N. Giancaspro**-1; C. Private**

*ENEA Area Sperimentale Monte Aquilone, S. S. Garganica 89, km 178+700;
71043 Manfredonia (FG), Italy

**ENEA R. C. Portici, Loc. Granatello; 80055 Portici (NA), Italy
(1)corresponding author: tel +39 0884 543493; e-mail aiancaspro@Dortici. enea. it

An innovative small size hybrid photovoltaic-electrical fuel generator was planned and realized in the ENEA Monte Aquilone Test Site, located in the South of Italy, near Manfredonia (Apulia, Italy). This generator utilizes an innovative equipment, a modular battery inverter Sunny Island made by SMA, provided with a bi-directional converter to charge and discharge the batteries and with a new electronic switch that defines the current operating mode:

— in current-controlled operation mode, the converter is synchronized to an external power supply unit, that can be a public grid or an electrical generator;

— in voltage-controlled operation mode, the converter supports and controls an optimal voltage at industrial frequency.

The Sunny Island is an equipment capable to control the battery loading and at the same time to optimise the voltage and frequency of the bus AC on which different electrical loads and energy systems are connected. A simple single-phase power supply can be arranged with one single Sunny Island and a standard battery pack. The energy systems can be selected between several generators depending on availability: photovoltaic

plants, small wind plants and fuel units.

The goal of this project is to demonstrate that the use of bidirectional converter allows to reduce the size and the cost of a PV system for island application and, at the same time, to enhance the reliability of the produced electrical power by using an additional back up electrical fuel generator.

In order to make a comparison of the management and realization costs among several systems, the sizing of components was determined in the different cases: classic PV system and hybrid with electrical fuel generator.

The experimental activity consists in the data acquisition and analysis of the system working in real conditions. First results confirm the planning which promised a cost reduction of the produced electrical kWh.

A Standard Evaluation Report contains several time series, histograms and many more performance indicators for each of the components of the RES. The task of matching data of a component in a given RES to any other RES therefore represents a multidimensional vector analysis with weighted dimensions and the recognition of patterns in time series. The definition of categories of similar use based on Standard Evaluation Reports requires a process which is independent of the persons carrying out the categorisation process and has to allow an automatic assignment of a RES to a category. In previous work /4/, a categorisation process has been carried out for a limited number of RES, mostly PV systems in Western Europe. The process which was used then is unsuitable for a large number of greatly varying systems and, in particular, does not allow an automatic assignment of a RES to a category.

A component in a RES has to fulfil a number of different performance requirements. Whether they are fulfilled by the component when new can easily be determined by means of most simulation tools. However, to determine the suitability of a component for a specific RES requires an analysis of the performance over the expected lifetime of the component. Components belonging to a certain category are characterised by a similar combination of stress factors and the product with the highest lifetime under the same combination of stress factors is the most suitable product for this category. Stress factors on a battery, for instance, are the requirements of the application (e. g. power requirement and ambient temperature), installation and operating conditions (e. g. temperature management, charging voltage, strategy of operating genset in hybrid systems) and product size (Ah-throughput, current rates). Figure 4 shows these relationships. It is advantageous to make recommendations and the selection of the most suitable product on
the basis of stress factors as these affect the product most immediately. Factors which are very closely linked to operating conditions will change the focus of the analysis to categories of similar operating conditions. Although the differences are probably small due the close relation between the various factors, the focus on stress factors seems most relevant.

An alternative to categories based on similar combinations of stress factors are categories based on a similar combination of ageing mechanisms. Such a definition of categorisation is, however unsuitable. Ageing mechanisms are the effect of stress factors. A battery which is highly suitable for a certain RES is in fact characterised by ageing processes which proceed slowly under the given combination of stress factors and/or where the ageing processes have little impact on the ability to fulfil the required function. The same combination of stress factors may lead to different corrosion rates and the same corrosion rates may lead to a different capacity loss. The most suitable battery is of course that battery which — given the same combination of stress factors — has the lowest corrosion rate and the smallest amount of capacity loss as a result of corrosion. Categories based on ageing mechanisms would make the selection of batteries for a specific RES very difficult.

