PERGAMON Renewable Energy IS (1998) 512-518

AUTONOMOUS POWER SUPPLY WITH PHOTOVOLTAICS:

PHOTOVOLTAICS FOR RURAL ELECTRIFICATION
- REALITY AND VISION -

Hansjorg GABLER
Fraunhofer Institute for Solar Energy Systems ISE
Oltmannsstr. 5, D-79100 Freiburg, Germany
Tel. i-49(0)761/4588-229 Fax +49(0)761/4588-217

Keywords: Rural Electrification, Solar Home Systems, Hybrid Systems, Photovoltaics

Photovoltaic electricity - inexhaustible, clean, available everywhere in the world, modular in low power
quantities - is an ideal option to supply rural areas which are not yet electrified. Nevertheless, two decades
had to pass after the first demonstrations of solar technology in the seventies until stable markets with
high growth rates developed in different parts of the world for one specific technical solution for rural
electrification, the Solar Home System. Indonesia, Mexico, Kenya, China and India are excellent
examples for the establishment of such markets /l/. Today we estimate that between 100,000 and 200,000
Solar Home System units were sold in 1997. The total accessible user potential is valued at 10 to 100
million units. Three further applications, photovoltaically powered pumping systems for drinking water,
photovoltaic lights and photovoltaically powered battery charging stations have found much wider
dissemination than just as individual demonstration systems. Because of space restrictions, these three
applications will not be discussed in detail in this paper. The text concentrates on the successful Solar
Home Systems and larger hybrid systems for power supplies, which are the subject of much discussion
currently.

The Solar Home System as a small, individual solution

The designation “Solar Home System” has become a well-defined concept. A 50 W photovoltaic module
typically supplies power for three lamps and a black and white television in a single household. A lead-
acid battery with a charge controller stores the energy from the day for the night and tides over two to
three overcast days. Depending on the size of the local market, customs duties, taxes and the share of
locally manufactured components, the cost of a Solar Home System is between US$500 and US$ 1500.
Under favourable conditions, Solar Home Systems offer cost advantages over the classic alternative for
electrification, the extension of electricity grids into sparsely populated regions and connection of power
lines to all rural households. The World Bank assesses the average cost world-wide of grid extension to
be US$900 per household. In unfavourable cases, with long distances, difficult terrain and low population
density, a power line for one household can cost many thousands of dollars.
The investment needed for this cannot be financed via the low energy consumption of rural households,
which ranges between fractions of kilowatt-hours and a few kilowatt-hours. Political decisions to electrify
rural areas despite this imbalance have contributed to electricity utilities becoming indebted in developing
and threshold countries, so that necessary investments to modemise distribution grids and power stations
cannot be financed. The result is the poor quality of the electric power supply, which can be observed in
many regions around the world.

0960-1481/98/S-see front matter 0 1998 Elsevier Science Ltd. All rights reserved
PII: SO960-1481(98)00215-8
 WREC 1998 513

Solar Home System

Fluorescent
Lights

Television b/w

Radio/Cassette or
Other Small
Power Appliances
Fig. 1: Solar Home System - individual power supply for single households

Increased privatisation of the power supply is intended to provide a remedy to the described situation.
Decentralised electrification with privately financed Solar Home Systems fits in perfectly with this aim
and is therefore actively supported by the energy planning authorities of various countries and by
multilateral organisations such as the World Bank.
Light, allowing information and communication for private users, light, which lengthens the working time
into the evening in farms and shops, schools and community centres, cannot be underestimated in its
importance. Solar Home Systems can provide this light.
Electricity also solves other supply problems. Workshops, food storage and processing, infrastructure
tasks in health services and administration, and the development of tourism demand higher daily amounts
of energy and power than can be supplied by a Solar Home System. The next electrification stage consists
of larger individual systems and then local networks.

