The GAA is situated in a small peninsula located in the southeastern edge of the Greek
mainland. The central part of GAA is the Athens basin that covers an area of 450 km2 with

high population density, 8000 inhabitants per km2.

The climate of GAA is typical "Mediterranean”, with mild winter and dry hot summer. The monthly mean temperature varies between 9.3° in January and 27° in July and the annual precipitation is 376 mm. The prolonged sunshine duration is characteristic of the regional climate, with annual value 2884 hours.

Hourly values of ambient air temperature and humidity were measured in 23 experimental stations, being installed in the Athens urban and suburban region for a period of two years, 1997 and 1998. The sites were selected as a way to study areas with different building density and traffic load that are located along the north-south and east-west axes of the Athens basin and to get information about the boundary conditions around the basin. Seven stations were placed in the central area of Athens while fifteen stations were placed in urban areas and in a radial configuration around Athens centre. Station 2 was situated in the slope of Hymettus mountain (at an altitude of about 500 m), in an almost rural, non built-up region with moderate vegetation and no traffic. The location of the station was selected in a way that the effect of the local flows is substantially reduced. Since this station was nearly free from urban climate modifying effects, it was used as the rural station in this study, and is mentioned as "reference”.

The specific climatic conditions measured in the reference as well as in the urban stations are given in Tables 1-2. These specific years have been selected as both present extreme characteristics. In fact, 1997 is characterised by the highest heat island intensity measured in the city, while 1998 by the lowest one. Thus, use of the selected climatic data may provide the upper and lower limits of the energy cost and ecological footprint.

Jochen Benz, Felix Holz, Tim Meyer, Michael Muller1, Werner Roth

Fraunhofer Institute for Solar Energy Systems ISE
Heidenhofstrasse 2, D-79110 Freiburg, Germany, www. uesp. de
Phone: +49 761 4588-5227, Fax: +49 761 4588-9227, werner. roth@ise. fraunhofer. de

1Steca GmbH, Solar Electronics

Mammostr. 1, D-87700 Memmingen, Germany, www. stecasolar. com
Phone: +49 8331 8558-530, Fax: +49 8331 8558-12, Michael. Mueller@steca. de

Further Project Partners:

Aquasolar AG, Feldkirchen; Elektro-Peter (Pro Regio Bundelfunk GmbH & Co. KG),
Baden-Baden; EMU Unterwasserpumpen GmbH, Hof/Saale; FEG — Fertigungs — und
Entwicklungsgesellschaft mbH, Sommerda; FMA GmbH, Marktheidenfeld;
Hoppecke Batterien GmbH & Co. KG, Brilon-Hoppecke;

Signalbau Huber GmbH, Unterensingen

Abstract

A new approach for DC coupled PV-hybrid systems is shown, which leads to simplified planning and installation and a more cost effective and reliable operation of PV-hybrid systems by means of distributed intelligence and standardised interfaces.

1. Introduction

Today, whenever a PV-hybrid power supply system is installed, it is generally tailor-made for the specific application. Effort for development, design and sizing is large and makes the system more expensive. Also from an operator point of view it is difficult to maintain a number of such systems due to their individual configurations and differences in handling or interfaces. Extension of existing systems today is difficult.

To allow PV-hybrid systems gaining marked share and becoming a viable alternative to stand-alone diesel generators or connections to the mains, the design and installation phase needs to be drastically simplified. Furthermore, much more flexibility to easily include all classes of components is needed. For these ends, standardisation of interfaces and a „Mount-and-Forget" philosophy need to be rigorously applied for both, the power system and the energy management (including information exchange). In particular the ability to handle extensions and changes of the system, while ensuring continuous, unattended, reliable and economic operation are features which essentially depend on the system concept. Concerning energy management, today’s approaches do not enable flexible and Mount-and-Forget solutions. Changing software and modifying parameters becomes necessary as soon as energy supply or demand change significantly (e. g. by changing or adding components to the system). The main task of UESP is to tackle this

This project is funded by the German Ministry for the Environment, Nature Conservation and

