By Klaus Eberhardt

The growing demand for PV modules requires future large-scale and highly-integrated manufacturing sites with annual capacities well above 500 MW. The corresponding challenges for an EPC company (Engineering, Procurement and Construction) in order to design and build such facilities is described and discussed below.

4 key areas enabling considerable potential for reducing CAPEX (Capital Expenditure) and OPEX (Operating Expenditure) have been defined:

-Process equipment

-Scaling

-Benchmarking, value engineering & industrial engineering

-Integrated PV fab

The respective key elements for each area are shown in Figure 1.

Technology improvements in order to increase the conversion efficiency of PV modules as well as high throughput process equipments with high uptime and increased yield are main factors for future cost reduction. This is driven by the process equipment industry and will not be investigated in more detail within this article. The other elements are well within the scope of services from an EPC company and they will be presented and discussed within this paper.

Scaling

Back in the year 2000, a typical annual production capacity of a single wafer-based cell or module line was in the range of 25 MW. Most of the facilities have been designed for a total capacity between 25 MW and 50 MW. As of today the annual capacity of a standard line is in the range of 50-60 MW mainly driven by an increase in cell efficiency as well as throughput and yield enhancements of the process equipments. Currently several of these lines are installed next to each other to get the desired manufacturing capacity. For a case study a 100 MW solar cell line is being compared with a 600 MW line. KPIs (Key Performance Indicators) are the manufacturing area (MA) and the total gross building area (TA) with respect to the manufacturing capacity (CA). By de-bottlenecking, grouping of the process equipments to so-called ‘farm’ layouts and by reducing the overall numbers of process equipments due to higher utilization factors, the manufacturing area productivity (MA/CA) can be improved by 27%. Due to smaller footprint of the Facility systems by capacity, the total gross building area (TA/CA) can be reduced by almost 40%. The improvements are depicted in Figure 2.

The total investment costs for the building and facility by capacity could be reduced by more than 20%1) for the given example. Beyond improved utilization of building areas and facility systems, the economies of scale result in further cost reduction in the areas of process equipment invest and reduced material costs through volume purchase agreements and through the dilution of fixed costs.

The optimal annual production capacity for a single story building varies between 200 MW and 400 MW. The average manufacturing area by capacity for a thin-film fab is higher compared to Wafer-based technologies: The main differences for thin-film versus wafer-based technologies are:

-Large glass substrates instead of 6-inch wafer

-Integrated manufacturing (front end and back end cannot be separated.)

-Lower efficiencies.

By using multi-level buildings, the capacities within one building can be increased accordingly. But construction time and construction cost by area increases for multi-level buildings. Land availability is a main criterion for the decision whether to construct a one-level or multi-level building.

However there is a natural limitation for further scaling up:

-Manufacturing areas get too large and it is difficult to meet safety standards like distance to emergency exits.

-Costs for inter building logistic increases.

-Costs for shipping the manufactured products to the consumer increases in case the local market cannot absorb the entire production.

Beyond that a further increase of manufacturing area will not result in major cost savings by area. Innovative approaches are required for further cost reductions, as described below.

Smart Engineering: Value Engineering, Benchmarking and Industrial Engineering

The technology drives the overall design. The building and the facility systems have to be designed according to the requirements of the process equipments. Figure 3 shows the interdependences between the technology (equipment & automation, production environment) and the required utilities. The facility systems have to supply and dispose the utilities and the building is the overall ‘umbrella’  Industrial engineering work has to be applied for optimized process equipment layout, material and process flow as well as for improved logistic concepts.

It is obvious that the EPC company has to be aware of the requirements from these equipments in order to design the facilities in an optimal way. This is the prerequisite for a straightforward concept design for the building and for an accurate definition and sizing of the facility systems. During the concept design, value engineering steps for subsequent optimization of the concept have to be performed in order to define cost reduction potentials. A major KPI is the electrical power consumption of the process equipments. The customer or equipment supplier provides a so-called Tool Utility Matrix (TUM) with all the utility consumptions to the engineering company designing the facility. This matrix contains the average and maximum consumptions for utilities like electrical power, water, gases and chemicals as well as the requirements for cooling water and the specification for the exhaust streams and waste water. These values are the basis for the design, both for facility systems that supply the process equipments with the necessary utilities as well as systems for the heat removal, disposal and treatment of gaseous and liquid waste streams. The electrical power consumption of the process equipments significantly influences the design of various facility systems like transformers, cooling towers and chillers. In a first approximation, it is assumed that the electrical power consumed by the process equipments is being completely converted into heat. This heat must be removed from the production environment. The majority of the heat is conducted to the process cooling water, but part of the heat is dissipated into the environment of the building through thermal dissipation. The closed cooling water loops are recooled with chilled water. The air is recooled as well to keep the predefined temperature level in the manufacturing area. During the design phase, the cooling water loop and chillers are sized according to the electrical power consumption of the process equipments as defined in the TUM. Experience from designing high-tech fabs for other industries like Flat Panel or Semiconductor showed, that the utility consumption data as provided from the equipment supplier is sometimes higher compared to the actual consumptions. Reasons for that are:

-Safety factors

-Duty factors

-Tool utilization is overestimated.