Hybrid systems for power supply in the kilowatt range

The usual supply levels, classified according to small and medium loads, and local, regional and national
supply with the corresponding voltage levels and power ranges, are illustrated in fig. 2. Photovoltaic
power technology is cautiously expanding into the range “supply of small and medium loads” at present.
To increase supply reliability (back-up with different generators), or to reduce the investment for very
large photovoltaic arrays or very large battery storage units, systems to supply daily loads of several
kilowatt-hours or more am typically constructed as hybrid systems: a diesel generator, a photovoltaic
generator and possibly other generators for wind energy or hydroelectric power complement each other in
supplying power. A battery bank and possibly other units for short-term energy storage ensure that power
is available at all times. Power is distributed to the loads with AC voltage of the usual frequency and
amplitude.
The concept of hybrid systems is not new. During the past two decades, several hundred hybrid systems
were constructed, operated and evaluated with the support of the German Federal Ministry for Education,
Science, Research and Technology BMBF and the European Union EU, European utilities, and national
and international funding bodies, also on other continents. A number of reports on installed systems were
presented at the 13th OlTI Symposium on Photovoltaic Energy /3/, /4/.
514 WREC 1998

D Solar Tlrermal
Wind

* Hyclro
6 011. Cm. Cilomase
i _ Star.,~s Nenns

/’
Reglonal Power Sunphl ,

3.. / 110 JBOkV

3- I ZOkV ,..
, ?~~,, 3 .. I 4ODV r . .
Voltnge Form and Level

low IDDW 1kW lO&W 1CIOLW 1YW 1 DMW 1OOMW **

??

Power Ranges

Fig. 2: Classification of power supply technology for electrification with renewable energy - energy
sources, application classes and trends. (Kleinkauf and Raptis, ISET Kassel, 1996 /2/).

Fig. 3: The Rappenecker Hof. A power supply for the hikers’ inn in the Black Forest, consisting of a
photovoltaic generator, a wind energy converter, a battery bank and a diesel generator, started
operation in 1987 (Photo: Fraunhofer ISE).

Hybrid systems should meet the following power supply specifications /5/:
- continuous (24 hour), highly reliable power supply with single-phase or triple-phase AC voltage.
provision of sufficient power to operate electric motors.
Compared to a diesel generator, which is operated on demand for several hours a day, a hybrid system
should:
 WREC 1998 515

considerably reduce the fuel consumption, the number of operating hours and thus the replacement
and maintenance costs for the diesel generator.
Many of the installed systems have successfully met these goals. The technical feasibility has been proven
beyond doubt. However, there are still unanswered questions regarding larger hybrid systems. more
obviously so than in the case of the Solar Home Systems discussed in the first section.

Configuration alternatives for photovoltaic hybrid systems

The majority of lower power hybrid systems now installed falls into one of the two following categories:
DC-connected systems or AC-connected systems (fig. 4).

Fig. 4a: A DC-connected hybrid

system with a PV generator, wind
energy converter and diesel
generator.

Fig. 4b: An AC-connected hybrid

system with a PV generator,
hydroelectric turbine and diesel
generator.

The system illustrated in fig. 4a, with the input energy flows connected on the DC side of the inverter, has
proven its functionality in many systems installed in the field. Reliable components are commercially
available. However, the need for individual charge controllers, or rectifiers including a charge-controlling
function, can lead to systems with relatively complex structure and thus to higher investment costs.
Concentration of all load flow control tasks in a bidirectional inverter, as shown for the system in fig. 4b,
can lead to simpler system structures. However, at present only a few inverters are available on the
market, which have proven their reliability in operation and could be used for this complex task.
Therefore, it is not yet possible to compare costs for both concepts on the basis of field experience.
The concept of AC connection can be developed further, if the photovoltaic generator is also equipped
with its own inverter. Then the photovoltaic power would also be fed into the AC side of the bidirectional
battery inverter.
 516 WREC 1998

Photovoltaic hybrid systems - a cost comparison with conventional solutions

A comparison between the cost of extending electricity grids m rural areas and that for local power supply
solutions, whether for individual consumers or small supply networks for several consumers, cannot be
made in this paper. However, it is certainly correct to assume that for many decades, there will be several
hundred thousands of power supply situations, in which an electricity demand of several kilowatt-hours to
several hundred kilowatt-hours has to be met daily by local generation.
In the following, a simplified cost comparison of three alternative system solutions will be made to
estimate the future significance of photovoltatc hybrid systems for local, grid-independent electricity
supplies. The table m the appendix summarises the assumptions on cost and operating lifetime of system
components for a small PV/diesel/battery system at a very good solar location. It also includes the cost of
a diesel/battery system and a simple diesel system for comparison.
PV/Batteryl Battery/Diesel Diesel
Diesel
Cost per kWh 3.3 DM/kWh 3,0 DM/kWh 1,6 DM/kWh
Proportional costs:
PV-Generator I4 %
- Battery 41 % 46 8
- other system components 35 % 32 % 32 %
and installation
- Operating costs 10 % 22 % 68 %
(mamtenance and fuel)
Tab. I: Energy costs per kWh for three alternative systems. Fuel price: 1 DIM/I. The proportional costs
are calculated for a system lifetime of 20 years and include the replacement of components wtth a
limited lifetime. The calculation is documented in the appendix.