A full description of the categorisation process is contained in /5/. Certain combination of stress factors results in certain risk of individual ageing mechanisms. Six stress factors have been chosen in the categorisation process. The number of stress factors is kept low to maintain a good overview of the categorisation process, to have independent stress factors and to have a clear view of the effect of factor combinations on ageing mechanisms. Stress factors are calculated from the data contained in the Standard Evaluation Report and their independence has been analysed. The stress factors are contained in a second parameter file which THESA generates.

The categorisation process is technology neutral; however at the moment the process has only been applied to lead acid batteries, the component with the highest life cycle cost in RES. The individual stress factors selected in the categorisation process are:

1. Charge factor as average over a year

The charge factor is defined as Ah charged / Ah discharged. In certain combinations with other stress factors a high charge factor may lead to overcharge resulting in water loss/drying out, temperature rise, high gassing, a thermal runaway, increased corrosion, and increased active mass shedding. On the other hand lower charge
factors may lead to undercharge causing hard/ irreversible sulphation and electrolyte stratification.

2. Ah throughput (cumulative Ah discharged) normalised to nominal capacity

A high Ah throughput may influence active mass degradation, shedding, and electrolyte stratification.

3. Highest discharge rate

This factor characterises the highest current rate in which 1% of the Ah throughput was discharged. This parameter is used to indicate the power requirements to the battery. Concerning the ageing mechanisms a high discharge rate may lead to less homogenous current distribution, higher battery temperature and thus higher water loss, preferential and incomplete discharge, and lower efficiency and on the other hand to better preservation of the original AM structure (crystals size, porosity) /6/, and reduced electrolyte stratification.

4. Time between full charge

This factor is expressed as an average time in between recharge above 90% SOC, because in many RES a higher SOC is hardly ever reached. A long time between full recharge may influence hard/irreversible sulphation and electrolyte stratification.

Shorter times between full charge in combination with other stress factors may be important for corrosion, water loss/drying out and shedding.

5. Time at low SOC

This factor expresses the length of time per year (%) in which the battery remained below 35% SOC. A long time duration may influence hard/irreversible sulphation.

6. Partial cycling

The cumulative Ah throughput in (%) is expressed in the following SOC ranges:

TOC o "1-5" h z 100 — 85 % : A 85 — 70 % : B

70 — 55 %: C

55 — 40 %: D

40 — 0 %: E

A single partial cycling factor is calculated by means of a weighting function:

PC = (A*1 + B*2 + C*3 + D*4 + E*5)/5

High factors result from partial cycling at low SOC level. Low SOC partial cycling may influence hard/irreversible sulphation (particularly at the bottom of the electrode), electrolyte stratification, and a preferential discharge leading to accelerating ageing of certain active mass areas. On the other hand low factors may express shallow cycling (very mild discharge) and mostly operation as a fully charged standby battery.

Each stress factor is a numerical value representing particular segments of the time series of the Standard Evaluation Report or certain averages of the data. An intensity level must be assigned to each individual stress factor in order to consider the influence of the stress factors on ageing mechanisms. A five level intensity index was selected from 1, low intensity, to 5, very high intensity. An individual intensity index does NOT simply indicate a good or a bad value. For instance, both a very low and a very high charge factor are undesirable. The combination of certain intensities of stress factors may indicate an increased potential risk of certain ageing mechanism. If a particular ageing mechanism is performance limiting and thus life limiting, then obviously the effects of battery technology, design and quality as regards this ageing mechanism are particularly important.

Parameters for the stress factors intensity evaluation were determined on the basis of expert knowledge. The parameters were chosen such that the distribution of values measured in real existing RES systems corresponds to the range of intensity levels which were defined.

The stress factors are visualised by means of a radar plot. In a single radar plot the six stress factors and their intensity levels can be depicted. The individual stress factors in the radar plots are grouped to allow easy visual comparison of RES. Six distinct categories can be recognised and described by means of individual bands of stress factor intensities. The allowed intensity band of each stress factor is marked by the dark colour area (see figure 5).

Figure 5: Radar plots visualising the combination of stress factors for each of the six different categories which have been found as result of the categorisation process.

