

module, 3700 cm2 (4 ft2) in size, with an initial effi-ciency of 8.9%. The company started commercialproduction of these modules and plans to increaseproduction to 10 MW per year.

Utility Power Group (UPG), a small business ofChatsworth, California, and its subcontractor,Advanced Photovoltaic Systems, Inc. (APS), asmall business of Lawrenceville, New Jersey, manu-facture single-junction a-Si alloy modules. UnderPVMaT, these companies are increasing moduleperformance and reducing manufacturing costs.UPG is increasing production capacity, and APSis developing two new module prototypes.

By FY 1994, UPG had substantially streamlined itsmodule-production processes. For example, thecompany eliminated 30 steps from the process ituses for encapsulating and connecting the electri-cal leads of the module, reducing the cost of thisproduction step by 81%. Overall, UPG reducedmodule production costs by 28% and increasedits manufacturing capacity by a factor of five.APS developed prototypes of new modules andreduced processing time for producing modules,resulting in a significant reduction in cost.

Solarex Corporation plans to increase its capacity to manufacturemultijunction modules made from amorphous silicon to 10 MW per year.

Sol

arex

Cor

pora

tion

Siemens Solar Industries is increasing the efficiency of materials use and automating itsproduction lines through participation in the Photovoltaic Manufacturing Technology project.

Sie

men

sS

olar

Indu

strie

s

The Utility Power Group produces single-junction amorphous silicon modules.

Util

ityP

ower

Gro

up

Fiscal Year 1994 5

Phase 2B

Golden Photon, Inc., of Golden, Colorado, is devel-oping a new PV technology consisting of modulesmade from cadmium telluride (CdTe) material.This promising thin-film technology can substan-tially reduce the cost of PV modules. DuringFY 1994, the company completed design of itsmanufacturing line with a production capacity of2 MW per year.

Solar Cells, Inc. (SCI), a small business of Toledo,Ohio, is developing a high-throughput manufac-turing line of PV modules also made from CdTe.During FY 1994, SCI improved the mechanism forconveying raw materials and controlling pressureand temperature during processing. Overall, SCIincreased production capacity by a factor of 100.

Solarex Corporation’s, Crystalline Silicon Division isadvancing its technology for manufacturing PVmodules made from cast-ingot polycrystalline sili-con. During FY 1994, Solarex engineers modifiedthe design of one casting station, resulting in a73% increase in capacity, and began using a newwire saw for cutting wafers from cast ingots.

Texas Instruments, Inc., (TI), of Dallas-Fort Worth,Texas, is developing its Spheral Solar™ PV tech-nology. TI’s innovative approach to module designuses small crystalline silicon spheres bonded toaluminum foil. During FY 1994, TI engineersdeveloped equipment capable of producing sev-eral MW per year, demonstrated cell yields as highas 90% on a pilot production line, and produceda pilot PV module, 4000 cm2 in size, with an effi-ciency of 8.3%. (TI closed its PV operationsJanuary 26, 1995.)

Phase 3A

Spire Corporation, a small business of Bedford,Massachusetts, is developing automated equip-ment that assembles and solders silicon solar cellsinto cell strings. Forming the strings is an interme-diate step in producing PV modules. The equip-ment can be programmed to handle cells ofdifferent sizes and thicknesses from various manu-facturers. By using this equipment, module manu-facturers will be able to reduce handling of cells,speed processing, and thus reduce overall costs.

In FY 1994, Spire developed and conducted dem-onstration tests of a prototype machine using sili-con cells from several manufacturers. Spire willtest the machine in a production setting inFY 1995.

Springborn Materials Science Corporation, of Enfield,Connecticut, is improving encapsulants for PVmodules. In the past, some encapsulants madewith ethylene vinyl acetate prematurely aged whenexposed to ultraviolet (UV) insolation.

During FY 1994, Springborn scientists discoveredsome glass products used by PV module manufac-turers transmit significantly less UV radiation thannormal glass. Through accelerated UV testing,the scientists determined that the new glasssignificantly reduces premature aging of theencapsulants.

PV Integrated Into BuildingsIn addition to helping U.S. industry improve itsmanufacturing capabilities, the PV Program ishelping position U.S. industry to profit from thehuge potential market for PV-integrated buildings.For example, PV Program analysts predict a mar-ket on residential buildings of 20,000 MW by 2020.

PV can be integrated into buildings a number ofways. For example, modules can be mounted on

These modules, produced by Siemens Solar Industries, are mounted on the roof of theSolar Energy Research Facility in Golden, Colorado.

War

ren

Gre

tz,N

atio

nalR

enew

able

Ene

rgy

Labo

rato

ry

6 Photovoltaic Energy Program Overview

the roof, or constructed as part of the roof orbuilding facade.

PV can also be incorporated into new, innovativeproducts. For example, semi-transparent thin-filma-Si can be coated onto glass using manufacturingprocesses similar to those used to coat glass forcommercial buildings. PV-coated glass could beused for skylights or other locations where coated,semi-transparent glass would otherwise be used.

When PV is used in this way, it becomes part of thearchitectural structure of the building. InstallingPV roofs and PV windows is more cost-effectivefor the builder, and can decrease the net cost ofenergy from PV.

PV:BONUS

The PV Program is developing four prototypeproducts for PV-integrated buildings under thePV:BONUS project, including: PV roofing, PVmodular homes, alternating current (ac) modulesfor buildings, and utility-dispatchable PV systems.Each of these products is being developed by ateam of firms with a range of buildings- and PV-related expertise. In addition, NREL distributed anewly developed curriculum to architecturalschools dealing with PV-integrated buildings.

PV:BONUS is administered by DOE’s GoldenField Office with technical support from NREL.

PV Roofing

A team of eight firms, led by ECD, is developingPV systems based on ECD’s a-Si technology thatdirectly integrates into conventional roofing mate-rial. The team is developing PV roofing materialsfor two types of roofs—shingles and flat metalroofs. In FY 1994, the team completed construc-tion of the prototype roofs and began outdoorperformance testing at NREL.

In FY 1994, one member of the PV-roofing team,the National Association of Home Builders, begandesigning a prototype townhouse made with aPV-metal roof. The association plans to constructthe prototype at its research center in UpperMarlboro, Maryland, beginning in FY 1995.

PV Modular Homes

Another team, led by Fully Independent Residen-tial Solar Technology, Inc. (FIRST), of Hopewell,New Jersey, is developing modular homes with PVmodules incorporated into the roofs. Modularhomes, mostly constructed in a factory and

shipped to the site for final assembly, are becom-ing increasingly popular because they cost lessthan homes built entirely on-site by constructioncontractors.

In FY 1994, the FIRST team constructed andbegan testing two prototype homes in Ottsville,Pennsylvania. In FY 1995, the team anticipatesconstructing pre-production models of PV modu-lar homes.

AC Modules

Almost all appliances use ac electricity exclusively;however, existing PV modules produce direct cur-rent (dc) electricity, requiring power-conditioningunits to convert the dc to ac electricity.

