DEPARTMENT OF MECHANICAL ENGINEERING

COLORADO STATE UNIVERSITY

MATERIALS ENGINEERING LABORATORY

Mass Production of Photovoltaic (PV) Modules

    MARKET NEED | THE TEAM | PV MANUFACTRUING TECHNOLOGY |
PERFORMANCE THE PROTOTYPE MANUFACTURING TECHNOLOGY | FUTURE  EFFORT | MARKET SUMMARY
CONTACTING US | REFERENCES | SELECTED PUBLICATIONS 
ANNUAL    RESEARCH REPORT FOR RESEARCH SPONSORED BY THE NATIONAL RENEWABLE ENERGY LABORATORY

Shipments of photovoltaics (PV or solar cells) continue to increase at 35% a year. Large-scale implementation of PV can lead to significant economic and global environmental benefits. For PV to contribute significantly to the generation of electricity, reliable modules with high efficiency must be manufactured at large throughputs with low costs. Several major corporations have expressed the strong need for a technology to enable the mass production of thin film PV modules. The development of technologies for the mass production of PV modules has been the main focus of our efforts since 1991.

2. THE TEAM
Our team has spent a combined 25 years developing many innovations to improve the manufacturing efficiency and enable the mass production of PV modules. These innovations encompass production processes, production hardware and PV modules
design. These PV manufacturing technologies are demonstrated on an automated, working prototype.

Our team consists of Robert "Al" Enzenroth, Dr. W. S. Sampath, and Kurt L. Barth. We are mechanical engineers with background in R and D for industrial manufacturing, and materials science. We are general faculty at the Materials Engineering Laboratory of Colorado State University (CSU). We have considerable experience in developing manufacturing technology innovations for Fortune 200 companies. Our team goals are to achieve:

• A significant increase of worldwide PV manufacturing volume by 2010
• The reduction of PV module prices to enable widespread application of PV power
• A good return on investment for all the stakeholders
  

The PV manufacturing technology has been developed at CSU with grant funding from several federal sources. Grant obligations to these agencies have been fulfilled. The associated intellectual property has been officially released to our team in accordance with federal agency and CSU policies. Detailed domestic and foreign patents describing our process and hardware has been filed with our names as both inventors and assignees. The first domestic patent has been approved and will issue in summer 2002.

3. PV MANUFACTRUING TECHNOLOGY
We have demonstrated a working prototype of a manufacturing technology for fabricating polycrystalline thin film CdS/CdTe PV modules. This technology offers a significant improvement in capital productivity, labor productivity and overall manufacturing efficiency.

A. Automated, Low Cost Manufacturing Equipment:

• All semiconductor processing steps are fully automated and performed in one chamber operating at modest vacuum. An automated conveyor belt extends through all the processing stations in vacuum.
• The hardware for these steps is self-similar, readily scalable and is fabricated using common industrial materials and techniques. The vacuum processing steps include: glass heating, all semiconductor depositions, heat treatments and ohmic contact formation.
• Metallization to form the back electrode and module interconnection is performed using a robust, low cost, high throughput industrial spray processing outside of vacuum.
• Capital costs are low because (1) only modest vacuum is required for processing and (2) the simplicity of the hardware.

B. High Processing Speed:

• The manufacturing process is synchronous with a cycle time of 2 minutes for all vacuum processing steps (including the deposition of the semiconductors) on a fully automated conveyer system.
• One completed device emerges from the system every 2 minutes
• Novel interconnection technology is based on industrial processes and is low cost and high throughput. Monolithically interconnected mini-modules have been fabricated.

C. Other Advantages:

• Materials usage is excellent (near 100%).
• Raw materials are readily available for very large volume production (hundreds of megawatts) [ASARCO]. The raw material cost is estimated to be less than $0.40/Wp.
• No liquid waste and virtually no solid waste are generated.
• The systems are designed to be operated by personnel with secondary education.
  

4. PERFORMANCE THE PROTOTYPE MANUFACTURING TECHNOLOGY

A. Device Performance and Yield:
Over 3000 films and devices have been fabricated using the working prototype system. The 3.6 x 3.1 inch substrates used are low cost, soda lime glass, commercially available from Pilkington (TEC 15). Devices show excellent photovoltaic properties and have been consistently fabricated with 10.5 to 13% conversion efficiency [Barth]. A mean cell efficiency of 14 devices within one substrate of 12.6% with a standard deviation of 0.26% has been demonstrated. NREL verified efficiencies as high as 12.44% have been demonstrated with this system.

