Part 31 - cd3wd431.zip - Offline
MICROFICHE
REFERENCE
LIBRARY
A project of Volunteers in Asia
By: Bernard McNelis, Anthony Derrick & Michael Starr
Published by: intermediate Technology Publications
103/l 05 Southampton Row
London WClB 4HH
U.K.
Available from: Intermediate Technology Publications
103/l 05 Southampton Row
London WCIB 4HH
U.K.
Reproduced with permission.
Reproduction of this microfiche document in any form is subject to the same
restrictions as those of the original document.
E
ICITY
rvey of Photovoltaic
er in eveloping
ountries
Bernard
McNelis,
Antholly
Derrick
and Michael
Intermediate
Technology Publications
1988 in association with UNESCO
Starr
CONTENTS
1.
2.
INTRODUCTION
1
1.1 Background
1.2 Objectives and scope
1.3 Approach
1
1
2
OVERVIEW OF PHOTOVOLTAIC
AND APPLICATIONS
2.1
2.2
2.3
2.4
3.
4.
Why photovoltaics for developing countries?
Photovoltaic technology
Systems and applications
Photovoltaic manufacturers, markets and prospects
19
3.1
3.2
3.3
3.4
19
22
27
28
REFRIGEItATORS
Introduction
Principal case study - India
Other PV refrigeration projects
Conclusions
LIGHTING
Introduction
Principal case study - South Pacific
Other lighting projects
Conclusions
RURALELECTRIFICATION
6.1
6.2
6.3
6.4
7.
Introduction
Principal case study-Mali
Other water pumping projects
Conclusions
PHOTOVOLTAIC
HEALTHCARE
6.1
5.2
5.3
5.4
6.
6
6
5
11
12
WATERPUMPING
4.1
4.2
4.3
4.4
6.
TECHNOLOGY
Inizoduction
Principal case study-Indonesia
Other rural electrification projects
Conclusions
OTHER APPLICATIONS
7.1
7.2
7.3
7.4
7.5
Agricultural applications
Water treatment
Telecommunications
Cathodic protection
IJnusual applications
FOR RURAL
33
33
38
39
45
61
51
62
54
66
69
69
62
66
66
71
71
72
72
74
74
8.
9.
CONCLUSIONS
77
8.1
8.2
8.3
8.4
77
81
82
82
Summary of experience
The technology
The economics
Social and institutional factors
RECOMMENDATIONS
9.1 Identification of appropriate applications
9.2 Strategic approach to development
9.3 Staged development
86
85
85
86
1. INTRODUCTION
1.1 Background
The provision of adequate supplies of energy in suitable forms and at acceptable
prices is an essential prerequisite for most development activities. Energy supply
problems were brought to the forefront of attention in 1973 with the steep
increases in the price of oil. The impact of higher oil prices has fallen particularly
heavily un developing countries, as the cost of their energy imports constitutes a
much higher proportion of export earnings than for industrialized countries.
One of the ways the industrialized countries responded to the energy crises
was to initiate, or greatly expand, research, development and demonstration
(RD and D) programmes in new and renewable sources of energy and in energy
conservation. Many developing countries also introduced energy RD and D
programmes, but continuing financial constraints have inevitably limited the scale
of these activities. There have nevertheless been many projects and development
programmes initiated in developing countries over the last 10 years or so, many
with technical and financial assistance from the industrialized countries.
One important area of renewable energy RD and D has been in solar
photovoltaics, the direct conversation of solar energy into direct current electricity
by means of solar cells. Photovoltaic (PV) systems first came into prominence in
the late 1950s for powering space satellites. After the oil crisis in the early 197Os,PV
systems were developed for a wide range of terrestrial applications. When
correctly designed and installed, PV systems will operate for many years, requiring
little supervision and only occasional simple maintenance. They need no fuel
supplies and give rise to no pollution.
At first the costs of PV systems were very high, but with improved technology,
cheaper materials and higher volume production, prices have been steadily falling
in real terms. The stage has now been reached when PV systems are both
technically and economically suitable for many applications, particularly those
involving relatively small amounts of power in remote locations, where the cost of
operating and maintaining a conventionally powered system is high.
Over the last 10 years, PV systems have been installed in developing counties
to supply power for water pumping, refrigeration, lighting, village electrification,
communications and other applications. Many of these systems were installed as
part of development and demonstration projects and it is now appropriate to make
a comprehensive evaluation of the experience gained.