Existing dc modules create several constraints inbuilding design, including the requirement fordc wiring in parts of the building and ac wiring inother parts; the need to size inverters based onprojected electrical loads and locate them some-where in the building; and the limitation on theshape, orientation, and number of PV modulesused in the building. An ac module would free

The PV:BONUS Team Developing PV Roofing

Energy Conversion Devices, Inc., a PV cell manufacturer

United Solar Systems Corporation, a PV module manufacturer

National Association of Home Builders of Upper Marlboro, Maryland, a buildingtrade association

Solar Design Associates of Harvard, Massachusetts, an architectural and systemsengineering firm

Arizona Public Service Company, of Tempe, Arizona, an electric utility

Bechtel Corporation, of San Francisco, California, an engineering and constructionfirm

Detroit Edison Company, of Detroit, Michigan, an electric utility

Minoru Yamasaki Associates, of Rochester Hills, Michigan, an architectural firm

Team Developing PV Modular Homes

Fully Independent Residential Solar Technology, Inc., ofHopewell, New Jersey, a nonprofit organization

Bradley Builders of Tempe, Pennsylvania, a home builder

AvisAmerica of central Pennsylvania, a builder of modular,manufactured homes

Fiscal Year 1994 7

architects and building designers from suchconstraints.

In FY 1994, a PV:BONUS team led by SolarDesign Associates, a small business of Harvard,Massachusetts, produced five pilot inverters forthe ac module. The inverters, the key electricalcomponent of the modules, are constructed intosmall junction boxes that fit inside the modules.The ac-module team began testing the pilotinverters for performance in laboratories at theSolarex manufacturing plant in Rockville,Maryland.

Dispatchable Peak-Shaving PV

Also under PV:BONUS, Delmarva Power andLight Company, of Wilmington, Delaware, is devel-oping PV systems with batteries that can be calledupon to supply power dispatched by a utility. Theutility would dispatch these systems during its peakperiods—when it is experiencing the greatestdemand for power from its customers.

The Delmarva Power team designed and installeda prototype dispatchable system on the roof ofone of the utility’s buildings in FY 1993. Theteam tested performance of the system during theutility’s peak demand periods, which take placeduring the summer. Based on its performance,the team improved the design of the system inFY 1994, and performed a detailed market studyfor these systems inside Delmarva Power’s serviceterritory.

New Curriculum for Architecture Schools

In FY 1993, NREL, the American Institute ofArchitects and Associated Collegiate Schools ofArchitecture Research Corporation of Washington,D.C., developed a course for architecture schoolson how to incorporate PV into building designs.The corporation tested a pilot version at severalarchitecture schools in FY 1994 and subsequentlydistributed them to 50 schools in North America.In FY 1995, it will modify the course into a formthat will be useful to practicing architects.

A prototype modular home, partially constructed in a factory and assembled on site atOttsville, Pennsylvania, in FY 1994, has PV modules comprising part of the roof.

Lyle

Raw

lings

Team Developing the ac Module

Solar Design Associates, an architectural and systemsengineering firm

Solarex Corporation, a PV module manufacturer

ASE Americas, Inc. (formerly Mobil Solar Energy Company),a PV module manufacturer

Steve Strong, president of Solar Design Associates, exhibits a prototypealternating-current module incorporated into a window-wall developed underthe PV:Building Opportunities for PV in the United States project.

Sol

arD

esig

nAs

soci

ates

8 Photovoltaic Energy Program Overview

New Laboratory CapabilitiesIn FY 1994, the PV Program inaugurated newlaboratory capabilities at Sandia and NREL thatincrease the services provided to industry.

Solar Energy Research Facility

First among the new facilities was the SERF,opened at NREL in FY 1994 to support PV R&D.Throughout the year, NREL moved 37 laborato-ries and 170 staff members into the facility withminimal impact on program projects. For exam-ple, researchers from the PV Device PerformanceLaboratory missed only 5 working days as a resultof the move. The laboratory is a world-renownedfacility for standard measurements of solar celland module efficiencies.

In addition to housing state-of-the-art laboratories,the SERF is a model example of energy-efficientconstruction. Design features include daylighting,on-demand lighting, indirect evaporative cooling,variable-speed fan motors, passive solar design,and PV modules integrated into the roof of thebuilding.

Power Processing Laboratory

In FY 1994, Sandia reactivated its power-processingtest capability as part of the PV Systems EvaluationLaboratory. Projects under way at the laboratoryinvolve improving the performance of such com-ponents as charge controllers for PV-battery sys-tems, trackers that allow PV modules to follow thesun during the day, and inverters that convert dcto ac electricity. The projects are designed toimprove the performance and reduce the costof producing the balance-of-systems componentsconnected to PV systems.

In FY 1994, Sandia engineers began a project atthe laboratory to develop inverters in support ofthe U.S. Department of Defense procurement for

mid-size PV systems. In order to run acceptancetests for these systems, the engineers designed atest bed for benchmark testing of the inverters.Benchmark testing involves evaluating hardwaresuch as inverters from different manufacturersunder a single set of performance parameters.The engineers can use this information to helpprocure the best inverter for a particular project,

Vice President Al Gore visited the newly inaugurated Solar Energy Research Facility (SERF)at the National Renewable Energy Laboratory in FY 1994. The SERF houses the premierfacilities in the nation for PV research and development.

War

ren

Gre

tz,N

atio

nalR

enew

able

Ene

rgy

Labo

rato

ry

Companies Evaluating Inverters atSandia in FY 1994

Abacus Controls, Inc., of Somerville, New Jersey

AES/Skyline Engineering, of Temple, New Hampshire

Omnion Power Engineering Company, of Mukwonago, Wisconsin

Pacific Inverters, of Spring Valley, California

Trace Inverters, of Gilington, Washington

An engineer runs benchmark tests of an inverter at Sandia NationalLaboratories’ new Power Processing Laboratory.

San

dia

Mat

iona

lLab

orat

orie

s

Fiscal Year 1994 9

or to help direct research toward areas where allmanufacturers need to improve.

As a result of the benchmark tests in FY 1994,Sandia engineers identified two areas for improve-ment in FY 1995: reducing radio frequency inter-ference (RFI) and reducing noise.

Although RFI from most inverters was within Fed-eral Communications Commission standards,some of them could interfere with nearby broad-cast communications equipment such as televi-sions. In FY 1995, Sandia engineers will work withinverter manufacturers to design better RFI filtersto reduce interference and noise during opera-tion.

Outdoor Test Facility

In FY 1994, NREL completed design and beganconstruction of a new Outdoor Test Facility forperformance and exploratory testing of PV mod-ules. Testing modules and small PV systems out-doors at such a facility under field operatingconditions is important for module manufacturersand potential users to verify performance.

One of the projects already under way at the Out-door Test Facility is the establishment of a stand-ardized performance energy rating testing (PERT)system. Such a system is useful for the PV industrybecause it allows module performance to be com-

pared based on standardized tests. For example,the tests help users determine the amount of en-ergy a module will actually deliver over time. InFY 1994, NREL engineers began establishing aPERT data base of existing modules and estab-lished an industry review committee to develop cri-teria and methods for rating energy delivery ofmodules.

The project underscores the technical basis forthe PV Program’s support of industry becausesuch methods form the basis for recommendedpractices and standards in the industry. Engineersthen use these practices and standards to specify,procure, and install systems in the field.