B. Device Reliability:
Nearly 500 devices have been studied in accelerated indoor stress testing to evaluate the reliability. Currently a randomly selected set of 17 devices, processed at the optimum condition, has demonstrated an average efficiency of 11.15% after 5100 hours of stress. One of these devices had an initial efficiency of 12.15% and has maintained an efficiency of 11.97% after 5167 hours of stress. The reliability of the devices is very promising. The stress conditions consist of 1 sun illumination, 65º C, OC with 5 hours of illumination out of an 8 hour cycle. Devices are also stressed at 77º C and 100º C. Testing is ongoing. Processing conditions have a great influence on device stability. As data is collected, empirical acceleration factors are being refined. However, these indoor tests are estimated to represent decades of field conditions.

Devices are also subjected to outdoor field conditions to evaluate reliability. After nearly a year of outdoor exposure the devices have little or no change in performance on average within the error of measurement.

C. Current Efforts:
Our group is currently working on the demonstrating:
(i) Multiple 8-hour runs with consistent device performance.
(ii) Developing larger process modules to fabricate 10 watt, monolithically integrated panels.

The process modules in our pilot system are being upgraded for operation without service, maintenance or reloading for at minimum 1 shift (8 hour) operation. Upon completion, multiple 8-hour long runs will be conducted and process capability (initial efficiency distribution, stability) will be quantified. This will demonstrate the manufacturing capability of our technology. These modifications are near completion. The film thickness produced from one of the process modules was evaluated over 10 hours of operation. The film thickness uniformity over time is shown in Figure 7.

Following the demonstration of a robust, manufacturing worthy process capable of 8-hour operation, we will upgrade the system to perform these processes on full 16x16 inch substrates. Recently, 14x14 inch deposition of device quality CdTe was deposited on 16x16 inch Pilkington glass substrates. Though testing and optimization continues, initial film quality appears excellent. Film uniformity has been characterized by optical transmission and is estimated at +/- 5% with a minor decrease in thickness in the film corners.

5. FUTURE EFFORT:
Our technology has been demonstrated on a working prototype, the next step is to demonstrate a pilot production manufacturing line. This line will incorporate the advances of the current working prototype and will process 10 watt modules. The production prototype for producing 10 watt modules will be built at our facility to operate at full production speeds of one module every two minutes. Only a small portion of the hardware subsystems needed for the production prototype are not available off the shelf. These subsystem components are robust (similar designs are used in many industries) and will be easily fabricated. The current working prototype system is being used to determine the designs of these subsystems and to test the hardware. The current working prototype is also being used to determine optimum process conditions (substrate temperature, vapor flux and residual gas composition) at the film growth interface. These same conditions will be maintained precisely in systems processing larger substrates.
Based on this production prototype, large scale manufacturing plants for PV modules can be designed and built. These plants will have capacities greater than 10 MWp/year. The hardware for these plants will be highly standardized and optimized for high manufacturing efficiency. Thus our technology is suitable for adding significant capacity rapidly. These standardized plants can be replicated to meet the needs of the market. It is expected that these plants will have a many fold improvement in capital, labor and time for installation, compared to current technologies.

6. MARKET SUMMARY:
Market Segments: Current yearly world sales of PV is approximately 380 megawatts (MW) [PV News] up from 40 MW in 1990. This is a 36% increase over 2000. The current market for PV systems (the panel plus related power conditioning hardware) is approximately $2 billion. PV shipments have grown at an annual rate of 38% in 1997, 18% in 1998 and approaching 20% in 1999 [SU, Beadle]. The terrestrial market for PV can be divided into four main market segments.
1. Grid Connected: These systems provide power during daylight for use on site or to supply the electrical utility grid. Larger 100 to 500 kilowatt systems are installed along utility lines operating near full capacity preventing the need to enlarge power lines or transformers.
2. Industrial: Industrial PV systems supply power for back-up or off grid telecommunications, telemetry and cathodic protection equipment.
3. Rural Habitation: Rural habitation PV systems provide power for off grid locations such as villages in developing nations or remote homes and cabins.
4. Consumer Goods/Indoor: This segment includes off grid PV for portable radios, computers and garden lights etc. as well as small electronic applications such as watches and calculators.
The grid connected, industrial, and rural habitation market segments account for 89% of the PV shipments; the market analysis for these three segments will be detailed below.

