The drastic increase of humanity’s energy consumption results in an exponential rise in CO 2 emissions related to current predominant types of energy generation technology. The radiation exchange balance between the Earth’s surface and space has been altered by a significant increase in CO2 contents within the Earth’s atmosphere as observed over the last decades; now the balance occurs at higher surface temperatures. This is caused by the reduction of optical transmittance of the Earth’s atmosphere in the infrared range, at which thermal radiation from the Earth occurs, while the Sun’s solar spectrum reaching the ground remains relatively unchanged. Effects of that temperature increase, such as an increase of floods and hurricanes, cause additional damages in the vicinity of 50 billion US$ each year and are reflected in the statistics of Munich Re-insurance already today. On this back ground the need to examine the present energy supply systems concerning their carbondioxide intensity during their life cycle becomes obvious.
This work examines energy – and CO2 balances of photovoltaic power plants during their live cycles, including production, operation, dismantling and re-use.
Parameters influencing this balance such as raw materials, production and operation conditions, yield, and recycling-quotes are considered:
• Taken into account are the expenditures on the production of PV power plants, considering their specific material composition and conditions of production processes, even that of the raw materials.
• During the operation phase, all parameters having an influence on the electrical yield for the PV power plant such as cell reaching irradiance and operation temperature are modeled and examined in detail. Several suggestions for improvements are given and tested.
• The use of recycled material has a crucial impact on the energy expenditures of manufacturing: The exclusive use of secondary materials results in an energy saving of 92% for aluminum, for copper 73% and for glass 67%. This potential is not just a theoretical one; currently industrial recycling-quotas (in Germany) are already in the vicinity of 31% to 35% (up to 87% in the electrical engineering sector) for aluminum, 48% to 55% for copper and 45% for glass.
If the photovoltaic power plant is getting recycled after dism antling at the end of its usable lifetime, the material related energy-expenditures reduce themselves once more. Initial examinations revealed that the energetic expenditure for solar cells can be lowered by one magnitude if recycled material is used. Due to insufficient data at the manufacturing plants a detailed analysis of this effect could not be carried out yet. The modeling of the CO 2 balance is more complex. For the production of a PV power plant, its components and its raw materials, the carbon dioxide intensity of the country of production has to be taken into account. Already within Europe the corresponding specific CO2 emissions can differ by a factor of
27.6 (Netherlands at 442 g/kWh vs. Norway at 16 g/kWh). Increased global trade, where locations of production and application of most materials, com ponents and products are rarely in proximity, in conjunction with the tendency to change suppliers frequently, is making a solid, lasting statement on CO2 balance enigmatic.
Under present basic conditions, an exemplary comparison of Germany with Brazil shows that the highest CO2-reduction (26,805 kg per kWp of installed PV power plant) is observed for the production of PV power plants in Brazil and the local substitution of off-grid diesel-generators by PV-power plants based on singlecrystalline silicon technology. For production and installation in Germany, power plants based on multi-crystalline solar cells have some better CO 2-reduction abilities: 8,677 kg/kWp vs. 7,792 kg/kWp with single-crystalline technology.
Such calculations are significant for carbon trading: under unfavorable side conditions, for example production in Germany (with its relatively high CO 2 intensity of 0.56 kg/kWh) and installation with a grid-connection in Brazil (with its very low CO2 intensity), application of PV may even cause a negative CO 2 reduction: -1,009 kg/kWp (worst case).
All data are computed with conservative assumptions for the recycling quote (25%) as well as for the manufacture of raw materials (only local production). The balance can be influenced positively by importing raw materials from countries with a favorable energy-mix (e. g., Norway, Iceland).
To further improve CO 2 reduction by photovoltaic power plants, further measures have been proposed, simulated and tested: Via an accurate yield prediction, considering all relevant optical, thermal and electrical parameters, the planning – expenses can be reduced considerably by minimization of yield uncertainties. Beside the modeling, practical measures, primarily of optical and thermal nature, have been proposed to increase the electrical yield of a PV power plant.