Program Services for Industry

Both NREL and Sandia have made additional fa-cilities and services available to the private sector,including:

• Measurement and Performance

Both NREL and Sandia have laboratories thatperform thousands of state-of-the-art measure-ments for the PV industry. These include analyz-ing materials, characterizing devices, evaluatingfabrication problems, and modeling perform-ance with computers. In FY 1994, for example,program scientists performed more than17,000 characterizations of cells and modules

In FY 1994, the National Renewable Energy Laboratory connected these modules at its newOutdoor Test Facility to the local utility, Public Service Company of Colorado. The modules,made from amorphous silicon, were produced by United Solar Systems Corporation, of Troy,Michigan.

War

ren

Gre

tz,N

atio

nalR

enew

able

Ener

gyLa

bora

tory

In FY 1994, the National Renewable Energy Laboratory opened the Clean Room at the SolarEnergy Research Facility to incorporate advances in semiconductor science into PV technology.

War

ren

Gre

tz,N

atio

nalR

enew

able

Ener

gyLa

bora

tory

10 Photovoltaic Energy Program Overview

for more than 100 research groups in industry,universities, and other research organizations.

• Materials and Processing

NREL has facilities for crystal growth, thin-filmdeposition, and cell component and materialsprocessing development. The material anddevice research capabilities extend from crystal-line silicon to thin-film materials (for example,a-Si, copper indium diselenide [CIS], andCdTe) to gallium arsenide and other III-Vmaterials.

Sandia’s Photovoltaic Device Fabrication Labo-ratory (PDFL) is available to integrate new c-Sisolar cell designs and fabrication processes intomanufacturing facilities.

Program scientists produce advanced crystalline silicon cells using processingmethods similar to those prevalent in today’s PV industry at Sandia NationalLaboratories’ Photovoltaic Device Fabrication Laboratory.

San

dia

Nat

iona

lLab

orat

orie

s

Fiscal Year 1994 11

Advances in PV science continued in FY 1994through research encompassing both crystal-line and thin-film materials.

Crystalline SiliconAbout 95% of today’s commercially available PVmodules are made from c-Si cells—both single-crystal and multicrystalline silicon.

Single-Crystal Silicon

In FY 1994, researchers at the University of NewSouth Wales (of Australia) fabricated a new cellwhich broke the 4-year-old efficiency record for aone-sun single-crystal silicon cell. The cell, calledthe Passivated Emitter, Rear Locally Diffused cell,has an efficiency of 24.0%, the highest efficiencyof any silicon cell yet recorded. It is fabricated on10-cm-diameter wafers made from float-zone sili-con, the highest quality and purest form of silicon.Sandia has supported development of the designof this cell for several years.

In another project using float-zone silicon, HondaR&D Company Ltd., of Siatama, Japan, incorpo-rated c-Si cells produced by SunPower into apower module with an efficiency of 21.6%. This isthe highest efficiency of any flat-plate module yetproduced. SunPower provided the back-contactcells for the module, the cells for which it sharedthe prestigious R&D 100 Award in FY 1994 withSandia. By producing a c-Si module with an effi-ciency greater than 20%, Sandia passed an impor-tant research milestone of the PV Program.

Sandia scientists are also working with solar-gradeCzochralski (Cz) silicon, which is slightly less purethan the float-zone-type silicon. Cz silicon is, how-ever, representative of that used by c-Si cell manu-facturers. Improvements in production processes

using Cz silicon can be quickly incorporated intothese companies’ production lines.

In FY 1994, engineers at Sandia’s PDFL produceda cell, 4 cm2 in size, using Cz silicon supplied bySiemens. At 18.3%, this cell efficiency is morethan 1% higher than that of the best cell pro-duced by Siemens, and provides directions forimproving the company’s processes.

Also in FY 1994, Siemens and Sandia began a col-laborative project to fabricate thin, single-crystalcells (250 µm). Producing thinner cells wouldreduce manufacturing costs through more effi-cient use of single-crystal material and reducedhandling.

Also in FY 1994, Georgia Tech scientists fabricateda cell produced from float-zone silicon with an effi-ciency of 16.9%, using “rapid thermal processing.”Georgia Tech scientists produced these cells usingtwo short processing steps, each less than 10 min-utes, carried out at a temperature of 880°C. At thistemperature, they can perform several processingsteps at once, which can eventually lead to reduc-ing the time and cost of producing the cells. Ulti-mately, these scientists hope to fabricate such cellsin an hour or less. In FY 1995, the scientists willbegin applying these techniques to producingmulticrystalline silicon cells.

Multicrystalline Silicon

Multicrystalline cells can be produced and proc-essed at a lower cost than single-crystal cells be-cause carefully producing a large, single crystal ofsilicon is not required. In FY 1994, PV Programresearchers produced new cells and modules withrecord efficiencies and faster, lower cost produc-tion processes.

Advancing PV TechnologyPV Program scientists and their research partners in industry anduniversities expanded the understanding of the fundamental mechanismsgoverning the formation and performance of PV materials.

12 Photovoltaic Energy Program Overview

In FY 1994, scientists at Sandia designed and builta prototype module made with multicrystalline sili-con with an efficiency of 15.3%. This is the firsttime a module with 15% efficiency has ever beenmade with this material.

The Sandia engineers produced the module usingcell-processing techniques similar to those devel-oped at the Georgia Institute of Technology(Georgia Tech), of Atlanta, Georgia. DOEdesignated Georgia Tech as a University Centerof Excellence for PV Research and Education. InFY 1994, Georgia Tech researchers produced asmall-area multicrystalline silicon cell with an effi-ciency of 17.8%, the highest efficiency recorded todate for research cells made from this material.

The cell produced by Georgia Tech and the mod-ule produced by Sandia used high-quality multi-crystalline silicon produced by Crystal Systems,Inc., a small business of Salem, Massachusetts. InFY 1994, engineers from Crystal Systems and scien-tists from NREL and Sandia improved the heatexchanger method used to fabricate the com-pany’s multicrystalline silicon. Using this method,heat flow is carefully controlled during processing

so that the silicon solidifies along a nearly planarsolid-liquid interface. The resulting material ismade up of grains relatively large in size and withsuperior electrical properties.

Basic Science

The PV Program also developed advanced scien-tific and measuring equipment to further advancePV technology. One example of high-technologyequipment developed by PV Program scientists isthe Scanning Defect Mapping System (SDMS),now being developed as a commercial product byU.S. industry.

The SDMS, designed and built originally atNREL, won the R&D 100 Award in FY 1993.The SDMS accurately and quickly—in less than1 hour—measures the number and type of crystal-line defects in a single-crystal or multicrystallinesilicon wafer. Before the SDMS was built, scientistsused laborious processes to characterize thecrystalline properties of silicon, taking as muchas 2 weeks for a sample.

NREL joined forces with Labsphere, a small busi-ness of North Sutton, New Hampshire, to licensethe SDMS. The company will market the SDMS tocompanies working with silicon in the PV andsemiconductor industries to monitor and controlquality in their production processes. Labspherelicensed the technology and produced a

In FY 1994, Labsphere began commercial production of this Scanning Defect MappingSystem. The system, originally developed at the National Renewable Energy Laboratory,won the prestigious R&D 100 Award in FY 1993.