1.2 Objectives
and scope
The main purpose of this survey is to review the present state of knowledge
regarding photovoltaic applications in developing countries and to assess future
prospects. Many lessons have been learned from projects including those where
photovoltaic-powered systems did not perform as well as expected. It is now vital
to disseminate the information available so that valuable resources are not wasted.
In addition to reviewing the experience gained with PV systems in developing
1
countries, the report also provides advice on the selection of appropriate
equipment, taking into account the various technical, economic, social and
institutional factors involved. The recommendations are intended to help decision
makers identify which photovoltaic applications are suitable for the specific
conditions obtaining in the regions for which they are responsible and to give
guidance on how to implement the necessary projects.
A further objective of the report is to identify areas where further development
and demonstration activities are needed. Such activities include the training of
local personnel in the design, installation and evaluation of systems and the actions
needed to provide the basis for local manufacturing facilities, as well as field
demonstrations for selected applications.
Chapter 2 first addresses the question of why photovoltaics are of particular
interest for developing countries and then presents a summary of photovoltaic
technology and application=. Chapters 3 to 6 deal with each main applications area
inllur- li, starting with water pumping and proceeding to vaccine storage and other
medical applications, lighting and rural electrification. Other applications are
considered in Chapter 7.
The conclusions are presented in Chapter 8, with an overall summary of the
experience followed by specific conclusions relating to the technology and the
economics. The applications that are considered to be the most appropriate for
developing countries are then identified. The chapter concludes with a review of
the issues which are important for the successful implementation of a photovoltaic
project.
The recommendations arising out of the study are listed in Chapter 9. The tist
section covers the methodology that should be adopted for identifying appropriate
PV applications. The second section deals with the overall approach to the
development of photovoltaics in a developing country, and the final section lists the
priority topics for research, development and demonstration needed for the
implementation of photovoltaics in developing countries.
1.3 Approach
The information on which this survey is based comes from two main sources:
published reports, technical papers and articles; and in-house knowledge available
with the staff of IT Power, built up over many years of experience of photovoltaic
projects worldwide.
In considering each application, first a general introduction is given, covering
the main issues involved and reviewing the status of development worldwide. This
provides the background for the principal case study based on a specific country or
region and covering the technical, economic and social/institutional aspects in
detail. This is followed by a briefer review of significant projects in other countries.
The conclusions arising from the review of the experience for each application
are then listed and discussed in relation to the following headings:
(a)
Technical
Reliability
Availability
Durability
Ease of operation and maintenance
00
Economic
Capital cost
2
Operation and maintenance costs
Life cycle cost/benefit
Comparison with conventional alternatives
(c)
Social/mstitutional
Availability and quality of institutional support
Demand for product or services
Compatibility with social requirements of user
Availability of skills needed for operation and maintenance.
c
2. QVERVIEW OF PHOTOVOLTAIC
TECHNOLOGY’ AICIU’D
APPLICATIONS
2.1 Why photovoltaics
for developing
countries?
Energy is needed for practically all the activities that are basic to human survival,
such as cooking, water pumping and food production. After basic needs are
satisfied, further energy is required to improve the quality of life, through lighting,
transport, telephone communications and consumer tools such as refrigerators,
radios and televisions. As a country develops, still further inputs of energy are
required for industries and for commercial and public buildings. In urban areas, the
necessary energy supplies may be readily provided through oil products, coal and
networks for electricity and natural gas. In rural areas, traditional sources of
energy, principally firewood, agricultural residues and cattle dung, continue to be
of major importance, supplemented by commercial sources such as electricity and
oil products in areas where the physical infrastructure makes this possible.
The majority of the population of all developing countries live in the rural
areas. The combined effect of population growth and supply problems of
commercial fuels is putting ever-increasing pressure on the traditional fuel
supplies. Deforestation resulting from over-cutting of trees, sometimes aggravated
by long-term climatic changes, is becoming a major problem in many countries. The
use of agricultural residues and cattle dung as fuel reduces the amount of nutrients
returned to the soil.
Photovoltaic systems are widely recognized as an attractive means to address
some of the rural energy problems, since they offer the following advantages:
0 Being built up from solar cell modules, they are able to provide relatively small
amounts of electrical power at or close to the point of demand
9 No fuel requirements
# Relatively simple operation and maintenance requirements, within the
capability of unskilled users
l No harmful pollution at the place of use
0 Long life with little degradation in performance.
The remainder of this chapter provides a summary of photovoltaic technology
and a general review of systems, applications, markets and prospects. This is
intended to provide the background for the subsequent chapters of this report
which deal with specific applications in developing countries.