Nuclear Safety, BMU under contract number 032 99 22 A. The author is responsible for the

content of this publication.

problem. Concerning the power system, UESP uses DC technology as the backbone of the power system. One alternative is the AC bus bar concept which works well for large systems with a number of AC loads and AC generators. However, the technical effort for the parallel operation of many AC components with strongly differing characteristics is rather high. Beside these technical constraints, in many applications (telecommunication, data acquisition stations, small village power supplies) DC loads need to be supplied anyway. A PV generator delivers DC current and the batteries are always DC (and in future scenarios fuel cells will also supply DC), so DC coupling is the most cost effective, efficient and most reliable solution for small and medium sized systems avoiding a cascade of inverters and converters.

Philipp Schramek; Muhlbergstr. 26, 82319 Sternberg, philipp. schramek@retarus. net

David R. Mills, Damien Buie, Anne Gerd Imenes; Solar Energy Group, School of Physics, University of Sydney, NSW 2006, Australia

To use a photovoltaic receiver with concentrated solar radiation one has to ensure that interconnected cells are uniformly illuminated to maximise the electrical output. At the same time, as much solar radiation collected with the primary lens or reflector as possible should be utilised. Instead of adapting the surface of the reflector to the receiver, we propose to design and locate a special surface for the receiver such that it is uniformly illuminated. The details of the resulting receiver shape will be explained. The applicability of the technique to other types of concentrating reflector will be discussed.

Introduction

A Photovoltaic (PV) receiver for concentrated radiation requires relatively uniform illumination over the whole surface to achieve best performance. Some types of thermal receivers may also benefit from more even illumination over the absorbing surface. An example of an application for concentrated radiation other than a PV receiver could be the absorber surface of a volumetric solar receiver (Pressurised closed Volumetric Cavity Receiver (Buck et al. 2000) or Open Volumetric (Hoffschmidt et al. 2002)), where the reduction of damaging hot spots in the receiver material is a priority.

One approach to achieving even illumination uses a kaleidoscope-like secondary reflector (Lasich, 1996; Ries et al., 1997) which transforms the incoming radiation in such a way that a flat absorbing surface (made, for example, of PV cells) would be uniformly illuminated. However this involves the use of a reflector which absorbs some potentially useful radiation, and some reflector materials may not be resistant to high concentration radiation flux or may require cooling to survive. Another way to achieve uniform illumination is to design the optical device or concentrator (dish concentrator) in such a way that the absorbing surface is uniformly illuminated. Some approaches combine a secondary reflector with a specially designed concentrator. In Horne (1992) a PV receiver for a dish concentrator is suggested which is broken up in cylindrical rings which are positioned so that the illumination on all cells is approximately uniform. This approach involves complex construction and incurs optical losses.

Another concept might be to construct a dish concentrator out of flat reflectors whose sizes are slightly smaller than the PV receiver. Jorgensen et al. (1992) suggest a dish made out of a number of reflector rings, each ring with an own design to achieve uniform illumination in the focus. Burkhard and Shealy (1975) presented a design procedure for a continuously differentiable reflector surface to achieve a specific illumination (e. g. uniform illumination) on a given receiver surface. This procedure is applicable both for rotationally symmetric reflectors and receivers and to linear
reflector and receiver systems. The company OEC designed a dish with a special surface to achieve uniform illumination on a rectangular (not rotationally symmetric) receiver (Ries et al. 2001). One disadvantage of the concentrator specifically designed to create uniform illumination of a receiver is that this would be a very special concentrator which could not be used for other systems where a higher concentration but not such a high uniform illumination is needed.

H. H.C. de Moor1, N. J.C. M. van der Borg1, B. J. de Boer1
H. Oldenkamp2

1Energy Research Centre of the Netherlands, ECN
P. O. Box 1, NL 1755 ZG Petten, the Netherlands
www. ecn. nl e-mail: demoor@ecn. nl
2OKE-Services

Introduction

Building integration is at present the most important market segment for PV [1]. Surely governmental support was important, but especially in Europe it is the first major market to become cost-effective, as the solar electricity has to compete with the relatively high tariffs for private consumers.