War

ren

Gre

tz,N

atio

nalR

enew

able

Ene

rgy

Labo

rato

ry

In FY 1994, Sandia National Laboratories researchers built the firstprototype multicrystalline silicon module with an efficiency greater than15%.

San

dia

Nat

iona

lLab

orat

orie

s

Fiscal Year 1994 13

pre-commercial prototype in FY 1994; the com-pany plans to begin commercial production andsales of the machine in FY 1995.

Concentrator Cells and ModulesIn FY 1994, the PV Program provided assistance tocompanies developing concentrator PV modulesand produced new cells for concentrator modules.Concentrator modules use optical lenses to focussunlight on smaller, highly efficient PV cells.

New Modules

The Solar Energy Applications Corporation(SEA), of San Jose, California, installed concentra-tor modules into four arrays under a commercialcontract with the Sacramento Municipal UtilityDistrict (SMUD), of Sacramento, California, inFY 1994. SEA uses low-cost acrylic concentratorswith a low concentration ratio (approximately 10suns)—the area of the concentrators divided bythe area of the cells, measured in “suns.”

In addition, ENTECH produced the largestPV module in the world in FY 1994 under thePVMaT project. The prototype module has aconcentration ratio of approximately 20 suns andhas a rated power output of 430 W.

The company installed four arrays in FY 1994under a commercial contract with two utilities inTexas: one array in Fort Davis, Texas, with Centraland South West Services, Inc., of Dallas, Texas,and three arrays with Texas Utilities of Dallas,Texas. Each has a rated capacity of 25 kW.

Both ENTECH and SEA developed several genera-tions of prototypes under PVMaT and the Concen-trator Initiative during FY 1992 – FY 1994. SEA’ssubcontract to Sandia under the ConcentratorInitiative is scheduled to be completed in FY 1995.

Record-Setting Cells

Cells used in concentrator modules are differentthan those used for flat-plate modules in severalways. Because there are fewer of them in a singlemodule, higher quality, more efficient cells canbe used more economically. They also provide ahigher power output per cell.

In FY 1994, NREL scientists produced a small-area(0.1 cm2) concentrator cell made from (GaInP2/GaAs). The cell had an efficiency of 30.2% undera concentration ratio of 180 suns. By producing

a cell with an efficiency greater than 30%, thescientists passed an important research milestoneof the PV Program.

The monolithic, two-junction cell was producedby growing epitaxial layers of GaInP2 immediatelyafter the growth of a conventional crystalline GaAscell. The scientists carefully chose the materialsbecause their crystalline lattice structures arematched (to improve the monolithic crystal qual-ity) and because they absorb a different portion ofthe solar spectrum.

Compared to a simple, single-junction GaAs cell,the efficiency of this cell is almost 20% higher.However, because of the monolithic design of thiscell, processing takes only a few minutes longer.

In FY 1994, Applied Solar Energy Corporationand Spectrolab, both of Los Angeles, California,announced plans to make this cell for applicationsin space.

Thin Films

Thin-film PV consists of modules made of verythin layers of PV materials. The layers are usuallyabout 1µm–10 µm, or about 30 to 100 times thin-ner than cells and modules made from wafers. Inaddition to using small amounts of material, these

By producing a concentrator cell with an efficiency greater than 30%, National RenewableEnergy Laboratory scientists passed an important milestone of the PV Program.

War

ren

Gre

tz,N

atio

nalR

enew

able

Ener

gyLa

bora

tory

14 Photovoltaic Energy Program Overview

modules can be manufactured in a single series ofsteps (monolithically) with less processing andlower costs than conventional PV modules.

The PV industry and the PV Program bothmade significant strides in advancing thin-filmtechnology in FY 1994. Two thin film companiespassed the milestone of producing 10%-efficientprototype modules. By the end of FY 1994, at leastfive U.S. companies were either building manufac-turing plants or planning to do so.

In FY 1994, PV Program scientists developedadvanced methods of producing thin films andachieved record efficiencies in laboratory cells sig-nificantly higher than previous records. They alsoproduced thin films that are stable in outdooroperating conditions, and furthered scientificunderstanding of the physics involved in the insta-bility of samples produced with existing methods.The advances took place both in a-Si and polycrys-talline thin films—CdTe and CIS.

Amorphous Silicon

Two important advances took place in a-Si technol-ogy in FY 1994. First, an a-Si manufacturer, UnitedSolar Systems Corporation, produced a prototypemodule with a stabilized efficiency of 10.2%. Thisis the first a-Si module yet recorded with a stabi-lized efficiency of more than 10%.

Second, NREL scientists used a novel method toproduce a-Si material samples whose efficiency isstable when exposed to sunlight. The currentmethod for producing a-Si materials is a relativelylow-temperature (250°C) vacuum process calledglow discharge. When a-Si modules produced withthis method are exposed to sunlight, their per-formance degrades. For this reason, performancevalues for a-Si modules are usually quoted in termsof “stabilized efficiencies.”

In FY 1994, NREL scientists produced samples ofa-Si materials using a hot-wire technique. The hotwire, which is similar to that used in a tungsten fila-ment lamp, produces temperatures of about2000°C near the wire. Scientists then exposed thesample to lamps in a laboratory for 120 hours. Pre-liminary results of the test show no degradation ofperformance. For comparison, the performanceof similar samples made with the glow-dischargemethod used in this test degraded 16%.

Scientists believe the higher temperatures pro-duce a-Si with superior electrical characteristics(measured in terms of reduced defect densities).

This research team from the National Renewable Energy Laboratory won the “Best ofWhat’s New” award given by Popular Science in FY 1994 for producing a thin-filmcopper indium diselenide cell with a record efficiency and a potential for very-low-costproduction.

War

ren

Gre

tz,N

atio

nalR

enew

able

Ener

gyLa

bora

tory

Perc

ent o

f Ini

tial E

ffici

ency

(%)

100

90

80

70

CD-SS24-A295507

a-Si Produced With a Hot Wire

a-Si Produced by Glow Discharge

Hours0 50 100 150

Light Induced Degradationin Amorphous Silicon

Scientists at the National Renewable Energy Laboratory used a hot-wire techniquein FY 1994 to produce an amorphous silicon PV device that remained stable whenexposed to simulated sunlight in a laboratory.

Fiscal Year 1994 15

In addition, the amount ofhydrogen that remains in thea-Si material produced by ahot wire is low, about 2% ofthe total. By comparison, a-Simaterial produced by the glow-discharge method consists ofabout 10% hydrogen.

NREL scientists have hypothe-sized for several years thatthe presence of too muchhydrogen causes instability.In FY 1994, scientists at theUniversity of Oregon con-firmed independently that a-Simaterial with a low percentageof hydrogen (produced by adifferent method) remainsstable when exposed tosunlight.

In FY 1995, NREL scientistswill conduct more tests andproduce a-Si samples with ahot-wire that have higherabsolute efficiencies.

Polycrystalline Thin Films

The PV Program is workingto develop two polycrystallinethin films, CdTe and CIS,because performance of labo-ratory cells has been improv-ing rapidly in recent years.Furthermore, NREL beganperformance testing inFY 1994 of the first arrays ofCdTe and CIS modules.