2.2 Photovoltaic
technology
Brief history
The photovoltaic effect was first observed by the French scientist Becquerel in
1839 who noticed that when light was directed onto one side of a simple battery
cell, the generated current could be increased. Work on the photovoltaic properties
of selenium in the 1870s led to the frst selenium photovoltaic cell in 1883. The
photosensitive properties of copper and cuprous oxide structures were
discovered in 1904. By 1905, it was known that the number and energy level of
electrons emitted by a photosensitive substance varied with the intensity and
wavelength of the light shining on it.
5
In the years that followed research work continued with the objective of
developing practical photovoltaic devices. Selenium and cuprous oxide
photovoltaic cells were developed, leading to several applications including
photographic exposure meters and other small light sensors. By 1941, selenium
devices had been developed with a light-to-electricity efficiency of about 1 per cent.
A new technique was later developed, known as a ‘grown p-n junction’, which
enabled the production of single-crystaI cells. Using doped silicon crystals,
American research workers in the mid-1950s were able to achieve solar conversion
efficiencies up to 6 per cent.
Western Electric began to sell commercial licences for silicon photovoltaic
technology in 1955 and there were some attempts to develop practical systems for
powering specialist eqmpment in remote areas. It ~2323 not until :ih+~late 1950s
however that crystalline silicon solar cells were developed with %gh enough
conversion efficiencies for their use in powrer generators. A major impetus for the
development of these cells w& the space programme. The first solar-powered
satellite, Vanguard I, was launched by the USA in 1958. Practically all satellites
launched since then have been powered by solar arrays made up of many
thousands of crystalline silicon photovoltaic cells.
Following the 1973oil crises, interest in photovoltaics as a terrestrial source of
power increased greatly and many countries, including several developing
countries, instituted photovoltaic research, development and demonstration
activities as part of wider energy research programmes. Total world expenditure
from all sources on photovoltaic research, development and demonstration
activities is probably running at between $200 and $300 million per annum.
Over the last ten years, there has been more than a tenfold reduction in the real
price of photovotlaic modules. This has been achieved through a combination of
improved cell technologies and larger manufacturing volumes. Starting from
virtually zero in 1974,sales of photovoltaic systems have grown to about 25 MWp in
1985,with a total value of at least $800 million. Worldwide there are over 20 module
manufacturers of significance and there are several times this number of firms
designing and marketing photovoltaic systems using bought-in components.
Research efforts are continuing on a broad front to develop better photovoltaic
devices and lower cost systems.
Much has already been achieved and many thousands of systems are operating
reliably today, ranging in size from a few watts to several megawatts. From being an
exotic, highly expensive technology for very specialized situations, photovoltaic
generators are now an appropriate solution for a growing number of applications
and in time could become a major factor in world energy supplies.
Solar radiation
The energy generated continuously by the sun is radiated as a stream of photons of
various energy levels. At a point just outside the earth’s atmosphere, the intensity of
the solar radiation incident on a plane normal to the sun’s rays is known as the solar
constant. The average value of this is 1353W/m’, with seasonal variations due to the
elliptical nature of the earth’s orbit. As the solar radiation passes through the
earth’s atmosphere, a considerable amount is lost by scattering and absorption,
some wavelengths being affected more than others. The amount of energy lost
depends on the path length of the direct solar beam through the atmosphere and
the amount of dust and water vapour at the time.
The solar irradiance at ground level is made up of a direct component and a
diffuse component. The sum of these two components on a horizontal plane is
termed the ‘global h-radiance’. The diffuse component can vary ‘from about 20
6
per cent of the global on a clear day, to 100 per cent in heavily overcast conditions.
On a clear day in the tropics, with the sun high overhead, the global h-radiance can
exceed 1000W/m2,but in northern Europe it rarely exceeds 850 W/m*, falling to less
than 100 W/m2 on :Lcloudy day.
Knowledge of tke soiar radiation reaching a photovoltaic cell is important,
since not only is the total QYG;&~from the cell dependent on the intensity of the
incident radiation, but also &&rent types of cell show varying levels of response to
the different wavelengths of incoming energy. To compare solar cells, it is normal
to quote the maximum power output in peak watts (Wp) at Standard Test
Conditions (STC), defined as an h-radiance of 1000 W/m2,with a reference sunlight
spectral energy distribution and a cell temperature of 25°C.
