Eoin Hodge1, Chris Gibbons2

1: Energy Engineering Group, Mechanical Engineering Department, Cork Institute of Technology, Bishopstown, Cork, Ireland, email: ehodge@cit. ie 2: Energy Engineering Group, Mechanical Engineering Department, Cork Institute of Technology, Bishopstown, Cork, Ireland.

Introduction

The European Union’s long term strategy for energy supply security is geared to ensuring, for the well being of its citizens and the proper functioning of the economy, the uninterrupted physical ability of energy products on the market, at a price which is affordable for all consumers, while respecting environmental concerns and looking towards sustainable development, as detailed in Articles 2 and 6 of the treaty on European Union. Environmental concerns, which are now shared by the majority of the public and which include damage caused by the energy supply system, whether such damage is of accidental origin (oil slicks, nuclear accidents, etc.) or connected to emissions of pollutants, have highlighted the weaknesses of fossil fuels and the problems of atomic energy. The European Union target is to double the contribution of renewable energy sources from 6% to 12% of total energy consumption by the year 2010. An integral part of the EU renewables strategy is to achieve one million PV systems and five million square metres of solar collectors, by the 2010 date [2].

From an Irish perspective, an average of 3kWh/m2 of solar energy reaches Ireland everyday, being equivalent to more than 100 litres of oil per m2 per year, and is 600 times the total national energy consumption. With the recent consequence of being Europe’s fastest growing economy manifesting itself in the form of huge rises in energy use and thus pollution, we are now the second largest producers of CO2 emissions per capita, and already far exceed our emission limits under the Kyoto protocol [3].

Most solar cells presently on the market are based on silicon wafers, the so-called first generation technology. As this technology has matured costs have become increasingly dominated by material costs. In the last ten years, continuous work has brought the efficiency of standard cells to the 25% region

[4] . A switch to second generation or thin film technology cells now seems imminent. Thin film technology eliminates the silicon wafer and offer the prospect of reducing material and manufacturing costs, but they exhibit lower efficiencies of around 10% for a commercial device. Third generation or tandem cells are currently at a ‘proof of concept’ research level, with a theoretical conversion rate of 86.8% being asserted [4]. Whatever the material construction and manufacturing method of cells, the thermal effect of overheating will prevail in the semiconductor and it is accepted that a lowered temperature will bring about an increase in conversion efficiency [1].

The aim of this project is to improve the efficiency of PV electrical output, by convectively cooling the cells through perforations in them. As the cells heat

up they lose efficiency. As the panel heats up a loss in efficiency of 0.5% per °C increase in temperature has been recorded [1].

Optical, electrical and technological as­pects have to be considered for reasonable choice of the shape and dimensions of the microstructure. Calculations considering geometric — and wave-optical aspects for a prismatic structure are presented in section "micro prisms — optical simulations".

The deposition of organic thin films from solution by spin-coating or dip-coating is strongly dependent on the structure dimen­sions, the viscosity, the adhesion and co­hesion forces and the deposition process.

Same film thickness as for the planar solar cell architecture should be aspired on the Figure 2: Cross section of a replicated mi — micro prism substrate, following the shape of cro prism structure (SEM-Image)

the structure as precise as possible. The

lattice distance of the microstructure is mainly restricted by the conductivity of the p — conducting polymer film (PEDOT CPP105d). An optimum period of the prism structure is obtained by electrical calculations presented here. First investigations were undertaken on retroreflecting prismatic structures with a groove angle of 90° and a period of 100^m.

G. Oreski1, G. M. Wallner2, A. Skringer3, P. Pert!3, A. Plessing3

1) Polymer Competence Center Leoben GmbH (PCCL), Austria

2>Institute of Materials Science and Testing of Plastics, University of Leoben, Austria

3) Isovolta AG, Werndorf/Graz, Austria

E-Mail: gernot. oreski@stud. unileoben. ac. at

Multi layer films, produced by ISOVOLTA AG (Werndorf/Graz, Austria), are used as backsheets for novel, flexible photovoltaic (PV) modules. Backsheet materials have to fulfil electrical insulation and moisture barrier properties. Multi layer polymer films allow a significant weight reduction and more flexibility of PV modules (Plessing, 2003). A specific failure mechanism frequently observed in multi layer films is the phenomenon of delamination. This paper focuses on the description and evaluation of the delamination behaviour of several tested multi layer films using two different methods.