This performance recordcontinued in FY 1994, whenNREL scientists produced asmall-area cell in the labora-tory made from copperindium gallium diselenide

with an efficiency of 16.8%. This is the highestefficiency yet recorded for any thin-film cell.

Also in FY 1994, scientists at the University ofIllinois at Champaign-Urbana produced a single-crystal sample of CIS in the laboratory. Havingsingle-crystal CIS samples available for analysis willallow for the orderly, scientific characterizationof the properties of CIS in the future.

Max

imum

Sm

all-A

rea

Cel

l Effi

cien

cies

(%)

17

15

13

11

9

7

51976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996

CD-SS24-A295505

Univ. of Maine

Boeing

Boeing

EuroCIS

NREL

ARCO

Record CIS Cells

Scientists at the National Renewable Energy Laboratory developed small-area copper indium diselenide cells in FY 1994 withefficiencies of almost 17%, more than double those of a decade ago. These cells have the highest efficiencies of any thin-filmcells in the world.

Max

imum

Sm

all-A

rea

Cel

l Effi

cien

cies

(%)

17

15

13

11

9

7

51976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996

CD-SS24-A295506

Matsushita

Kodak

BP Solar

AMETEK

Photon Energy

Univ. ofSo. Florida

Univ. ofSo. Florida

Monosolar

Record CdTe Cells

In the last several years, there has been a remarkable increase in performance of small-area cadmium telluride cells in thelaboratory measured at standard conditions.

16 Photovoltaic Energy Program Overview

Industry progress in FY 1994 has been equally re-markable. Siemens Solar Industries produced thefirst CIS power module, 3892 cm2 (4 ft2) in size,with an efficiency of 10.3%. This is also the high-est stable efficiency for a thin-film module to date.

Siemens shipped a 1-kW array of CIS modules toNREL to begin performance testing outdoors inFY 1994. Similarly, SCI shipped a 0.4 kW arraymade from CdTe modules. The details of the per-formance tests are provided to the individual com-panies; nevertheless, the output of the modulesremains stable in outdoor environments.

NREL scientists verified in FY 1994 that modulesproduced by SCI had a power output of 55.4 Wand an efficiency of 8.1%, a record for this mate-rial. Several other SCI prototype modules under-went performance testing outdoors at NRELthroughout FY 1994. So far, output has remainedstable.

The Thin-Film PV PartnershipProgram

In FY 1994, DOE announced the formation of apartnership program to support development ofthin-film PV. The 5-year program, called the Thin-Film PV Partnership Program, will be funded forabout $180 million. DOE will provide about two-thirds and private industry will provide the rest.

The purpose of the thin-films partnership is to de-velop a continuous stream of pre-commercial pro-totype products. These prototypes start as ideas inresearch laboratories and end as fully tested tech-nologies ready for private industry to license anddeploy in the marketplace.

NREL began negotiating the first series of 11 con-tracts in FY 1994. The contracts, worth $29 mil-lion, are with six commercial partners in industryand five research partners in universities. Of thesix commercial partners, four are small businesses.Each of the partners will supplement 10%–50% ofthe value of these contracts with its own funds.

The reason for combining all thin-film productR&D under one umbrella is because the thin-filmPV industry is making particularly good progressin commercial development. By the end ofFY 1994, at least five U.S. companies were eitherbuilding new manufacturing plants or planning todo so.

Furthermore, many analysts believe thin filmshold a leading edge for the future of very-low-costPV. By combining the talents of scientists perform-ing fundamental research with engineers develop-ing prototype modules under one organization,the PV Program will help participating companiesposition themselves to be market leaders in the21st century.

ac Power Output

CD-SS24-A295508

Jul 94 Aug 94 Sep 94 Oct 94 Nov 94

Pow

er(W

atts

)500

300

100

The alternating current power output of a 0.4-kW array cadmium telluride moduleproduced by Solar Cells, Inc., remained stable during performance testing at theNational Renewable Energy Laboratory in FY 1994.

Golden Photon, Inc., is completing construction of a facility for manufacturing cadmiumtelluride modules in Golden, Colorado. The Thin-Film PV Partnership Program will supply acontinuous stream of advanced thin-film technology to companies such as Golden Photon toincorporate into their manufacturing facilities.

Gol

den

Pho

ton,

Inc.

Fiscal Year 1994 17

Sandia Shares R&D 100 Award

For the fourth year in a row, the PV Program received the R&D 100Award in FY 1994. Every year, R&D magazine designates the awardsfor the 100 most important scientific advances of the year. This year’saward was for supporting development of a high-efficiency c-Si cellcalled the back-contact cell.

The unique design of this c-Si cell leads to its high efficiency. Unlikeconventional cells, which have metallic grids on the front and rearsurfaces of the cell, all of the electrical contacts on this unique cell areat the back surface of the cell.

Another unique design feature is the construction of the electrical fieldinside the cell. Unlike conventional cells in which an electric field isformed by doping different layers of the c-Si with minute quantities ofphosphorus and boron, large numbers of highly doped regions withmuch smaller junctions between them are on the back of this cell.

As a result, the back-contact cell has increased efficiency for atleast two reasons. First, there is no top grid to prevent sunlightfrom entering the portion of the cell that is covered by the grid.Therefore, more charge carriers—oppositely charged electronsand holes—are generated in the cell. Second, the contacts arecloser to where the charge carriers are generated. As a result,these charge carriers have less distance to travel and lessopportunity to recombine. More charge carriers are collected inthe electrical contacts, more electricity is generated, and theefficiency is higher.

DOE shares the award with SunPower, the Electric PowerResearch Institute (EPRI), of Palo Alto, California, Sandia, andAmonix, Inc., of Torrance, California.

SunPower Corporation has entered into a commercial agreement with Honda to incorporate its back-contact cell into a new commercial module. Aprototype module, pictured here undergoing performance testing at Sandia National Laboratories, has a record efficiency of 21.6%.

San

dia

Nat

iona

lLab

orto

ries

18 Photovoltaic Energy Program Overview

The PV Program inaugurated a numberof projects in FY 1994 to demonstratethe value of PV systems for several

important market segments by verifyingtheir performance. The projects includedomestic utilities, exports, and governmentapplications for the U.S. Department ofDefense and the National Park Service.

DOE sponsors only part of each of thesedevelopment projects with other agenciesand private companies, thus effectivelyleveraging program funds. DOE partici-pates in these projects as part of a nationalplan put forward with U.S. industry todevelop PV technology and expand world-wide markets.

Domestic UtilitiesUtilities play a major role in all large-scale powergeneration, and, as a result, their activities takeon strategic importance. In FY 1994, UPVG con-firmed that utility involvement in large-scale PVmarkets is strong.

UPVG was formed in 1992 to accelerate cost-effective PV applications. It consists of membersfrom 89 utilities, the American Public PowerAssociation, of Washington, D.C., the EdisonElectric Institute, of Washington, D.C., EPRI, andthe National Rural Electric Cooperative Associa-tion, of Washington, D.C. Collectively, these 89utilities produce almost one-half of the electricityconsumed in the United States. The UPVG issupported by a grant from DOE’s PV Program.