A drawback is the usually non-optimal performance due to non-uniform illumination and failures in the modules, wiring and inverters. In this paper it is proposed to go to parallel wiring as this reduces the losses compared to series connected modules.

1 Lay-out strategy and performance loss

Building integration has a number of implications that may influence the energy performance of the PV-system, such as:

• partial shading; if only a part of the PV-array is shaded the energy loss can be over-proportional compared to the loss of incident solar energy

• multiple orientations; the PV-array may cover area’s with different orientations or more in general non-optimal orientation as the PV modules are part of the building envelope.

As the installation and maintenance is part of the building process two factors have to be taken into account in the system lay-out

• ease of installation; the cabling and mounting of the modules most be standardised to avoid costly engineering and allow roofers to install the PV system

• accessibility; the system components are often not accessible for inspection or repair requiring a lay-out that is tolerant to system failures

Partial shading and different orientations within one system are especially important in larger buildings [2], but also not negligible in family houses. These differences in the irradiance on the various PV-modules of the total PV-system, cause a loss of power of the array of modules compared to the sum of their potential individual power values. This so — called mismatch effect is the phenomenon that PV-modules connected in parallel or in series cannot operate in their individual maximum power point because their voltage (parallel) or current (series) is forced to be equal.

The maximum power point of the various interconnected modules may differ from each other due to possible individual differences in the modules, due to differences in soiling, module temperature and irradiance.

The amount of annual energy loss due to mismatch can be influenced by the electrical and geometrical layout of the PV-system. Accessibility, determined by the mechanical structure, is also of importance for the system performance. The effect of a non-functioning
module might not be restricted to that module itself but it can jeopardise the performance of other modules as well. The effect of a bad module on the amount of annual energy loss depends on the electrical layout of the PV-system; the presence of bypass diodes and the lengths of the strings (series connected modules).

Some possibilities for the electrical layout of a PV-system are the following [3]:

• Central inverter. All the modules are grouped in a number of strings. Each string consists of modules in series and the strings are connected in parallel on the central inverter.

• String inverter. This is the same a central inverter but it is based on one string only, typical in the range between 1 and 2.5 kWp.

• Multi-string inverter. This is a special application of a central inverter in which the various parallel strings are equipped with their own MPP-tracker.

• Parallel connection of single modules.

In this concept the modules are coupled to a DC-bus. The total DC-power is then converted to AC by a central DC/AC-inverter [4].

A similar approach was proposed using electromagnetically coupled modules [5].

• AC-Module inverter. In this concept each module has its own inverter [6]. The output of all parallel inverters is fed into the 230 V AC-bus.

The number of modules per string is an important design choice; on the one hand lower ohmic losses (higher voltages) for longer strings, on the other hand higher output for shorter strings (see below) and more simple safety measures due to the lower voltages. The installation is simple if all the modules are in series or all parallel. A mixed lay-out becomes complex for large systems.

An attempt was made to quantify the effects of multiple orientation (by modelling), of partial shading (by experimenting) and of accessibility (by reasoning). This has been done for realistic but arbitrarily chosen PV-systems. Therefore the results are not generally applicable but they give an indication of the order of magnitude of the addressed effects. Furthermore in the concept choice, stringing and size of inverter, more items play an important role such as the price, efficiency and reliability of the inverters. These items are not addressed in this paper.

Long-distance transportation. Biofuel supply from Russia to the EU members is associated with high transportation costs. Consequently, transportation of raw biomass is extremely unprofitable, and Russian biomass export to the EU should be based on mainly advanced refinery/conversion technologies and products: from fuel pellets to bio-diesel, pyrolysis oil, pyrolysis gas and electric power.

Poor infrastructure for collection, sorting and storage of biomass for energy production.