To analyse the peel tests, on the one hand a conventional load based method with the result of the peel force per unit width is applied (according to ISO 11339). On the other hand, a more sophisticated energy based fracture mechanics approach has been developed (Kinloch et al., 1994). The energy based fracture mechanics method
yields the adhesive fracture energy GA [J/m2], which is supposed to be a material parameter.

The peel force indicates how difficult it is to peel one layer from another, while the adhesive fracture energy describes how well the two layers are stuck together. When the two peel arms are made of different materials with different stiffness, the peel angles will be ф < 90° and 0 > 90°.

To separate the two peel arms from each other energy in form of external work has to be provided. The adhesive fracture energy is calculated from this external work, which contains several deformation and failure processes.

Uext is the external work and Us the stored strain energy. Udt und Udb refer to the dissipated energy through irreversible tensile deformation and irreversible bending deformation of the peel arm. B is the specimen width and da the peel fracture length. In an ideal case (peel arms of infinite modulus, no irreversible bending deformation) Ga is related to the peel force P and the peel angle ф (index 1 and 2 refer to peel arm 1 and peel arm 2).

The energy dissipation Gdb is a complex function which is described by Kinloch et al. (1994). The adhesive fracture energy of the laminate is the sum of the values of the two peel arms.

The advantage of the energy based fracture mechanics method is that the results are totally independent of specimen geometry and testing parameters.

To calculate the adhesive fracture energy two experiments are required. First a T-Peel test has to be carried out to measure the peel force P and one of the peel angles. Second a tensile test of each peel arm has to be performed. The stress-strain curve is approximated to a bilinear model (s. Fig. 2), that describes the real material behaviour of many polymers quite well. The results of the bilinear model

are the elastic modulus Ei, the plastic modulus E2 and the yield strain ey, which is the intercept of the two lines.

Fig 2: Stress-strain curve of PET and bilinear model approximation

Experimental

Three multi layer film types (laminates) were characterized as to their delamination behaviour. The slash (/) indicates the investigated interface.

• poly vinylidene fluoride / polyethylene terephthalate (PVDF/PET)

• poly vinylidene fluoride / silicon oxide barrier layer — polyethylene terephthalate (PVDF/SiOx-PET)

• poly vinyl fluoride / polyethylene terephthalate — poly vinyl fluoride (PVF/PET — PVF)

• polyethylene terephthalate / ethylene vinyl acetate — poly vinyl fluoride — polyethylene terephthalate — poly vinyl fluoride (PET/EVA-PVF-PET-PET)

Whereas the investigation of the first three laminates focuses on the adhesion between various backsheet material layers, for the latter laminate type the adhesion between the EVA encapsulation and the PET backsheet material is characterized.

For the delamination tests rectangular specimens were used. The two parts of the laminate have already been adhered but there is a region of unadhered material. Both, the peel tests and the tensile tests were done using an Instron 4505 tensile testing machine. To eliminate the kinetic energy associated with a moving fracture the peel tests were conducted at a crosshead speed of i0mm/min. The peel force P and the peel angle ф were recorded. The peel tests were carried out using two different configurations. In configuration A (s. Fig. 1) the stiffer peel arm 2 is at the bottom. In configuration B the specimen was revolved through 180° (peel arm 2 at the top). The adhesive fracture energy was calculated using the T-Peel. exe program developed by the Adhesion, Adhesives and Composites Group at Imperial College London (http://www. me. ic. ac. uk/AACgroup/index. html).

For the tensile tests rectangular strips of a width of 10mm and a length of 100mm were used. The tensile tests were conducted at a testing speed of 10mm/min.