The photovoltaic process
The material most commonly used to make photovoltaic cells for power
applications is crystalline silicon, either in mono-crystalline or, more recently,
semi-crystalline form. The essential features of this type of cell are shown in Figure
2.1. It is made from a thin wafer of high purity silicon, doped with a minute ~~ran#y
of boron. Phosphorus is diffused at a high temperature into the active surkcti of 9~
wafer. The front electrical contact is made by a metallic grid and the back c~&~t
usually covers the whole surface. An anti-reflective coating (ARC) is applied k~ the
front surface.
The phosphorus introduced into the silicon gives rise to an excess of what are
known as conduction-band electrons and the boron an excess of valence-electron
vacancies or ‘holes’, which act like positive charges. At the junction, conduction
electrons from the n (negative) region diffuse into the p (positive) region and
combine with holes, thus cancelling their charges. The area around the junction is
thus depleted in charge by the disappearance of electrons and holes close by.
Layers of charged impurity atoms (phosphorus and boron), positive in the n region
and negative in the p region, are formed either side of the junction, thereby setting
up a ‘reverse’ electric field.
When light falls on the active surface, photons with energy exceeding a certain
critical level known as the bandgap (1.1 electron-volts in the case of silicon)
interact with the valence electrons and elevate them to the conduction band. This
process leaves ‘holes’, so the photons are said to generate ‘electron-hole’ pairs
which are generated throughout the thickness of the wafer in concentrations
depending on the intensity and spectral distribution of the light. The electrons
move throughout the crystal lattice and the less mobile holes aIso move by valenceelectron substitution from atom to atom. Some recombine, neutralizing their
charges, and the energy is converted to heat. Others reach the junction and are
separated by the reverse field, the electrons being accelerated to the negative
contact and the holes towards the positive. A potential difference, or open-circuit
voltage (Voc), is thus established across the cell which is capable of driving a
current through an external load.
The current-voltage relationship (I-V) characteristic for a typical cell is
dependent on h-radiance and temperature, as illustrated in Figures 2.2 (a) and wj.
For crystalline silicon cells, when illuminated by light with intensity 1000W/m2AM
1.5 direct spectrum, at 25”C, the open-circuit voltage (Voc) is about 0.6V and the
short-circuit current (1s~) about 30 mA/cm2.
As the cell temperature increases, the current increases slightly as the voltage
decreases significantly, in consequence, the maximum power decreases. It is
therefore desirable to operate the cells at as low a temperature as possible.
7
\
\\’
Sunlight
ElectronFlaw
,I,\
I
1
c
r
J
==j Load 1
Motor,Pumpset,
I inM
$ll~,l\Cll
Conventiinal
DirectionOf
Current
DJ&pfotor
I!
’
n-Silicon
\
u
T-otol
Thickness
1
250450pm
\
4Antireflection Cooting
-Met01
Boseplote
(Cutawaynot
to scale)
Figure 2.1 Crystalline
Silicon Photovoltaic
Cell
The cell efficiency is the ratio of the maximum power to the product of gross
cell area and it-radiance, usually expressed as a percentage. The photovoltaic
process, like other energy conversion processes, is subject to a maximum
efficiency dependent on the physical characteristics of the materials. The
achievement of improved working efficiences, closer to the practicable maximum,
is therefore a major objective of research and development work. For example, the
maximum practicable conversion efficiency for conventional crystalline silicon
cells is about 25 per cent, but the efficiency actually achieved for mono-crystalline
cells commercially manufactured is typically about 14 per cent, although 20 per
cent has been reported for cells made in a research laboratory.
Crystalline silicon cells
The mono-crystalline silicon solar cell is a highly stable device and is based on wellestablished semi-conductor technology developed over many years for integrated
circuits. Wafers about 250-350 pm thick are cut from long single crystal ingots
75 mm, 100 mm or even 150 mm in diameter. The ingots are sometimes made by the
‘float zone’ (Fz) process, but more usually the Czochralski (Cz) process is
employed, whereby an ingot is slowly drawn out of a melt of doped silicon in an
inert atmosphere. The atoms of silicon solidify into a perfect cubic lattice following
the structure of a seed crystal. Commercial photovoltaic cells made from the
wafers typically have efficiencies in the range 11-15 per cent.