It is important to note that the battery voltage, an obvious factor with relevance to corrosion, drying out (VRLA batteries only) and deterioration of active mass has not been chosen. Firstly, the upper charging voltage is highly influenced by the setting of the charge controller and is therefore closely linked to the operating strategy of the system. Secondly, when calculating average values of the battery voltage or any other figure of merit based on voltage measurements, a number of assumptions have to be made because a simple arithmetic average of the battery voltage is difficult to interpret and depends severely on
the averaging process which has been applied during the measurement. Another important stress factor, battery temperature, has also been excluded from this analysis because battery temperature is more than anything else a parameter which characterises the quality of the installation. Battery temperature is an important aspect in each category and is a stress factor which needs to be considered separately for each category.

The measurements have been carried out by using the instrumentation defined from "HP 84110EM EMC Preproduction Evaluation System”; in particular:

-HP 8594EM EMC Spectrum Analyzer 9 kHz — 2.9 GMz — HP 11947A Transient Limiter

-HP11967D Model 3810/2 Line Impedance Stabilization Network, (LISN)

-HP 85878A EMI Report Generator — Tl insulation transformer

In the following tables, 4-8, the preset values for the instrumentation are reported.

SHAPE * MERGEFORMAT

Table 8. Limit Line 2, average values

The measurements have been carried out at acceptable values of irradiance.

At first we tried to characterize the background level since the standards prescribes that its values must be 6 dB less than the lowest boundary line within all the range.

The electronic energy band diagram of MIS/IL is shown in Fig.2. The dependence of of surface potential Ys on the value of external back bias V of the inversion grid is calculated from the equation of charge neutrality [9],

where

npo, ppo are the equilibrium densities of electrons and holes respectively, p=q/kT, dox is the thickness of the interfacial layer, Dit is the interface traps density (states/cm2- eV), V is the external back bias applied to the inversion grid, фm is the metal work function, %s is the electron affinity of the semiconductor and Eg is the energy gap of the semiconductor.

The potential barrier height of the MIS/IL structure ФЬр is the sum of the band bending and the value of Fermi level Vp in the bulk as shown in Fig.2.

Where Nv is the effective density of states in the valence band. Therefore, the back bias increase the value of ФЬр. The potential barrier height is related to the saturation current density Js of the MIS/IL structure by[10],

Г qtbp_

Js = А“т2в^ kT >e^(-q^TN (4)

Where A** is the effective Richardson constant and qфT is an average barrier height in the insulating layer with thickness d. The dependence of x on the potential Y [7] is given by,

The thickness t of the IL by calculating of the cross point of Y(x) with the Fermi level. The value of Yb is taken as the lower bound of the integral in the equation (4), so we get,
1

1 2

dy

J

The MIS/IL solar cell photocurrent can be represented as the sum of the electron and hole currents, generated by the light flux F in the field of the base and induced IL [9] and is given by,

where

e-(V+al) _ e_(at+vnl)

2 sinh(v-l — v-t)

e (v-t +«l) _ e-(at +vj)

2 smh( y,1 — v-t) (a + h)e~Vp‘ — (vp

(yp + h)e’p — (yp (a + h)eVp‘ + (vp

(yp + h)epp — (yp

vn and vp are are the inverse to the diffusion length of electrons and holes, S is the surface recombination velocity on the frontal surface of the solar cell, Dp is the hole diffusion coefficient, R takes into account the part of the photon flux reflected from the surface of the substrate with thickness l and a is the optical absorption coefficient.

The MIS/IL solar cell parameters will optimized by maximizing the efficiency using genetic algorithm given that [11],

I X V X FF

sc oc

P,

where

Vm and Im are the voltage and current corresponding to the maximum power, respectively, FF is the fill factor and Pin is the power of the light for the AM0 spectrum (135.3 mW/cm2).

For the best location of the planned system the possible sites were analyzed in the GodOlo campus of the University. Here the most promising three possible locations ere introduced with the advantages and disadvantages of them. Each of these sites make possible the south orientation of the PV panels, and — as one of the most important aspect — they are without shadowing during the whole year.