Figure 1: Schematic overview of the elements of the PVSAT-2 procedure.

To determine the expected energy yield of a PV system, satellite data and ground based irradiance measurements give information about the solar resource at the site of the PV system. These data are used because local measurements by pyranometers are costly for operators and need periodic maintenance.

First, surface irradiance is derived from METEOSAT-7 images (later METEOSAT-8) with the Heliosat method (Hammer, 1999). Satellite images have a very good spatial and temporal resolution what makes them very suitable for PV applications. To achieve a higher accuracy the satellite measurements are combined with ground measurements by the geostatistical method kriging-of-the-differences (Betcke and Beyer, 2004). These irradiance values are input for a PV simulation (Beyer et al, 2004). According to the submitted plant description, in the second step a PV simulation will determine the expected energy yield of the day.

Here, the central decision support tool carries out the daily performance check. It compares the expected and recorded energy yields on the daily and hourly basis individually for each PV system. It decides on the occurrence of a failure. The detectability of a malfunction depends on the analysis of hourly values (e. g. shading) or is detectable by evaluation of daily values over a longer period of time. In case an error occurred, the
footprint algorithm, part of the error detection routine, determines its cause. This whole procedure of error detection is hardly influenced by the accuracy of the input data.

If the malfunction is found, the operator of the PV system will be informed automatically as part of the automated procedure.

ASTM 5470 is a standard test method for thermal transmission properties of thin thermally conductive solid electrical insulation materials.(ASTM, 1995). The flow meter was designed to this standard. The design incorporated some improvements deemed necessary from earlier testing on a similar meter at the University of Limerick, Ireland.

Figures 1 and 2 show photographs of the Guarded Heat Flow Meter in use designed specifically for Thermal Interface Resistance measurements.

Fig 2 Guarded Heat Flow Meter

Figure 3 shows the heart of the apparatus, the test column, which comprises of the Upper and Lower Meter bars, machined from aluminium, and the reference

calorimeter. The material selected for the reference calorimeter was 302 Stainless Steel. The reference calorimeter is used to determine the rate of heat flow through the specimen.

The heat source was a guarded thermally controlled heater unit. The unit was made up of an inner heater block and an outer guard heater block. The heater block was surrounded with a layer of epoxy thermal insulation and encased within the outer guard heater. Cooling was achieved by a continuous flow of mains water through the cooling unit.

1.1.1 Are these mature products or are they just funny toys?

Looking at the whole diversity of PV powered products; one might wonder what added value is there in such solar gadgets, are they really sustainable and meaningful?

1.1.2

Matching appearance of the product with the appearance of the PV cell In Figure 2 there is apparently a discrepancy between the shape and colour of the products and the shape and colour of the incorporated PV cells and panels. In the left figure the red colour of the lamp mismatches with the blue colour of the PV module, but also a rectangular PV module is put on a round object. It is less disturbing somehow in the right figure in which a round form is combined with a round PV module

In a common research project of Flabeg GmbH & Co. KG, Merck KGaA and the Fraunhofer Institutes ISC and ISE, a production process for AR solar glass was developed, allowing the coating of panels up to 1.5 x 2.5 square meters. After edge processing and surface cleaning, the porous SiO2 AR layers are obtained by sol-gel dipping of the glass followed by a thermal curing at 630°C taking place simultaneously with the glass toughening, and therefore minimising the production costs. The critical steps in the preparation of the glazing are the glass cleaning and a good control of the SiO2 sol particle size, in the range of 30 nm. A precise control of this size is necessary to achieve films with good adhesion properties and the desirable refractive index. Typical AR layers are about 130-140 nm thick, with a refractive index of 1.27, and are deposited on both sides of the panel. The AR coating process can be applicable to any kind of glass (float glass or rolled glass with texture). Fig. 1 shows the light transmittance through low iron patterned glasses with and without the AR layer. The broad transmission maximum at 99 % around 600-650 nm indicates that the % X condition for reflection minimum is fulfilled in this range. Note that the reflection measured through a glass is roughly the sum of the first air-glass reflection and of the second glass air reflection (which adds 4% to the primary air-glass reflection).