PV Potential

In FY 1994, UPVG carried out a study of potentialPV use by U.S. utilities. UPVG’s members identi-fied nearly 9000 MW of capacity that utilitieswould be willing to purchase at a price of $3,000per kW as cost-effective systems.

The size of the potential market, 9000 MW, wasmuch greater than expected. With a domestic util-ity market of this magnitude, UPVG outlined howthe PV industry can begin a self-sustaining processwhereby PV eventually becomes competitive with

Meg

awat

ts(M

W)

80

60

40

20

01988 1989 1990 1991 1992 1993 1994

Total

US

CD-SS24-A295510

World PV Module Shipments

Annual worldwide sales of PV have increased at a rate of 17% per year for the last 10 years. In 1994,U.S. industry’s share of this market increased to 36%, up from less than 30% in the late 1980s.

Expanding PV MarketsAs PV technology continues to advance in worldwide markets,the PV Program helps U.S. industry enhance its leadershipposition.

Sys

tem

Cos

t ($/

W)

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

CD-SS24-A295503

20

15

10

5

0

Carrisa 1-B

Austin PV300

PVUSA US-1

PVUSA "Kerman"

SMUD PV-2

SMUD (UPG)

SMUD (APS)

Price of Utility PV Systems

In FY 1994, the Utility PhotoVoltaic Group documented the steady reduction in price ofutility PV systems during the last decade.

Fiscal Year 1994 19

conventional power generation by consolidatingexisting markets and price reductions.

In FY 1994, 62 utilities expressed formal interestin participating in demonstration projects withUPVG. The initial proposals from these utilitieshad an aggregate capacity of more than 8 MW,with a combined utility investment of more than$38 million.

The Value of Grid Support

The PV Program helped lay the groundwork forthe UPVG study through its demonstration pro-jects organized under the Photovoltaics for UtilityScale Applications (PVUSA) project. For example,the PV Program helped sponsor the world’s firstT&D grid support system through PVUSA. The sys-tem, rated at 500 kW, was installed near Fresno,California, in FY 1993. In FY 1994, the PV Programprovided technical assistance to SMUD to install a

second grid-support system, rated at 200 kW, inSacramento, California. Also in FY 1994, PVUSAcompleted a study of the value of such PV systemsfor other utilities.

Sandia supplemented this work by sponsoring astudy with four other utilities in FY 1994 to deter-mine the value of PV for grid support. Each of theutilities participating in the study operates in a dif-ferent economic environment and has a differentsolar resource than the others.

In related work, Sandia engineers helped developa spreadsheet computer program in FY 1994 tohelp utility engineers quickly identify where theycould locate cost-effective PV systems in theirgrids. Developed by the NEOS Corporation, asmall business of Lakewood, Colorado, the spread-sheet allows utilities to perform quick cost-benefitstudies that prove a PV system is worth more exten-sive analysis.

The Potential Utility Market for PV

In its groundbreaking study, UPVG identified a domestic utility market of 9000 MW when PV systems reach $3,000 per kW. The followingare the five main components of that market.

T&D area grid support connects a PV system with a capacity of 200 kW–1000 kW to a T&D substation to supply added power to the areaserved by the substation. When the area experiences load growth, the PV system reduces excess loading on all the T&D lines and otherequipment in the area.

T&D radial line support connects a PV system with a capacity of 50 kW–200 kW directly to a transmission line experiencing load growth,and therefore alleviates problems such as voltage drop and poor power quality. All together, there are more than 200,000 distribution linesin the United States; as many as 5% could be connected to a PV system.

PV-friendly pricing is a rate structure used by several utilities that allows grid-connected customers to choose PV as their main power supply.The utility installs and maintains the system on the customer’s property, and charges a premium for the service. SMUD has the largestPV-friendly pricing program.

DSM for residential buildings is where PV systems located on a residence are controlled (electrically) by a utility for demand-sidemanagement (DSM). Thus, PV systems become part of the DSM industry, which currently has annual sales of more than $3 billion.

Remote residential buildings can use PV-generated electricity provided by a utility less expensively than connecting to a distribution line.

CD-SS24-A295504

Source: Utility Photovoltaic Group

PV Megawatts0

Indicates the EstimatedRange of PV Potential

200–240

215–260

1020–1130

3100–3430

3800–4200

3000 40003500 45002500200015001000500

T&D—Area Grid Support

T&D—Radial Line Support

"PV Friendly" Pricing

DSM for Residential Buildings

Remote Residential

20 Photovoltaic Energy Program Overview

Each chose a substation that was experiencingload growth and evaluated the economics ofinstalling a PV system at the substation. Eachdetermined the “break-even” cost—the price atwhich PV systems would be economical at thatsubstation.

The results showed that the cost-effectiveness ofPV for grid support is strongly dependent on loca-tion and utility economics. But for some utilities,PV is already cost effective for grid support in cer-tain locations. Based on the results of this study,these utilities now have tools to help evaluatewhen PV is cost effective for grid support.

PV for Demand-Side Management

The PV Program is also verifying the value ofusing PV for DSM applications. Such a PV systemcan be valuable to utilities to supply capacity dur-ing peak periods to the extent that its output corre-sponds with the times the utility experiences itspeak demand. The extent of this correspondencehelps utilities determine the value of the PV forDSM.

NREL is developing a data base that utilities canuse to evaluate the extent of the correspondencefor their systems. NREL is collecting data fromits own PV-DSM system on the SERF and fromprojects under way with the University ofDelaware’s Center for Energy and Environmental

Policy. Furthermore, the State University of NewYork at Albany is looking at utility load profile andsolar availability correlations.

The majority of the data base will be provided byprojects sponsored by the U.S. Environmental Pro-tection Agency (EPA). In FY 1994, EPA installedPV systems on the rooftops of buildings owned by20 utilities throughout the country. The systems,which vary in capacity from 4 kW–12 kW, operatein different types of climates and in different typesof utility load patterns. EPA is evaluating theextent to which PV can reduce emissions. NREL iscollecting data from the installations for 3 years toanalyze their usefulness for DSM.

NREL is also supporting development of analyticaltools to analyze the economic viability of PV forDSM. Working under subcontract with NREL, ana-lysts with the University of Delaware developed aspreadsheet in FY 1994 that both building ownersand utilities can use to evaluate the economicperformance of PV-DSM systems in buildings. InFY 1995, NREL will use the spreadsheet in casestudies of PV-DSM systems in buildings with fourparticipating utilities sponsored under thePV:BONUS project.

PV For ExportCurrently, 75% of this country’s PV production isfor export. In fact, the export market, mostly indeveloping countries, is the fastest-growing exist-ing PV market. Furthermore, its potential size ishuge. Greater than 40% of the world’s popula-tion—approximately 2 billion people, mostly inthe developing world—currently live withoutelectricity.

Most of this potential market is for people who donot currently have a reliable source of electricity.Often, these people require small amounts ofpower for such applications as indoor lighting,television, pumping water, and refrigeration ofvaccines for medical uses. PV is often the most cost-effective way to provide electric service.