Of course, the result of the comparison is determined by the assumed fuel costs. Typical diesel prices in
rural areas (including transport costs) range between 0.50 DIM/I and 2.00 DIM/I /6/. For a cost of 1 DM/I,
the battery/diesel system has a slight advantage over the photovoltaic system. This advantage is lost if a
fuel price of 2 DM/I is assumed. The diesel system taken for comparison, with neither battery nor inverter,
has a significantly lower supply quality (only a few hours of operation per day). The cost advantage of this
diesel system vanishes for a fuel price of 3 DM/I. The financing costs (interest) for investment are not
included in the calculation. Taking the financing costs into account favours the systems with higher
operating costs.
The price reduction for PV generators, which is expected in the medium term, will improve the situation
for photovoltaic systems. Naturally, there are further reasons for using solar energy: for example, the high
supply reliability for certain applications, which can be achieved with PV systems, the simplification of
supply logistics due to reduced fuel consumption, and the greatest possible avoidance of noise and fuel
exhaust. This last aspect, together with the positive image of photovoltaics as a clean and resource-
conserving energy supply, plays a central role for applications in infrastructure for tourism.

Open questions and technical development tasks

The results in table 1 make it clear that the photovoltaic generator in a hybrid system is not the major
factor when the lifetime costs are regarded. On the contrary, the costs for energy storage and the
remaining system components, including planning and installation costs, dominate.
The high storage costs are primarily caused by the limited lifetime of the commonly used lead-acid
batteries. Every development - intelligent battery periphery, intelligent operation management strategies -
which can extend the battery lifetime should therefore be supported to accelerate the dissemination of
photovoltaics. This area is being investigated by various institutions, also with the support of the German
Federal Ministry for Education, Science, Research and Technology BMBF. The state of the art was also
documented at the 13th OTT1 Symposium on Photovoltaic Solar Energy /7/. /8/. The same applies for
further development of design and operation management procedures, which aim to identify system
 WREC 1998 517

configurations that achieve high supply reliability with minimised system component dimensions. again
reducing the energy costs /9/, /lo/.
The investment, operation, planning and installation costs must and can be reduced by systems and
components becoming more modular and better standardised. It is undoubtedly the task of companies
which manufacture components and systems, together with research institutes, to stimulate this
development. First steps in this direction are described in 12.
Last. but not least, it is vital that the field experience with hybrid systems around the world be registered
rdprdly and comprehensively, and become available to manufacturers and institutes as input to further
system development. Also, the question of “introductory markets” for photovoltaic hybrid systems -
supply of individual objects or village electrification, isolated solutions or planning for future grid
connection - which is still very much open, would then be clarified.
Rural electrification is not just a technical challenge, but is tnterwoven with political and economic
development in an extremely complex way. It IS influenced decisively by acceptance issues and changing
technological cultures. These multi-facetted boundary condttions can and will ensure that very different
technical solutions arise to solve similar problems.

References

Ill H. Gabler and J. Beurskens

Rural Electrifxatlon wth Photovoltaics
14th European Solar Energy Conference, Barcelona, Spain (1997)

/?I W. Klemkauf and F. Raptls

Elektrifizwung mlt Hybrtdanlagen - Emsatz emeuerbarer Energlen zur dezentralen, netzkompattblen Stromversorgung -
(Electrlticatlon wth hybrid systems - Apphcation of renewable energy sources for a decentralised. grid-compatible
power supply)
Forschungsverbund Sonnenenergie. Annual Conference (1996)

I31 G. Bopp. R. Neufeld and M. Schulz

LanaBhnge Betrlebserfahrung mit 30 photovoltaischen Inselanlagen
(Long-term operating experience with 30 stand-alone photovoltaic systems)
13. Symposwn Photovoltaische Solarenergie. Staff&em (1998)

/4/ M. Landau
Modulare Stromversorpung mlt Photovoltaik fiir die Starkenburper Htitte
(Modular electricity supply wth photovoltaics for the alpine-lodge: Starkenburger Hiitte)
13. Symposum Photovoltalsche Solarenergie, Staffelstein (1998)