It is illegal but, unfortunately, still widespread practice that wood wastes in form of small — size trunks, branches, etc. are simply left or burnt right at a timber-felling site. In agriculture, unused straw, husk, etc. are usually burned, as well, or ploughed under. Insufficient predictability of the market rules. National and regional laws and regulations in the sphere of ownership and commercial use of natural resources, particularly forests, have not been completely established, leaving the situation with plant biomass availability somewhat uncertain. The federal low on ownership and use of forestry resources is currently under approval procedure by the State Duma and the RF Government.

Low motivation for investments in biomass conversion technologies inside the country. Russia, like most of the countries with transient economics, is characterised by relatively low profitability and high risks of long-term investment projects. This is the major cause of poor marketability of innovative efficient technologies in the field of biomass conversion developed by Russian R&D institutions. Here one can rely mostly on external investors who enter into the Russian bio-fuel market and find it effectual to apply these technologies to increase their business’s profitability.

The calculated cost of photovoltaic electricity is based on an assumed total capital investment of 5500 € per peak kW generating capacity. This figure represents low — estimate currently achievable equipment costs for roof mounted PV (compare [Cre 00] [SFV 02]). The operating costs are set at 1.5% per year of the initial investment, and a
service life of 20 years is assumed. The resulting average costs of electricity are 68 €ct/kWh in Germany and 61 €ct/kWh in EU countries overall. Optimum placement of the modules in locations unaffected by shadows allows generation costs to be reduced by about 18%. These lower cost assumptions apply also to the scenarios, since here the higher yield data form the basis of calculations. Electricity transmission from exemplary production regions with high solar irradiation (Morocco and Algeria) has been included into this consideration (s. Tab. 3). The transmission costs of 6.5 €ct/kWh are due mainly to losses responsible for 4 €ct/kWh, while the remainder arises from the capital investment for the high-voltage DC grid. Photovoltaic electricity generation is significantly more expensive than wind power by about one order of magnitude. Even imported photovoltaic electricity with its significantly greater cost efficiency does little to change this relationship.

• 5.3 Costs of Solar Electricity from Concentrating Parabolic Trough Plants

Cost calculations for this case are more difficult than for the previously treated technologies, mainly because of the high variety of possible plant configurations. The use of a heat storage medium enhances the output characteristics, reducing the losses resulting from unused excess heat and thus increasing the efficiency of the power plant [EC 94]. Appropriate scaling correspondingly lowers the price of electricity. A worldwide generation capacity of more than 7 GW would reduce the costs of the collector array, the primary component, by about half [KMNT 98]. In Tab. 3, representative calculations are provided that depict the electricity costs both locally and after their transmission to Germany at current and reduced costs of the mirrored troughs when storage is employed or not employed. A generous storage capacity insures that no heat will remain unused. This condition definitely does not lead to the most economical design, so that the cost data may be considered conservative. An additional assumption used for calculations of enhanced conservation is that 70% of the electricity has been generated from stored heat, resulting in relatively large average storage losses. The capital investments of very large solar power stations are 185 € per m2 of mirror array. (Concepts with more effective collectors are already envisioned that would reduce the costs of electricity by about 30 — 40%, and which are presently approaching the prototype stage [SM 01].) In a power plant without thermal storage, a mirror surface of approximately 6m2 per kW of electrical power (kWel) is required, whereas the addition of a heat storage with 14 FLH storage capacity raises this value to approximately 15m2/kWel. The cost of the storage medium itself lies at around 60 €/kWhei. (This value is also used for the scenario, although recent research has indicated that it would thereby be overestimated by a factor of 3, since more expedient configurations would allow two thirds of the original storage volume to be avoided [LS 02].) The capital investment for the conventional part of the thermal power plant is 525 €/kWel.

Solar thermal power stations may be used not only to generate electricity, but also to provide combined heat and power. In this case, a portion of the solar energy employed might be used for the desalinisation of seawater in order to provide an additional necessity of human existence that is often short in supply. The regional ecological, social, and economic utility of this technology is consequently improved. The turnover realized from water sales effectively reduces electricity production costs by 1-2 €ct/kWhel, thus approaching the threshold of cost-competitive generation [KNT 01]. This additional benefit is only mentioned at this point, since it has not been included into the assumptions made for the scenarios.