The priorities for the possible locations were:

— no shading,

— south direction of the surface or flat roof,

— easy connection to the electric grid,

— direct usage of the produced energy,

— a nearby covered, safe place for the control and data logger system,

— the possibility of the introduction of the PV system for demonstration.

Considering these priorities the most promising place is the plain roof of the main building of the Faculty of Mechanical Engineering (Fig. 2). The roof structure is divided into 2 separated parts. Both parts makes possible the south direction PV module installation. The lower (and bigger) part is 50 m long and 8,5 m wide, the smaller one is 35 m long and 20 m wide. These two parts are appropriate to the one-row installation of the different type PV modules. The main problem is that the smaller and upper part needs some static

investigation if its loadibility is proper or not (the weight of the panels are not so big problem, but the planned concrete fixing elements are heavy). If the upper part proves to be not proper for the installation, the modules can be placed on the bigger roof part in double row. The inverters and the data logger system would be installed directly below the roof structure and the grid connection is solved, too.

Advantages:

easy availability, simple maintenance, cheaper installation, no shadowing, easy connection to the grid,

the placement of the data logging and the inverters is possible near the PV system. Disadvantages:

the system is far from the department, difficult to take visitors to the system (ladder).

The second option is a student hostel building (see in Fig. 3). There are three student hostel buildings next to each other, but only the top of one building is proper for the system at this moment, as the other two needs renovation first. The buildings are faced almost to the west (the longer axis of the building is almost south to north, about 15° to the east from that).

The dimension of the roof is 15,7 m x 87,5 m. Only a part of the roof is applicable to the PV system installation as the other part gives some place to satellite antennas and exit building. On the other hand the available place is fairly enough to the installation of the PV system assured the necessary distance between the PV modules without shading. The installation of the inverters and the data logger system is possible in a the exit building (size ~2,5mx3mx 2m) near the PV system.

Advantages:

easy availability, simple maintenance, cheaper installation, no shadowing, easy connection to the grid,

the placement of the data logging and the inverters is possible near the PV system.

Disadvantages:

the system is far of the department, difficult demonstrations (ladder),

there are planes to build extra students rooms, one option for it is to make the building higher.

At last, but not at least a potential location is roof of the main building of the Szent Istvan University above the Department of Physics and Process Control. The roof above the department makes possible different forms of the PV system installation. Three different variation has been made to cover the whole roof surface (Fig. 4.a — c).

One big problem with this site, that before the installation the renovation of the roof is necessary. The installation of the bracket elements and the wiring can cause problems through the roof (leaking), and in case of defect the availability of the system is critical.

Advantages:

the system is very close to the department,

the placement of the data logging and the inverters is possible at the department,

no shadowing,

easy connection to the grid.

Disadvantages:

the roof needs renovation first, difficult availability, complicated maintenance, expensive installation, difficult demonstrations.

In every site the data logger system ca be connected to the LAN if the university, so the online supervising of the system is solved. So the measured data during the operation can be transferred easily to the department to the further data analysis and storage.

N. Forero, Licenciatura en Fisica, Universidad Distrital, Bogota, Colombia

J. Hernandez, Departamento de Fisica, Universidad Nacional de Colombia, Bogota.

G. Gordillo, Departamento de Fisica, Universidad Nacional de Colombia, Bogota.

Abstract

An autonomous monitoring system for PV Solar Plants is described. The system is able to collect data about the plant’s electrical output parameter including the plant I-V curve, as well as the solar irradiance and environment temperature.

The measurements and processing of the data are made through Virtual Instruments developed using Field Point modular distributed I/O systems and special programs that were developed using the graphic programming software, LabVIEW. Data of the environment temperature and irradiance that were measured in Bogota City during one year are reported.

Introduction

Automatic data acquisition systems are currently used for both, monitoring system performance and control of its operation. The obtained information can be used to evaluate the plant efficiency during long periods and to optimize future systems in term of performance and reliability.

The data are generally stored in situ using RAM for logger or using a hard drive for a computer system. In the majority of applications the recorded data are transmitted to a data collection central trough a physical connection [1]; however, data acquisition systems for remote monitoring have developed a cellular phone or a RF transceiver connection [2,3]. This feature is essential in PV systems, since they are usually installed in remote areas.