An important issue is the stability of the AR layer, because modules are expected to stay several tens of years in the field. To clarify this point, two main test procedures were performed: accelerated durability tests under severe conditions in accordance to IEC 1215 and real life outdoor exposure tests.

The following laboratory tests were passed successfully:

• Condensed water climate test at 40°C and 100% relative humidity (acc. DIN 50017): no significant degeneration or changes

• Mechanical resistance by Crockmeter-test (acc. DIN EN 1096-2): after 1000 cycles no significant visible changes, change in solar transmittance < 1%

• Humidity-heat-test of AR-PV laminate modules at 85°C and 85% relative humidity for 1000 hours (acc. IEC 1215): no visual degeneration, change in energy output < 5%

• Temperature-cycle-test of AR-PV laminate modules (-40°C to +85°C for 200 cycles, acc. IEC 1215): no visual degeneration, change in energy output < 5%

Besides, climate test in SO2 atmosphere (cycling: 40°C, 100% rel. humidity, 8 hours to 18- 28°C, 75% rel. humidity, 16 hours, 5 ppm SO2, 23 cycles), freezing test (-20°C, 48 days) and boiling test (100°C, 10 minutes) did not modify the optical properties of the layer. Laboratory tests even under severe conditions can only give an indication of the durability behaviour of new materials or systems. Therefore outdoor weathering tests have also to be performed to check the suitability of the layers in real conditions.

Test samples have been exposed at different locations spread over Germany and monitored for 3 years at the date of writing this paper. The AR glasses have been evaluated by visual inspection and solar reflectance measurement. During this test period no effects and changes could be observed. Fig. 1b shows a summary of the test data.

100 80 60 40 20 0

Fig.1. a): Optical transmission as a function of the wavelength for a low iron glass with and without the porous SiO2 AR layer (AR layer on both sides). b) Solar Reflectance (reflectance weighted over the solar spectrum) of AR-coated and uncoated glass as a function of outdoor exposure time.

In addition, the outdoor results showed no negative effect of contamination or soiling of the glasses due to the AR coating. A possible increased adherence of dust or particles in or on the porous AR structure is not observed. The porosity of the AR layer is in the range of a few nanometers and therefore much lower compared to the typical size of dust particles in the range of several microns. Because of these different dimensions, a particular interaction has indeed not to be expected.

As a conclusion, the mechanical resistivity, as well as the chemical inactivity and the nanostructure of the AR layer seem to combine to confer porous SiO2 AR glass the necessary stability for long term outdoor applications.

1.1. Electrical energy demand

Analysing the solar and wind potential of the regions the PV system is the on way in the electrification of the remote settlements that can be extended with another sources if the energy demand will be growing. For the present case study the energy consumption is presented in Table 1.

It can be see that the electronic equipment and the TV set appliances is the important portion of energy consumption and in critically meteorological case can be reduced the functioning time. The figure 3 shows the average daily energy demand of the application

2.1 Optimization of layer thicknesses

In dependence on the wave length an organic solar cell absorbs the incident light in different ways. While one part of the light energy is absorbed as heat, another part is

invested in the generation of electron-hole pairs, called excitons. This mechanism occures at any point inside the solar cell and the density of created excitons depends on the light absorption density at a given point. Excitons are separated only by utilization of the „built in field" at the pn-junction of the active layers (Fig. 2). Since the diffusion length of the electron-hole pairs is very short, the chance to reach the pn-interface and the separation as a consequence is very slim. It turned out that the average of the diffusion length is only 10nm in each direction from the interface [13]. For that reason most of the generated electron hole pairs recombinate almost instantaneously while only the excitons created near the pn-interface contribute to the generation of photocurrent at the end of the day. As a consequence one has to place a high amount of absorbed energy to this narrow area around the pn-junction interface what requires a proper distribution of the absorption density in the photovoltaic device. As shown in chapter (2) the absorption density ‘a’ does not follow a trivial function and due to the strong dependence on the conductivity ‘o’ of the optical parameters, ‘a’ does not correlate with the distribution of the electric field |E|2 in dependence on the position vector. While the square of the electric field |E|2 is continuous at the interface between two layers, ‘a’ is not. As a consequence the density of absorbed light energy differs