Since northern and central European regions are less suited for electricity generation using concentrating parabolic arrays, comparisons have been made between a region on the Iberian Peninsula in southern Portugal and areas both in southern Morocco and in

Mauritania. The transmission line load has been assumed equivalent to full capacity operation of the solar power plant during half its operating time, with the remaining 50% of the electrical energy divided in a power ratio of 2:1 in order to approximate average transmission loss. The results are compiled in Tab. 3 without consideration of possible cost reductions achieved through the additional production of fresh water. The cost of electricity from parabolic trough power stations for current component prices at good locations are comparable to the costs of electricity from wind power produced at locations capable of delivering about 1400 FLH. If the anticipated cost regression of 50% for the solar field can be realized, controllable solar power from concentrating solar power stations in northern Africa employing heat storage need not to be more expensive even after transmission to Germany.

Implementing Photovoltaics in the Local Urban Environment

The Swiss Editions

Marcel Gutschner* and Stefan Nowak
NET Nowak Energy & Technology Ltd
Waldweg 8, CH-1717 St. Ursen, Switzerland
Tel. +41 264940030, Fax: +41 264940034
marcel. gutschner@netenergy. ch, stefan. nowak@netenergy. ch

Nicole Zimmermann and Urs Wolfer
Swiss Federal Office of Energy
CH-3003 Bern, Switzerland
Tel. +41 313226511, Fax: +41 313232500
nicole. zimmermann@bfe. admin. ch; urs. wolfer@bfe. admin. ch

Municipalities play an important role in sustainable development and energy-related issues. One of the very key factors is renewable energy with e. g. photovoltaics providing more and more opportunities now and in the future. This is the starting point of the Solar ElectriCity Guide for Municipalities. The guide aims at showing ways of implementing photovoltaics (PV) in the urban environment on a communal level. The guide contains concise information and easy-to-use instruments to define, evaluate, plan and implement PV projects in the urban environment. The target audience are mainly local and regional authorities as well as related professionals (urban designers and developers, project developers and builders). Main topics and sections to be presented in the Solar ElectriCity Guide editions are 1) applications, 2) projects, 3) policy, 4) potential, 5) urban design, 6) building design, 7) finance and 8) contact points. The contents address technological issues but the focus is clearly on what PV can offer to cities and how PV can be successfully developed and integrated in the urban environment. Great emphasis is subsequently put on illustrations (pictures and examples), which make the guide a tangible and valuable source of inspiration and information for the key stakeholders in the development of solar power at a local urban scale.

Objectives

The objective of the Solar ElectriCity Guide is to provide local and regional authorities as well as related professionals (urban designers and developers, project developers and builders) with the necessary information and instruments to define, evaluate, plan and implement PV projects in the urban environment. The Swiss guide fits in the larger context of SwissEnergy for Communes, which is part of the SwissEnergy programme. An important objective of this programme is to increase the contribution of renewable energies. This guide shows to municipal stakeholders ways of implementing and promoting programmes and projects bringing together energy, building and planning issues. The Solar ElectriCity Guide has been developed within the EU project "PV City

Guide” and has involved partners from several countries. Different Solar Electricity Guides are (being) published on the European, national and regional level. Three Swiss editions (German, French, Italian) are being prepared.

Petroleum

Oil is the leading source of world’s energy needs. In Kenya it is mainly used in domestic, institutions and commercial establishments. Since no discoveries of oil have been made in Kenya to date, the country continues to rely heavily on imported petroleum for its local consumption. As a result, crude oil and imported refined petroleum products remain among the major Kenya imports.

Electricity

Electric power is an important source of energy. Kenya derives its electricity from hydro, thermal and geo-thermal sources. It is mainly used in commercial and industrial establishments; institutions and households lighting especially in the urban areas. Rural electrification Programme (REP) established in 1973 has supported a number of electricity schemes in rural areas with KPLC acting as the managing agent on behalf of the government.