This paper describes an automatic data acquisition system developed using a non­conventional design, based on Virtual Instrumentation, with devices commercially supplied by the National Instrument (NI).

The equipment allows the measurement of typical environmental and system variables of stand alone PV systems [4] (DC Current, DC Voltage, AC Current, AC Voltage, Energy, Power, Ambient Temperature, Solar Radiation), as well as the I-V curve of the PV plant. The data are transmitted and stored in a computer through two different interfaces (communication module-FP and data acquisition board). The monitoring variables are processed and displayed in the computer screen using Virtual Instrument (VI) developed with LabVIEW.

System Objective

The system was developed for monitoring a stand-alone PV Plant which supplies power to DC and AC loads [5]. It provides facilities to get information through three kinds of measurements.

■ Environmental and System variables (Current, voltage, energy, power ambient temperature, solar radiation)

■ Harmonic analysis of the AC signal through Fourier Analysis

■ I-V and P-V curves of the PV plant, as well as determination of its electric output parameters.

The user can display in a computer screen, instantaneous values of any of the above parameters and graphics of the I vs V and P vs V curves.

E. Kazarian, Professor at State Engineering University of Armenia, Yerevan:
erkazarian@yahoo. com

Nader Ali-PhD candidate at State Engineering University of Armenia, Yerevan. G. Grigoryan, Assistant Professor at State Engineering University of Armenia, Yerevan.

One of the perspective methods of utilization of organic conductors for conversion of solar energy into electrical is using organic semiconductor films (OSE), which are organic semiconductors on the base of (CH)x. In comparison with traditional photoconvertors (on the base of Si, Ge) this method has generated considerable interest from the point of view of potentially low — cost solar energy conversion, and the technology of production is relatively simple.

The change of the polyacetylene film conductivity is conditioned by changes of the electron and hole concentrations, also by their mobility. In this connection their relative influence on the value of the film photoconductivity can be very different.

After the fixing of the temperature equilibrium state (thermolization) the carriers mobility don’t change and the photoconductivity is conditioned only by electron n and hole p concentrations changes:

80= e (p pSp+pn8n) (1)

Where e is electron charge, p p and pn is carrier mobility.

In the stationary state the film photoconductivity is determined by (2)

(8 a )s = e (p p +pn)g Tf (2)

Where:

g — is optical generation rate, f is photoconductivity relaxation time.

Measuring the charged particle drift in the radiation field we investigate and evaluate the values, which influence the value of carriers mobility. The measurements show that the carrier mobilities are changed in 6-10 cm2/Vs interval.

The following scheme for photoconductivity measurements has been worked out in fig.1. (D — disk, PC — photoconductor)

Investigating polyacetylene photoconductor (PC), which is

connected in series to the current source (CS) and load resistance R is lighted periodically by means of rotating opaque disk, which is an access hole on it.

Periodic alternation of polymer conductivity in electrical schemes creates alternation photocurrent, which in increased and given to oscillograph. For the establishment of photoconductivity it is necessary to take into account electron — hole recombination and relaxation time. The convenience size of disk holes ensures the relaxation time.

The period of revolution of disk is carried out using the formula (3): 8o = (8 о )s[1-exp(-t/ Tf)] (3)

To get a stationary photo conductance (PC) the lighting time has to be much greater than the relaxation time, i. e. (T>> xf).

On Fig. 2 is shown the character of illumination photoconductor for the chosen period T.

The initial period 0-T of illumination doesn’t depend on the process of recombination and is defined only by generation process. The period corresponds to darkened state of photoconductor. By increasing the photoconductivity the value of the generation rate g can be estimated.

By the results of experimental investigations the generation rate g in case of charge mobility about (ц~ 0.5-1.5) cm2/Vs, also the quantum efficiency in the case of average threshold energy (E0 ~ 1.6) eV have been defined.

The results of the well-handled investigations have shown that the polyacetilene films can be used for cheaper and more economical solar photoconvertors.