In FY 1994, the PV Program identified the coun-tries most active in installing PV systems world-wide: India, Brazil, China, Russia, Mexico,Indonesia, and South Africa. In FY 1994, the PVProgram supported projects to stimulate marketsfor U.S.-made PV systems in each of thesecountries.

PG&E

1994 Study

1994 Actual Bids

APS Austin SRP Plains SMUD

Bre

ak-E

ven

PV

Cos

t ($/

kW)

CD-SS24-A295502

PG&EAPSAustinSRPPlainsSMUD

Pacific Gas & Electric Company's Kerman SubstationArizona Public Service CompanyCity of Austin ElectricSalt River ProjectPlains Electric G&T CooperativeSacramento Municipal Utility District

12,000

10,000

8,000

6,000

4,000

2,000

0

Break-Even Cost of PV Systems

When used for grid support, the cost-effectiveness of a PV power plantdepends strongly on its location and utility economics. Some utilities canobtain good value from PV installations for grid support at today’s prices.

Fiscal Year 1994 21

Delegation to India

DOE led a presidential mission to India in FY 1994in order to secure commercial deals for industryand major governmental agreements. Secretaryof Energy Hazel O’Leary headed the delegation,which included a number of representatives fromthe U.S. PV industry.

As a result of the presidential mission, Indian and,U.S. companies concluded commercial dealsworth more than $1 billion. India and the UnitedStates signed formal documents for bilateral con-sultations on energy and, specifically, on renew-able energy and PV. In FY 1995, the PV Programwill continue to support the U.S. PV industry’sincreasing involvement in the rapidly growingIndian market.

Rural Electrification

On the other side of the world, the PV Programand NREL expanded demonstration programsof rural electrification in Brazil and Mexico inFY 1994. The purpose of the PV Program’s involve-ment in rural electrification is to develop a modelof how developing countries such as Brazil andMexico can achieve their social and economicdevelopment goals while moderating the environ-mental impact that will inevitably accompany theireconomic expansion.

Global Potential for Off-Grid PV

Country Population(millions)

PopulationOff-Grid(millions)

PotentialPV Capacity(MW)

Africa (South of the Sahara) 310 280 14,000Africa (North of the Sahara) 135 56 3800Asia (Southeast) and Oceania 520 375 26,600Brazil 145 23 2300China 1070 400 28,000India 770 600 42,000Indonesia 175 80 9800Mexico 80 20 800Russia 280 5 550South and Central America (Spanish) 190 40 5250South and Central Europe 195 15 1180Source: Derek Lovejoy, “The Natural Resources Forum,” May 1992, p. 102

Department of Energy Secretary Hazel O’Leary led a presidential mission to India in FY 1994to secure bilateral agreements on energy. Through this mission, the U.S. PV industry was ableto conclude several important business agreements.

U.S

.Dep

artm

ento

fEne

rgy

22 Photovoltaic Energy Program Overview

Brazil

The rural electrification program in Brazil, whichbegan in FY 1993, is taking place in two phases.In Phase 1, Brazilian utilities in two states and thenational ministry of energy (called Centro dePesquisas de Energía Elétrica [CEPEL]) installedand will maintain PV lighting systems in750 homes and 14 schools in the outback ofnortheast Brazil. Installations were completed inFY 1994; CEPEL will begin evaluation of the effec-tiveness of the systems in FY 1995.

In FY 1994, DOE and CEPEL committed toexpand the project into six additional states inthe northeast and Amazon regions of Brazil. Theexpansion was in response to requests by the utili-ties in those states to become involved. Phase 2projects involve a greater variety of stand-alone,end-use applications, such as water pumping, tele-communications, and refrigeration. In addition,larger scale, hybrid village power systems will alsobe installed. Hybrid systems have at least twosources of power from, for example, PV, wind,diesel fuel, or some combination of the three,

and a large enough capacity to supply the minimalelectrical needs of a small village.

NREL began procurement of Phase 2 systems inFY 1994. NREL is supplying system integrationand design assistance, and specifying supply ofU.S.-manufactured components for use in theproject. Phase 2 installations will be completed inFY 1995.

The project is a model for other village electrifica-tion projects because of the involvement of thelocal and state utilities and national energy institu-tions such as CEPEL. Such projects establish insti-tutional and economic confidence in PV, and helpsolidify business relationships between U.S. andBrazilian firms.

In FY 1994, NREL received the Award of Merit forTechnology Transfer from the Federal LaboratoryConsortium for its work in the village electrifica-tion project in Brazil.

Mexico

The rural electrification program in Mexico,which began in FY 1991, is expanding in scope intwo states in northern Mexico. Each year, Mexicoinstalls thousands of PV lighting systems forremote households throughout the country undera program to promote economic developmentin rural areas. In FY 1994, Sandia helped twostates—Sonora and Chihuahua—install PV-powered water systems and train maintenancepersonnel.

Also in FY 1994, Sandia signed cooperativeagreements to develop PV systems with the stateUniversity of Sonora, the National Institute ofElectrical Research, and the Ministry of Energyand Mines.

The StatesThe PV Program also awarded cost-shared con-tracts to state agencies and working groups inFY 1994 to address relevant policy and technicalissues. Working through an initiative of 12 stateworking groups called Photovoltaics for Utilities(PV4U), PV Program analysts participate in a col-laborative discussion among stakeholders in thePV and utility communities to eliminate policyand technical barriers to the adoption of PV.

In FY 1994, PV4U established a bibliography ofimportant utility-related PV publications andlaunched a clearinghouse of PV documents. In

In FY 1994, 700 PV systems such as this one in Ceará, Brazil, wereinstalled by the Brazilian energy ministry with support from the NationalRenewable Energy Laboratory.

Rog

erTa

ylor

,Nat

iona

lRen

ewab

leEn

ergy

Labo

rato

ry

Fiscal Year 1994 23

addition, PV4U publishes a quarterly newsletterfor states active in PV development.

Federal AgenciesFederal agencies administer more than 30% of thetotal land area in the country. They have a largeneed for remote power that PV systems can fill eco-nomically and with minimal environmental impact.

Sandia continued progress on two projects inFY 1994 for increasing use of PV by federal agen-cies—PV for military bases and PV for the nationalparks.

PV for the Military

The military has used PV for many years for tacti-cal units in the field. PV generation provides apower supply for tactical and mobile combat unitsfor situations when generator noise cannot betolerated or when fuel supply lines cannot besecured.

Sandia engineers are assisting the military in devel-oping eight PV projects. The projects are fundedcooperatively by EPA and the Defense Departmentunder the Strategic Environmental Research andDevelopment Program (SERDP). ThroughSERDP, Sandia is assisting the U.S. Marine Corps,the U.S. Navy, and the U.S. Air Force to install PVsystems at military bases around the country. Byproviding assistance to other federal agencies in

their PV projects, DOE effectively leverages itsfunding.

The eight SERDP projects involve technicaladvances for PV systems used by the military. Thesystems—rated 75 kW–300 kW—are larger thanthose used by tactical units and are connected tothe utility grid. Seven are PV-diesel hybrid systemsdesigned to replace mobile diesel-powered gener-ators; the eighth is a grid-support system at the AirForce’s Yuma Proving Grounds in Arizona.