Wood-fuel resources

Wood energy is the most important source of energy accounting for 70% of the total energy demand in Kenya. According to the Ministry of Energy, about 80% of the population depend on wood-fuel for domestic energy need, providing 93% of rural household energy requirements and 80% in urban areas. In rural areas wood-fuel is mainly used in form of firewood whereas charcoal dominates in urban areas. The principal use of wood-fuel in Kenya is for household cooking and space heating. Government policy on wood-fuel is to ensure adequate supplies of wood are available to satisfy demand through sustained yields; while at the same time conserving the environment. This is done by encouraging use of improved cooking stoves and agro-forestry practices.

Alternative Energy resources

The sharp increase in energy prices and the need to develop substitutes for wood-fuel have aroused considerable interests in alternative energy sources in Kenya.

Solar power: Kenya has abundant solar energy resources. There are two primary ways solar energy can be utilised, namely:-

(i) Solar photovoltaics (PV)

(ii) Solar thermal devices

Solar PV converts solar energy into electricity and is used for variety of applications such as lighting, running appliances, pumping and powering refrigerators. This system is proving very popular and appears to be an attractive option especially for rural households who are unlikely to be connected to the national grid in the foreseeable future. Solar thermal devices convert solar energy into heat. It is commonly known as solar water heater and has been gaining widespread use especially in urban areas, hospitals and public institutions. However, high capital costs coupled with lack of appropriate technology and effective promotion strategies have hampered the exploration and use of solar energy.

Bio-gas: Biogas is a combustible mixture of methane gas and carbon dioxide produced by a biogas digester. This is commonly used for cooking, lighting and running generators. The use of biogas device can help to lower household demand for wood-fuel and commercial fuels, especially in rural areas. Bio-gas technology is cost effective if properly maintained. The operation of the plant is most suitable in high potential agricultural zones especially where farmers practise zero grazing. Lack of skills and effective promotion strategies hamper the use of this alternative energy source.

Wind-power: Wind is one of the greatest sources of natural energy. It’s commonly used for economical power to pump water or generate electricity. During the year 2001, wind power contributed 0.1 Giga-watt hour to the national grid. The major constraints hampering exploration of this type of energy has been lack of appropriate technology, lack of data on wind regimes, poor promotion strategies and lack of initial capital.

A general figure of merit for a renewable-fossil hybrid system designed to reduce carbon dioxide emissions ("avoided carbon”) is the resultant cost of one ton of carbon avoided (CCA). As before, it relies on the comparison of the hybrid system to an efficient, practical fossil fired power system, taken as the reference standard. It is the same standard as for GREF and FCR. A particular cost parameter is obtained, based on essential information on both the costs and performance summary of both systems. As such it is a combined cost-performance parameter:

COST OF AVOIDED CARBON by hybrid system =

COST OF HYBRID SYSTEM — COST OF STANDARD SYSTEM
AVOIDED CARBON (quantity)

The parameter COST means here the annual system cost (including both the annualized capital and operation costs). The standard system is the same baseline standard as previously discussed (Equation 1). The quantity of avoided carbon derives from the GREF which is strongly dependent on the same baseline standard. Equation 3 is useful in many ways. It can be used for monitoring the CCA as a function of the number of operation hours in the year, electricity price tariffs, etc. Also, for the hybrid system optimization; e. g., for deriving the minimum cost of avoided carbon as a function of the number of annual operation hours and several more variables.

Conclusions

PV hybridization is a practical way for enabling the PV hybrid plant to supply uniform power and operate for a long time throughout the year, and thus accelerating PV deployment. In order to meet environmental goals, a methodology has been developed to enable a simplified, while rigorous procedure for gauging and comparing the avoided fuel (emissions reduction), or green energy, for various plant configurations which is essential for directing development of improved systems.