Result: This section is small and recombination is not yet significant. It products the following result:

• It Field the current-voltage I= f (U) characteristic for pure and doped polyacetylene

• It approximates relaxation times and stationary photoconductivity

• The-quantum efficiency for polyacetylene film for different frequency of light is determined.

A. Schuler, C. Roecker, J.-L. Scartezzini

Laboratoire d’Energie Solaire et de Physique du Batiment LESO-PB, Ecole Polytechnique Federale de Lausanne EPFL, Batiment lE, 1015 Lausanne, Switzerland J. Boudaden, P. Oelhafen

Institut fur Physik der Universitat Basel, Klingelbergstr. 82, 4056 Basel, Switzerland

Multilayered interference filters of dielectric thin films have been designed for the application as energy-efficient coloration of collector cover glasses. The optical behavior of the designed multilayers is analysed by computer simulations yielding the CIE color coordinates, the relative luminosity, the degree of solar transmission, and a figure of merit which is a measure for the energy effectiveness of the coloration. A high performance should be achieved with a number of individual layers reasonable for large scale deposition. Constraints on the refractive index of the dielectric films are given by the availability of suitable thin film materials. The challenge lies in finding the best combination of material choice and layer thicknesses. We describe several types of multilayer designs for which the computer simulations yield promising results.

Motivation

The issue of color becomes more and more important for thermal solar collectors, and has attracted interest recently [1-3]. This might be related to a generally growing attention towards architectural integration of solar energy systems into buildings [4-7]. A recent opinion poll [1] showed, that 85% of architects would prefer different colors besides black, even if a lower efficiency would have to be accepted. Thermal solar collectors, typically equipped with black, optical selective absorber sheets, exhibit in general good energy conversion efficiencies. However, the black color, and sometimes the visibility of tubes and corrugations of the metal sheets, limits the architectural integration into buildings.

One solution to this problem is to color the absorber sheets. Optical selective absorber coatings are usually deposited by processes such as magnetron sputtering [8-10], vacuum evaporation [11], electrochemical processes [12], sol-gel technology [13], or as selective paint (thickness-sensitive or thickness-insensitive) [3,14]. Niklasson and Granqvist described the pioneering work in this area within a comprehensive overview [15]. Modifying the parameters of the deposition process can result in a colored appearance. Following this approach, the absorber surface combines the functions of optical selectivity (high solar absorption/low thermal emission) and colored reflection. Tripanagnostopoulos reports a different solution: his group used non-selective colorful paints as absorber coatings for glazed and unglazed collectors, and compensated the energy losses by additional booster reflectors [16]. Alternatively, we propose to establish a colored reflection not from the absorber but from the cover glass. This approach has the advantage that the black, sometimes ugly absorber sheet is then hidden by the colored reflection. In addition to that, the functions of optical selectivity and colored reflection are separated, giving more freedom to layer optimization. No energy should be lost by absorption in the coating: all energy, which is not reflected, should be transmitted. Therefore, multilayer interference stacks of transparent materials should be ideally suited for this purpose. A recent feasibility study showed encouraging results [17]. By employing optical methods such as real-time laser reflectometry, spectroscopic ellipsometry and spectrophotometry, the deposition of multilayered interference stacks can be monitored very precisely [18]. In this article we
describe a variety of multilayer designs, which can be employed to achieve the desired characteristics.

In order to investigate, whether fracture relevant micro-damages in a specific process step are induced in the wafers orare reduced by the process, the strength of a series of wafers is determined by bending tests such as concentric ring tests or four-point-bending tests. A growth or a diminution of micro-damages is indicated by a decrease or an increase of strength values, respectively. Concentric ring tests are sensitive to damages within the central surface region of a wafer, whereas four-point-bending tests comprehend also damages in the edges of the wafers.

For the evaluation of the bending tests must be taken into account, that a mathematical analysis on the basis of linear assumptions is not adequate for thin wafers. Rather a non­linear approach based e. g. on a finite element analysis must be applied [1 ].

Figure 1 shows an example for strength measurements applying concentric ring tests on samples taken after successive process steps in an industrial cell production line. Step 1 represents the raw wafers as delivered. Step 2, an etching process, leads to a significant increase in strength, because micro-damages are blunted or entirely removed. Step 3