The Yuma project involves the first grid-supportsystem installed by the U.S. military. It is signifi-cant because military bases often produce theirown electricity through a “mini-utility.” The grid-support systems contain battery storage for addedreliability and reduced overall demand. Sandiaengineers will document the performance of thissystem for possible use on other U.S. militarybases around the world. In FY 1994, they com-pleted specification and began purchasing thehardware; installations are scheduled for FY 1995.

PV for the Parks

The National Park Service is installing PV atnational parks all over the country because PV sys-tems have a negligible impact on these sensitiveenvironmental areas and because the systems pro-vide the most economic power supply. In FY 1994,Sandia and park service engineers conducted astudy of the use of PV systems in the national

This 80-kW, PV-hybrid power system is located at Mount Home Air Force Base in Grasmere Point, Idaho. The system provides power for mobile radarequipment.

Mou

ntH

ome

AirF

orce

Base

24 Photovoltaic Energy Program Overview

parks. They identified 455 PV systems, 97% ofwhich were operating without any problems. Fur-thermore, they found the users were satisfied withthe systems.

With the help of the Design Assistance Center atSandia, park service engineers completed evalu-ation of 367 potential sites for PV systems innational parks in FY 1994, and identified cost-effective opportunities for PV in 120 of them. Theengineers also identified 659 separate projectswhere PV would be the best, most cost-effectivesource of power for the park service.

By FY 1994, Sandia and the National Park Servicecompleted work at the Cholla Campground in theTonto National Forest in Arizona. The camp-ground is 9.6 km (6 mi) from the nearest electricutility grid, and estimates for extending a powerline to the campground exceeded $750,000. Onthe other hand, PV systems cost $240,000 andnow provide power for a number of applications,including area lighting, water pumping and disin-fection, building ventilation, and hookups forrecreational vehicles.

The park service will continue PV installations innational parks in FY 1995.In conjunction with the Design Assistance Center at Sandia National Laboratories, the National

Park Service is evaluating hundreds of sites in national parks for installing PV. At the ChollaCampground in the Tonto National Forest in Arizona, the Park Service saved $500,000 byinstalling PV.

San

dia

Nat

iona

lLab

orat

orie

s

Fiscal Year 1994 25

Produced for theU.S. Department of Energy

1000 Independence Avenue, SWWashington, DC 20585

by the National Renewable Energy Laboratory,a DOE national laboratory.

DOE/GO-10095-082DE94011871March 1995

Printed with a renewable-source ink on paper containing atleast 50% wastepaper, including 20% postconsumer waste

Key Contacts

U.S. Department of EnergyJames E. Rannels, DirectorPhotovoltaics Technology Division, EE-1311000 Independence Avenue, SWWashington, DC 20585Phone: (202) 586-1720Fax: (202) 586-5127

National Renewable Energy LaboratoryThomas Surek, Program ManagerPhotovoltaics Division1617 Cole BoulevardGolden, CO 80401-3393Phone: (303) 384-6471Fax: (303) 384-6530

Sandia National LaboratoriesChris Cameron, ManagerPhotovoltaics Systems Applications Department, MS-0753Albuquerque, NM 87185Phone: (505) 844-8161

Marjorie Tatro, ManagerPhotovoltaics Systems Components Department, MS-0752Albuquerque, NM 87185Phone: (505) 844-3154Fax: (505) 844-6541

NOTICE: This report was prepared as an account of work sponsored by an agency of the UnitedStates government. Neither the United States government nor any agency thereof, nor any oftheir employees, makes any warranty, express or implied, or assumes any legal liability or respon-sibility for the accuracy, completeness, or usefulness of any information, apparatus, product, orprocess disclosed, or represents that its use would not infringe privately owned rights. Referenceherein to any specific commercial product, process, or service by trade name, trademark, manufac-turer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, orfavoring by the United States government or any agency thereof. The views and opinions ofauthors expressed herein do not necessarily state or reflect those of the United States govern-ment or any agency thereof.

Printed in the United States of America

Available to DOE and DOE contractors from:Office of Scientific and Technical Information (OSTI)P.O. Box 62Oak Ridge, TN 37831

Prices available by calling (615) 576-8401

Available from:National Technical Information ServiceU.S. Department of Commerce5285 Port Royal RoadSpringfield, VA 22161(703) 487-4650

Information pertaining to the pricing codes can be found in the current issue of the following publi-cations which are generally available in most libraries: Government Reports Announcements andIndex (GRA and I); Scientific and Technical Abstract Reports (STAR); and publication NTIS-PR-360available from NTIS at the above address.

DEPART

MENT OF ENERGY

UNITED

STATES OFAM

ERICA

1994 PV ProgramAccomplishments

Improving PV Systems p. 2

The U.S. Department of Energy’s (DOE) Photovoltaic (PV) Program helped participating manufacturers reduce theircost of producing PV modules by a per-company average of almost 50% through the PV Manufacturing Technology(PVMaT) project. The PV Program issued a new solicitation for research and development (R&D) projects emphasizingproduct-driven manufacturing.

Through PV:Building Opportunities for PV in the United States (PV:BONUS), teams of private firms developedprototypes of new products for incorporating PV into buildings.

The National Renewable Energy Laboratory (NREL), of Golden, Colorado, inaugurated the Solar Energy ResearchFacility (SERF), the nation’s premier R&D facility for PV.

Advancing PV Technology p. 12

The PV Program shared a prestigious R&D 100 Award in fiscal year (FY) 1994, given by R&D magazine for the 100most important scientific achievements of the year. The award was for a back-contact PV cell made from crystallinesilicon (c-Si) material. SunPower Corporation, a small business of Sunnyvale, California, produced a prototype c-Simodule using these cells, achieving a record efficiency of 21.6%.

NREL produced a small-area cell made from crystalline gallium indium phosphide/gallium arsenide with a recordefficiency of 30.2% at 180 suns concentration.

Sandia National Laboratories, of Albuquerque, New Mexico, produced a large-area multicrystalline-silicon modulewith a record efficiency of 15%.

The PV Program established the Thin-Films PV Partnership Program in FY 1994 to coordinate U.S. research inthin-film PV, and to develop a stream of pre-commercial prototype products with U.S. companies (mostly smallbusinesses), operating in this field.

NREL fabricated a small-area cell made from thin-film copper indium gallium diselenide (CIGS) material with arecord efficiency of 16.8%, the highest efficiency recorded to date for this material.

Siemens Solar Industries, of Camarillo, California, produced a CIGS power module with a record efficiency of 10.3%.

United Solar Systems Corporation, of Troy, Michigan, produced a prototype module made from amorphous silicon(a-Si) with a record stabilized efficiency of 10.2%.

Expanding PV Markets p. 19

The Utility PhotoVoltaic Group (UPVG), representing 89 U.S. utilities, completed an authoritative study of near-termpotential markets for PV by utilities.

The PV Program completed studies of the value of connecting PV power systems to utility transmission anddistribution (T&D) grids: PV power plants connected to T&D grids at some locations are cost effective at today’s prices.

NREL completed support of the first phase of a large-scale PV project demonstrating the feasibility of PV systems inremote villages in Brazil.

Sandia began procurement and technical evaluation of PV systems for the military.