The problem of how to define avoided fuel has been elucidated and two helpful environmental power-parameters have been applied, the fuel consumption ratio (FCR) and the new parameter of green energy fraction (GREF). Their role and significance are illustrated for several PV solar electricity model hybrid systems. The FCR and GREF establish vital environmental metrics. They provide a summation of the fuel avoidance (thence the environmental consequences) of the whole hybrid power system, simple or complex, for the full or part of the year. Together with the CCA (cost of carbon avoidance, $/ton C) parameter the three metrics [(all defined with the same reference)] establish a unified gauging and evaluation criteria enabling standard assessments of various systems on an equal basis. This allows the comparative evaluation of renewable energy plants for upright clean energy. The issue of standards is explicated. The metrics and related equations provide useful yardsticks both for project valuation and for guidance in planning improved, cost effective PV-hybrids. The metrics and the practical procedures are also straightforward tools for tracking the actual clean energy performance of a particular performing plant.

By improved hybrid system design, large blocks of competitive, sustainable solar electricity plants can be widely installed to supply world electricity needs. As such they may lead to cost effective renewable power systems with improved sustainability. With this edge PV-hybrid systems will have the potential of playing a decisive role in massive supply of renewable energy in the near-term, while adhering to practical and well understood system sustainability criteria.

References

[1] Swezey, B., Bird, L. Buying green power — you really can make a difference. Solar Today, Jan/Feb 2003, pp.28-31 http//www. eere. energy. gov/greenpower/pdf/Buying Green__________________________

[2] Geyer, M. Panel 1 Briefing materials on status of major project opportunities, International Executive Conference on Expanding the Market for CSP, 19-20 June 2002, Berlin, Germany. p.4

[3] Wholgemuth N, Missfeldt F. The Kyoto mechanism and the prospects for renewable technologies. Solar Energy 2000; 69(4):305-314

A hydrogen-fuelled polymer-electrolyte-membrane (PEM) fuel cell, with a power rating of

1.2 kW, was selected. The fuel cell system was supplied by a project partner, Phocos, and is based on a Ballard fuel cell. PEM fuel cells require pure hydrogen and atmospheric oxygen to operate. Initially, the gas for the "Rappenecker Hof" is being supplied from gas cylinders. However, the flexible system concept allows for later inclusion of a reformer, producing hydrogen from fuels containing carbon compounds (natural gas, methane, biogas, ethanol, etc.).

The PEM fuel-cell module that is used is a fully integrated system. In addition to the fuel­cell stack itself, it contains all essential auxiliary aggregates and safety functions for operation. This includes all the control electronics, the hydrogen/ oxygen supply and the cooling of the stack. The module is dimensioned for a maximum output power of 1.2 kW at 24 V.

The following passive and active protective measures have been integrated into the system controls so that the fuel cell can be operated safely, fully automatically and without constant observation: [15]

• An air flow monitor and an additional hydrogen sensor are mounted in the exhaust air duct. These are independent of the internal safety sensors in the fuel-cell module and can disconnect the hydrogen supply via the main valve if necessary.

• A switching pressure manometer measures the hydrogen pressure and also disconnects the hydrogen supply to the fuel-cell module if the maximum pressure is exceeded.

• The fuel-cell module is surrounded by a closed, transparent housing, that guarantees a well-defined air input. Thus, a hydrogen leak would be detected early directly at the module.

3.1 The hydrogen supply

The fuel cell is operated with hydrogen of 5.0 purity (99.9990 %). The hydrogen is stored in pressurised gas cylinders in a room that is open to the exterior on one side, outside the main building. The room is dimensioned in a way that up to four groups of 12 cylinders can be accommodated. The content of one group of cylinders amounts to 106.80 Nm3 at a pressure of 200 bar. One group of cylinders is connected to a connection interface (to the fuel cell), which is located on the back wall of the storage room. The energy content of one group of cylinders is sufficient to operate the fuel cell at its rated power for about 100 hours which will be enough for approx. one month.

For the duration of the project, the hydrogen needed to operate the fuel cell is being supplied free of cost by the company, basi Schoberl GmbH & Co. in Rastatt.