/5/ G. Bopp. H. Gabler, K. Klefer, K. Presser and E. Wiemken

Hybrid Photovoltalc-Diesel-Battery Systems for Remote Energy Supply
North Sun ‘97, Espoo. Fmland (1977)

161 H. Dvenhart and G Hille

Economic analysis of Hybrid Systems for Decentral Electricity Generalton m Developing Countrles
I lth European Photovoltalc Solar Energy Conference. Montreux. Switzerland (1992)

A. Jossen
Bleibatterlen fir PV-Systeme - Ergebnisse emes SechsJIhrrgen Labortests
(Lead-actd batteries for PV systems - results of a six-year laboratory test)
13. Symposum Photovoltaische Solarenergie. Staffelstein (1998)

D.U. Saw and S. Schattner

Batteriealterung verstehen durch Modelherung und Simulatton
(Understandmg battery ageing by modelling and simulation)
13. Symposium Photovoltaische Solarenergie, Staffelstein (1998)

D.U. Saw, H. Puls and G. Bopp

Optimierungswerkzeug fix Auslegung und Betriebsfiihrung - Kostenrechnung und Batteriemodell inklusive
(Optlmlsation tool for design and operation management - including cost calculation and battery model)
13. Symposium Photovoltalsche Solarenergle. Staffelstein (1998)

R. Kaiser and G. Bopp

Betriebsfuhrung van PV-Hybridsystemen: ein Verglelch
(Operation management of PV hybrid systems: a comparison)
13. Symposium Photovoltaische Solarenergie. Staffelstein (1998)
Compneols dtmensmn spcctnc cost_5 lnYesmle”l COIL! Melime corn

P” generator 2 kW, “M 10 --/W, 20,0(K) -- 2o.om --

charge controller PV 2kW DM I 4w 2.Cw 4.cmo --

battery 30 kWh DM 500 --RWh I5030 -~ 60.000 -- I5snm --

recufier 5kW DM I --/w 5,LMO -- IO,000 -- 5,wo --

invener 3kW DM z.-~nV 6.ooO -~ 12.000 -- 6.000 --

diesel generator 5kW DM 2 --nV 10.000 -- \o,wo -- IO.lxQ --

mswllauon matena,. housing 3.003 -- 3,ooo -- 2.cw --

investmenl total 61.MM ,,9,ooo

-- 38.ooo

planwlg. transport, 20 B of ,he mvestmem COIL, ,2.?W-- 12,200 -- 7,600 --

mstallation

tolab investment and installauon 7s 131.200--

maintenance I %ofnhe inveslment cost ,2.2cm --

annually for the hybnd system

fuel EOSLS DM I 4Lirer 2.190 -- _

(fuel costs) * (DM 2 --/Liter) (4.380 --)

mhl: operating and mveslment costs 145.590 --

(tonal operating and investmen costs) (147.780 --j

energy costs DM 3 32lkWh

(energy COSLS, (DM 3 37,kWh)

??The COEDS
for a fuel pncs of 2 DM/lme arc gave” I” parentheses

E%DlilnatiQQ$

System conligurauon correspondmg 10 fig 4a DC connection

Dally energy demand 6 kWh
The photovollaic generator supplies 10 8 kWh dally at a IDCBIIO~ wlh very gaod solar madunce (5 4 kWhm~‘dal,y) With a cyslrm effic,ency of 50 9 (performance rauo = 0 51. 5 4 LWh xc uiable. reprercnnng 90 $(, ol the
dally energy demand
7he diesel generator runs 0 I2 h daily on average in the PVlbatleryldwsel system and t 2 h m the battery/dnesel sy,trm The efficiency value of the due, generator IS 20 9, ,I generaec 2 LWh e,cc!r+ from one ltfrc lur,
The Lola, operating bferlme of the system is 20 years PV generator lifetime 20 a, bfetune of eleclron~c components. 10 a. battery lifetime 5 a. dwsel generator lifeume 20.(100 operatmg hours
In the diesel system (without a battery or mverter), the diesel IS operated for 6 h daily. at 20 %t UT full load an averag The cffictency value of the dtcae, general IF 10 ?&
The dtesel system only achieves a markedly reduced wpply quabty (6 hours daily)
Ftnancing costs are not laken mto account