within adjacent layers even near their interface. Taking this into account and considering the diffusion length of the solar cell materials we calculated the absorption in a region of only 10nm from the active interface both in the p-layer and in the n-layer and summed it up to a value called „effective absorption“ Aeff. In order to gain a better understanding of the behavior of the absorption density we figured out the optimal layer thicknesses of the active layers copper-phthalocyanin (CuPC) and BBP-perylene of our solar cell models.

In our simulations we assumed a thickness of 140nm for the ITO layer as it was purchased on 1mm thick glass. We found an effective thickness of arround 140nm for the PEDOT layer of most of our solar cell models as well. For that reason we stayed with dITO=140nm and dPEDOT=140nm through all our simulations unless noted otherwise. Due to its almost constant refractive index throughout the visible sun light spectrum, the covering glass only decreases the amount of light absorption, but does not effect the location of the maximum and minimum of the absorption density inside the photovoltaic device. For that reason we left the glass out of consideration. Furthermore we assumed perpendicular incident light with a wave length of 550nm. In order to gain a tool for a quick determination of proper layer thicknesses we calculated Aeff for a given range of layer thicknesses and for given active materials. The figures 3-4 show the effective absorption Aeff in dependence on the thickness of the active layers dCuPC and dBBP-

perylene. As one can see there is a distinct maximum of Aeff, which corresponds to special thicknesses of the layers. The top-view (Fig. 4) provides the optimized layer thicknesses like a “map”. For this special model we found 55nm for dCuPC and 70nm for dBBP-perylene. In addition to that we found a big difference between the maximum and the minimum of effective absorption. A factor of almost twenty between the worst and the optimal case of efficient light absorption clarifies the importance of the optimal parameter setup for the solar cell layers.

Upon this thickness maps (Fig.4) which show the usability of given layer thicknesses we calculated the spatial distribution of absorption density in the whole solar cell. Figure 5 shows the absorption density in dependence on the position vector ‘z’ for a non — optimized organic solar cell with dCuPC=200nm and dBBP-perylene=150nm. As one can see ‘a(z)’ is almost at a minimum both in the CuPC layer and in the BBP-perylene layer. A solar cell with this configuration will never be a performer. No mater how good the transport properties of the materials are. Since ‘a(z)’ is a measure for the amount of generated excitons, one can see that most of the generated excitons will be lost by

recombination.

In contrast to that the following solar cell (Fig. 6) with the layer thicknesses dcupc= 55nm and dBBP-perylene=70nm is an optimal performer in terms of energy absorption efficiency. The absoption density in the BBP-perylene layer, which is the main absorber in this special case, reaches its maximum near the active interface.

The presented examples show the best and worst case for a given wave length of the incident light. According to given light spectra and taking into account geographical and geological aspects of the place of destination one is able to gain an optimal setup for any situation.

In this study computer software PVS2000 will be used for energy analysis. With the simulation program PVS it has a powerful instrument to plan and dimension grid — connected. With the program, the user is able to compute the PV energy output, characterize and evaluate the operation of the system, compare different variants of the system or optimize system parameters such as the tilt angle or the battery capacity. In developing the program special emphasis was put on ease of use, a clear and comprehensive display of the results, as well as short computation time. A time step procedure that represents the performance of a PV system correctly has been chosen for the simulation program PVS. It offers sufficient flexibility to treat special systems. The system behavior is simulated by balancing the calculated hourly energy flows. By this study only the energy produced from the PV system is needed to be known. The main initial inputs are:

— PV generator capacity 350 kWp and other PV module characteristics

— Inverter capacity 350 kW and other inverter characteristics

— Weather data (solar radiation, temperature)

— Install condition as follow the previous scenarios