Theoretical model-assumptions could be confirmed by practical measurements in Australia, Brazil, Germany and Zimbabwe. Examples: Irradiance reflection losses could be reduced by better matching the refractive indices, and the application of optical structures will allow to generate about 4% more electricity. Diminution of the operation temperature of the PV-Generators through large heat-capacities (water tank) results in an electrical yield gain of 6% to 12%. Additionally, the substitution of the conventional concrete-foundation by the water tank leads to a reduction of material and labor expenses. For the same yield a smaller generator becomes necessary, which further reduces material and energy requirements for production, and therefore increases the CO2 reduction potential during the life cycle of PV power plants.
 an increase of 10 K in temperature results in a doubling of the speed of chemical – biological reactions (RGT-rule), until an upper limiting temperature (ca. 60°C for enzymes), see Linder 1948/1977.
 i. e. beech wood: irradiance 3.7 PJ/(km2 a), storage in dry mass above ground: 570,000 kg (240,000 kg below the ground as roots and humus).
 from avg. irradiance and energy density of forest growth for moderate climates.
 maximum of solar yield is 5.4% for sugar beets (farmland in general: 0.3%).
 photosynthesis related to the global average.
 To measure the dark characteristic, an external variable power supply is necessary.
 Voltage loss of 0.3 V instead of 0.7 V at silicon diodes.
 Resistance for hail up to a diameter of 25 mm, torosion stability of the PV module for windspeeds up to 200 km/h.
 STC: cell temperature 25°C, irradiance 1000 W/m2 (perpendicular), sun spectrum equivalent to Air Mass 1.5 (see also IEC 60904-1, IEC 62145 and IEC 61215).
 SOC: as STC, but is using an actual measured cell temperature, occurring at an irradiance of 800 W/m2, an ambient temperature of 20°C and a wind velocity of 1 m/s. Common values are between ca. 42°C to 57°C. SOC are achieving more
realistic values for operation than STC (see also IEC 60891, IEC 61853 (draft) and IEC 61721).
 In remote areas electrical grid connection is expensive, conventionally powered generators such as diesel generators require additional transportation costs for fuel, oil and spare parts. In these remote areas renewable energy generating sets often are the most cost-effective, even in todays economical conditions (Kayne 1992, Vallve 1994).
 According to different examinations the public grid in many countries (as Germany) is capable of allowing PV grid injection in a vicinity of at least 15% of its nominal grid power.
The use of HVDC offers an additional advantage for this case: Aside from the transport, the frequency of 50 Hz by nine of the eightteen 715 MW generating units is transformed to 60 Hz. For political reasons – the power plant is located on the borderline between Brazil (public grid of 60 Hz) and Paraguay (public grid of 50 Hz) – half of the generators had to be built in 50 Hz, nevertheless Paraguay is just using 2% of the generated energy, while Brazil is using all the rest.
 IKARUS: Instruments for the development of strategies to reduce greenhouse gas emissions caused by energy use. Project by the German Ministry of Science and Technology, terminated in March 1995.
 bricks with vertical cavities holes and thermal insulation layer of styrofoam® (expanded polystyrene)
 bricks with vertical long cavities and font wall bricks with a thermal insulation layer of styrofoam® (expanded polystyrene)
CEE: Cumulated Energy Expenses
 Industrial multi-crystalline silicon cells of 100 cm2 area have reached 15.8% conversion efficiency (Sharp). The laboratory record is 17.8% for a 4 cm2 cell (UNSW). Commercial multi-crystalline cells are about 15% efficient (e. g., Q – cells).
To reduce energy losses by changing the temperature of the heating plate, two manufactures of laminators since 1996 (NPC and S. E. Project: 26 kW for four 83.6 Wp modules) keep the temperature of the hearing plates constant at 155 °C at initiate the cooling process by lifting the laminate off the hearing plate. According to the manufacturers energy consumption is only 20% of conventional laminators. More reasonable sees a saving of 30% in accordance to the relation of the heat capacity of the laminate to the heat capacity of the hearing plate.
 Spreng 1995 uses this method for energy balances, but the method can also be applied for CO2 balances.
 Due to the increase of electricity consumption, Brazil will require further power plants: While there are not any adequate locations for locations for large hydro power plants left, alternative scenarios will be developed, changing the energy mix and the CO2 intensity of Brazil.
24,408 kg/kWp (sc-Si, 25 a) 26,805 kg/kWp (sc-Si, 25 a) 25,372 kg/kWp (mc-Si, 25 a) 26,570 kg/kWp (mc-Si, 25 a) 19,860 kg/kWp (mc-Si, 20 a) 21,058 kg/kWp (mc-Si, 20 a)
 The use of stainless steeel would have been more adequate in terms of service life of the system, but for the limited duration of the tests (max. 3 months) galvanized steel was considered as to be sufficient.