If the facility systems are designed according to consumptions which are considerably higher than the real demand, most of the facility systems will be overdesigned. As a result, investment as well as running costs will be higher than necessary. Actual utility consumptions for various PV technologies have been measured after the production ramp-up. The outcome is a comprehensive database with as measured utility consumptions. This database is invaluable during the initial concept phase for new projects. Tool consumption data as received from equipment suppliers are benchmarked with the internal database. The comparison is presented to and discussed with the customer, sanity checks are performed and the final utility consumptions are jointly updated in case the initial consumption numbers can be reduced. The benefits are reduced investment costs for the facilities and building due to reducing size and number of electrical transformers, cooling towers, chillers as well as duct diameters. In addition, the running costs can be reduced, since the facility systems are operated at the optimum working point and less maintenance and service is required.

Another area for future cost reduction which is being explored currently is the recycling or reuse of chemicals and rinse water used in wet benches for wafer cleaning, saw damage etch or oxide etch.

Integrated PV Fabs

Wafer-based manufacturing can be separated in various parts of the added value chain. Therefore, most manufacturers started either with Cell or Module processing, to reduce the initial investment costs and related risks. Due to the market pressure to reduce costs, most companies are subsequently integrating upstream to wafer, ingot or even polysilicon manufacturing. Thin-film fabs are already highly integrated, but the manufacturing capacity of these fabs is steadily increasing and there are different requirements for exhaust and waste water treatment for the different thin-film technologies. As a consequence, smart site master planning as well as new building and facility concepts needs to be developed in order to respond to these differentiated market requirements. The overall trend is towards integration. The cost reduction potential for integrated PV fabs goes far beyond pure scaling: It offers the opportunity to centralize functions like gas farm, energy supply, water conditioning, waste treatment, chemical recycling, site logistics and others. Even functions which have been located historically off-site like glass manufacturing or power generation can be integrated in future wholly integrated, very large-scale PV fabs. Different scenarios have been investigated and evaluated. Figure 4 shows a vision for an integrated future PV fab complex.

The scenario assumes a fully integrated facility with 1 GW+ capacity within one site, even a coexistence of wafer-based and thin-film technologies is conceivable. A cogeneration plant provides electrical power, hot steam and chilled water. A float glass line with optional adjacent TCO (Transparent Conductive Oxide) coating is located at the same site. The capacity of the float glass line matches the demand of the module manufacturing. The glass properties (thickness, iron content, size) can be customized for the requirements of the front glass for wafer-based module manufacturing as well as for the glass-glass modules from the thin-film manufacturing. This will result in higher power for the modules due to decreased absorption in the front glass. To drive down costs for wafer-based manufacturing, possibilities to use frameless glass-glass modules are under investigation. In this case, the demand for glass may increase further.

Other synergies for integrated fabs like less transport costs between the different added value steps, optimized logistics, just in time delivery and reduced breakage will support to drive down the costs further.

Fast Time to Market

The fast return on investment is a key criterion for any new investment. If the financing of a new project is secured and the new project starts, the manufacturer has to make sure that the time to the production start is minimized. The process equipment suppliers standardize and modularize their equipment to reduce lead time and cost. The engineering company has to keep pace with this development and to make sure that the time for the design and permitting as well as for the construction of the building and for commissioning of the facility systems can be reduced as well. In order to accelerate the time schedule, it is advantageous to apply an integrated project approach2) in order to fulfill the requirements for cost-efficient design/build projects and to reduce the time to the production start. Key criteria are:

-Project Definition

    - Technology, initial capacity and expected expansions are well defined.

    - Utility requirements from the process equipments are well understood.

    - Site selection and site master plan is defined.

-Concept

    - Concept design including Industrial Engineering, Value Engineering and Benchmarking is being performed.

-Design

    - Final & permit design is executed.

-Procurement

    - Long lead items are being ordered.

-Construction

    - Execution design is performed while construction starts.

    - Facility systems are installed after the building is water tight, design for equipment hook-up starts.

-Operation

    - Equipment is installed and hooked up, all systems are commissioned.

    - Production start

This straightforward approach guarantees that there are no overlaps or missing links between the different steps and that the overall time is reduced. Figure 5 illustrates how this approach reduces the time to market significantly while benefiting from all cost reduction potentials as described above.

Overall project duration depends on size and complexity of the project. Experience shows that the overall time schedule can be reduced by 20% when the integrated approach is applied.

It has been shown that the innovative and experienced engineering (EPC) company can contribute in various areas to decrease the costs/Watt for PV manufacturers within all steps of the PV added value chain. The leverage from scaling is limited since manufacturing capacities in the range between 200 MW and 400 MW are an optimal capacity for a single story building. The technology and the expansion strategies are driving the design for the initial site master plan, for the building concepts as well as for the facility systems. All facility systems have to be sized according to the actual consumption of the process equipments. The experienced EPC company uses the lessons learned from previous projects to benchmark customer supplied tool consumption data. Technologies for either recycling or reuse of chemicals and water will become standard in future since it is costs effective above certain consumption volumes. Another area for cost reduction is the integration of various added value steps as well as of support functions within one site. Last but not least, the overall time schedule needs to be shortened to reduce the time to market. This can be achieved with an integrated project approach with certain overlaps between the different phases of the project. All these factors strongly support the PV industry on the way to sustainable growth and to reach grid parity. 

Klaus Eberhardt, Ph.D., is a veteran in the PV industry. He started the PV business at M+W in the year 2000. From 2006 to 2009, he supported Applied Materials Solar Division entering the PV Industry as an equipment supplier. Beginning 2010, he rejoined M+W group as Global Technology Manager for PV.