Market Forecast: Market forecasts vary but all call for near term markets to continue to increase rapidly. The PV industry road map, developed by members of the PV and utility industries, targets a 25% annual growth in PV shipments [Roadmap]. For the last three years, the annual growth in PV shipments has averaged over 35%.

Market Conditions: The current world PV production is currently worth over $2 billion annually. Additional markets are developing which require significant growth.

• Japanese Markets: Government policy in Japan calls for the installation of about 4,600 MW in the year 2010 [MOFA].
• German Markets: Germany has a continuing goal of installing 100,000 systems in 3-5 years totaling 400-500 MW [PVIR]. This represents the 2001 world production for Germany alone.
• Hawaiian Market: PV is cost competitive in Hawaii at current prices [DOE]. It will take over 100MW/yr. to supply the expected annual increase in Hawaii's electrical demand.
• Rural Markets: Nearly 2.5 billion people in developing countries have little or no access to electricity [UN]. Studies have identified a developing nation PV market of 400 million potential consumers and a total of 200,000 MW of capacity [Bonnet].
• Industrial Markets: PV will continue a steady 15-20% growth rate serving the expanding market segment.
  Different estimates for growth require up to 300 times current world production in the next ten years. When PV generated electricity is cost competitive with power from traditional sources the market is expected to grow "explosively".
• 9,000 MW capacity in only 10 domestic markets [Berger].
• Long-term forecasts are for a 5 million MW market by 2050 [Beadle].

7. CONTACTING US:
For more details including a list of our publications, please see the website www.engr.colostate.edu/me, click "Laboratories" then "Materials Engineering Lab". Members of our team can be contacted at the laboratory at (970)-491-8411. We can be reached by email at barth@engr.colostate.edu, or sampath@engr.colostate.edu. Demonstrations can be arranged by contacting us.

8. REFERENCES
ASARCO, Personal conversation with ASARCO Co. a large mining company and manufacturer of CdTe raw material.
Barth, K. L., R. A. Enzenroth and W. S. Sampath, "Continuous Belt Air to Vacuum to Air (AVA) Processing of Thin Films and Device", NSF Grantees Conference, January 2000. Published by the Society of Manufacturing Engineers
Barth2, Kurt L. Barth, Robert A. Enzenroth and W. S. Sampath, "Advances in Continuous, In-Line Processing of Stable CdS/CdTe Devices", 29th IEEE PVSC (refereed conference), May 2002.
Beadle: Beadle, Peter C., "Electricity From the Sun, Myth or Reality?", World Energy Vol.2, No.1, 1999. P. Beadle was president of BP Solar, Inc.
Berger, John, article based on book Charging Ahead, the Business of Renewable Energy and What it Means for America, May 1997 by Henry Holt and Company.
Bonnet, D., P. Meyers, "Cadmium-telluride-Material for thin film solar cells", J. Materials Res. Vol. 13, No 10, Oct. 1998.
DOE: "Customer-Sited Photovoltaics: Focusing on Markets that Really Shine", DOE/Go-10097-371, 1997.
MOFA: "Japanese Approaches to the Suppression of Greenhouse Gas Generation", Ministry of Foreign Affairs of Japan. Article at www.mofa.go.jp/policy/global/environment/warm/japan, chapter 1.
PVIR: Photovoltaic Insider' Report, "500 MW Market Created: New German Coalition Government Launches 100,000 Solar Rooftop Program", R. Curry editor, VOL, XVII, No. 12, DEC. 1998. 1011 W. Colorado Blvd. Dallas Texas 75208,
PV News, Vol. 21 No. 3, April 2002, Published by PV Energy Systems Warrenton, VA 20187, Paul Maycock editor.
Roadmap: "The Industry-Developed PV Roadmap, A Framework for U.S. industry and Technology Leadership", developed by current and past CEO's and Presidents of leading PV companies, leading academics and the president of Idaho Power, an electrical utility, Published on the Web at www.nrel.gov/ncpv/pdfs/27061.pdf.
SU: Strategies Unlimited pamphlet on report PM-47, 1999. 201 San Antonio Circle, Suite 205, Mountain View, CA 94040. Strategies Unlimited is a market research and strategic planning service provider.
UN: UN Chronicle, June 1995, Vol. 32, No. 2, page 69.

9. SELECTED PUBLICATIONS

10. ANNUAL RESEARCH REPORT FOR RESEARCH SPONSORED BY THE NATIONAL RENEWABLE ENERGY LABORATORY