State-of-the-art photovoltaic systems have Energy Pay Back Times as low as 1.7 years

Photovoltaic module production has shown a tremendous increase, with production of PV modules accelerating from ~80 MWp in 1995 to ~1700 MWp in 2005.

Introduction

Photovoltaic module production has shown a tremendous increase, with production of PV modules accelerating from ~80 MWp  in 1995 to ~1700 MWp in 2005. With such impressive growth figure, and in view of the large potential of photovoltaics as a renewable energy source, an analysis of present and future environmental performances of photovoltaic systems becomes increasingly important. This is typically carried out using the method of Life Cycle Assessment (LCA*).

LCA studies on PV have been published but these were mainly based on PV technologies from the late '80s. Since then, the PV industry has implemented many innovations in the field of solar cell and module technology. An update on the environmental profile of PV modules was therefore urgently needed to compare PV more realistically with the other energy options such as, e.g., biomass, wind, nuclear and coal. We need to know, for instance, the energy payback time and lifecycle greenhouse gas emissions related to energy technologies. These parameters are particularly important for PV since the energy demand for PV module production is the dominant factor for the environmental impact of PV.

The Copernicus Institute (University of Utrecht) and ECN Solar Energy (Petten) recently concluded a LCA study on crystalline silicon PV [1-3]. Three mainstream Si technologies were analyzed (see Table). Together, these PV technologies covered ~94% of the PV world market in 2004. The results of this study are thus representative for most of today's PV technology.

The study was carried out in the framework of the European Integrated Project CrystalClear, which aims to improve solar cells based on crystalline silicon. Highlights of this study were presented at the Materials Research Society Fall 2005 Meeting [1-2]. It attracted much enthusiasm since they were the result of a unique collaboration with 11 PV companies in Europe and USA. For the first time, up-to-date details about materials and processing steps were used over the whole production chain of PV systems (Fig. 1). The aggregate data set has been published on the ECN website for general use.

Figure 1: The different steps in the production process for three types of crystalline silicon solar cells and the “system boundary” of the new Life Cycle Assessment study.
Note that in the “ribbon” technology (process shown in middle) silicon wafers are pulled directly from molten silicon, without sawing., thus avoiding significant material losses.

Table 1: Assumptions for LCA study on crystalline silicon PV technologies¹  

Parameter	multi-crystalline Si	mono-crystalline Si	ribbon-silicon
Si wafer thickness (μm)	270-300	270-300	300-330
PV module life time (yr)	30	30	30
PV module efficiency (AM1.5, %)	13.2	14	11.5

¹A standard PV module type was used: 125 x 125 mm wafers, 72 cells/module, 1.25 m², module area, 3.6 mm glass with EVA/Tedlar encapsulation, modules, either unframed or framed with a 3.8 kg aluminum frame. The life time of the inverter was assumed 15 years. End-of-life disposal was not included in this LCA study.

Figure 2 shows the calculated energy payback time for grid-connected roof-top installed PV-systems. Three components make a PV system, (1) the laminated module, (2) the frame and (3) the inverter/cabling (Balance Of System, BOS). As can be seen in Figure 2, the EPBT is largely determined by energy requirements to make the PV module laminate, i.e, encapsulated wafers on glass. The EPBT's are between 1.7-4.6 years, the exact figure depending on the level of annual irradiation and type of silicon technology used. This is already much smaller than the lifetime of PV modules, which is normally 30 years or more.

In figure 3 the life cycle CO₂ emissions for various energy technologies are compared (in g-CO₂/kWh produced). As expected, PV performs very well in comparison with fossil fuel based technologies, but further improvements are still needed to compare PV with, e.g., wind energy. In figure 3 the life cycle CO₂ emissions for various energy technologies are compared (in g-CO₂/kWh produced). As expected, PV performs very well in comparison with fossil fuel based technologies, but further improvements are still needed to compare PV with, e.g., wind energy.

Figure 2: Energy Pay-Back Time (in yr) for a grid-connected PV-system under an irradiation of 1700 kWh/m²/yr (Southern-Europe) respectively 1000 kWh/m²/yr (Middle-Europe).

Figure 3: Greenhouse gas emissions of PV systems based on three silicon technologies, compared to a number of other energy technologies. N.B. The emission from a coal-fired power plant (1000 g/kWh) exceeds the Y-axis maximum! (Sources: Coal, CC gas, nuclear, biomass and wind data derived from the Ecoinvent database [4])

The figures for PV can be further improved by increasing module efficiency to 16%, reducing the wafer thickness down to 150 μm, and by switching to a new production process for silicon (based on Fluidized Bed Reactor technology). For this particular case, an EPBT of ~1.0 year and a life cycle CO₂ emission of 20 g/kWh were calculated for multicrystalline Si PV module when installed in S-Europe. In view of the present state-of-the-art on crystalline silicon PV, we can expect such competitive performance within the next 3 years.

Figure 4: Energy Pay-Back Time for future multi-Si and ribbon technology (two bars at the right), compared with today’s  status. The numbers below the X-axis give the respective module efficiencies. The systems are assumed installed in Southern Europe.

* In a Life Cycle Assessment study the environmental impacts of a product are evaluated by making an inventory of all energy and material inputs, as well as the emissions to the environment. Typically, a “cradle-to-grave” approach is used, i.e. the complete life cycle of a product from resource mining to waste treatment is taken into consideration. For PV systems the product use phase has negligible impacts and recycling processes only exist at pilot scale. Therefore these two stages have been omitted from the present study. (N.B.: A separate study on the pilot process for module recycling showed that implementation of this process would reduce the overall impacts [5]).

The Energy Pay-Back Time indicates the number of years the PV system has to generate electricity in order to compensate the energy invested during production of the system components. For wind turbines the energy pay-back time is in the order of 4 months, for (fossil) fuel plants the concept is less applicable because most energy input occurs during plant operation. The Life-Cycle Greenhouse Gas Emission of a PV system is calculated differently, namely by determining the total emission of greenhouse gases over the PV system's life cycle (i.e. mainly from component production) and dividing this by the total amount of electricity generated by the PV system over its life time. The life cycle greenhouse gas emission indicator can be compared between energy technologies to determine their potential contribution to greenhouse gas mitigation. The greenhouse gas emission of the present electricity supply system in the Netherlands is 570 g/kWh.

References

1. Alsema, E. and M.J. Wild-Scholten, Environmental Impacts of Crystalline Silicon Photovoltaic Module Production in Materials Research Society Fall 2005 Meeting. 2005. Warrendale, USA: Materials Research Society.

2. Wild-Scholten, M.J.de and E.A. Alsema. Environmental Life Cycle Inventory of Crystalline Silicon Photovoltaic Module Production. in Materials Research Society Fall 2005 Meeting. 2005. Boston, USA: Materials Research Society, see also http://www.ecn.nl/library/reports/2006/c06002.html

3. Fthenakis, V. and E. Alsema, Photovoltaics Energy Payback Times, Greenhouse Gas Emissions and External Costs: 2004–early 2005 Status. Progress In Photovoltaics: Research and Applications, 2006. 14(3): p. 275-280

4. Frischknecht, R., et al., Ecoinvent 2000 - The Swiss National Life Cycle Inventory Database. 2004, Swiss Centre for Life Cycle Inventories: Dübendorf, CH, www.ecoinvent.ch

5. Müller, A., K. Wambach, and E. Alsema. Life cycle analysis of a solar module recycling process. in Materials Research Society Fall 2005 Meeting. 2005. Boston, USA: Materials Research Society, Warrendale, USA.

---