Several groups have developed cast ingot processes which are less energy
intensive and which are more tolerant of impurities. A melt of doped silicon is
8
formed in a mould up to 300 mm cube and allowed to solidify under carefully
controlled conditions. The resulting ingot has a semi-crystalline structure which is
clearly revealed when it is sliced up to form wafers, usually 100 mm square. The
resulting solar cells typically have efficiencies in the range IO-12 per cent with some
manufacturers even being able to achieve more than 12 per cent using surface
passivation or gettering techniques.
All ingot processes, whether for mono or semi-crystalline silicon, have the
drawback that they involve sawing to form wafers. This is a time-consuming and
wasteful operation, with over half the material lost.
As an alternative to the ingot processes, several research teams have been
working for some years on the development of continuous sheet processes, which
do not need subsequent sawing. The main problem with all these processes has
been to achieve an acceptable quality of crystalline silicon sheet with a sufficiently
high rate of production to render the process economic. The only commercial sheet
process which has emerged to date is that developed by Mobil Solar Energy
Corporation (USA), which involves drawing out a nine-sided thin-walled polygon
from a silicon melt. Rectangular wafers are then cut from the walls of the polygon
-Maximum
Power
Denstty = 12.tmW/cm2
lrradiance
=
0.6
Voltage
EFFECT OF CELL TEMPERlTURE
30 -
1000
I
0.0
(V.)
ON V-l CHAR4CTERISTIC
W m”
q20.
f
4
Cell Temp = 25°C
u,
,
04
0.0
06
Vol taqe ( V. 1
EFFECT OF CHANGE I-N IRRADIANCE
Figure 2.2 Tgpical
ON V-I CHAItACTERISTIC
V-I Charucteristic
9
Curves
and made into cells. Cell efficiencies are reported to be comparable with those for
ingot processes.
Thin tllm solar cells
Thin film solar cells - due to low material needs and economic mass production
possibilities - have excellent prospects for the future. R and D-efforts are carried
out on cells based on amorphous silicon and on polycrystalline materials like XV-VI
compounds and chalcopyrites. Amorphous silicon-based thin film devices have
entered the consumer electronics market since a couple of years and are at present
undergoing commercialization for the power market. Thin film cells based on
CdTeICdS are as well entering the market. Present research efforts are being
directed to improving the cell efficiency and - especially with amorphous silicon
cells - long-term stability of thin film devices. Hn early 1986 these devices were
beginning to enter the market for developing country applications.
Modules and arrays
Solar cells can be interconnected in series and in parallel to achieve the desired
operating voltage and current. The basic building block of a flat-plate solar array is
the module in which the interconnected cells are encapsulated behind a
transparent window to protect the cells from the weather and mechanical damage.
One or more modules are then attached to a supporting structure to form a panel
and a number of panels makes up an array field which, together with the balance-ofsystem @OS) components, makes up the complete system. The array field may be
sub divided electrically into a number of sub-arrays working in parallel. A selection
of modules is shown in Figure 2.3. Flat plate arrays are normally fuced, with the
modules supported by a structure such that they are orientated due South (in the
Northern hemisphere) and inclined at or about the angle of latitude to maximize the
amount of solar radiation received on an annual basis. A steeper angle of
inclination will enhance the output in winter, at the expense of some reduced
output in summer.
For some circumstances, it is appropriate and cost-effective to mount the
modules on a support structure that tracks the sun through the day. Given clear sky
conditions, the output from the array tracked in this way is more uniform and can
exceed that from a fixed array by at least 20 per cent; moreover, the extra output
comes in the early morning and late afternoon, the times when demand for gridsupplied electricity is often highest. However, in view of the additional
complication and the need for more skilled operation and maintenance, tracking
collectors are generally not appropriate for remote sites, where fixed flat-plate
arrays are preferable.
Concentrator devices
Although a large number of concentrator photovoltaic devices have been
developed, +helong-term prospects for this approach are not favourable, at least for
high concentration systems for terrestrial applications. Apart from special
applications where the concentrator’s higher efficiency and potential for providing
thermal as well as electrical energy can be exploited to the full, the simplicity and
reliability of flat plate modules constitute attractive advantages. It should be noted
that in addition to maintaining the tracking system, the optical components of the
concentrating system (ie. lenses or reflectors) have to be regularly cleaned, adding
to the operational costs. However, until such time as very low cost flat plate
systems are developed, low concentration systems (eg. double mirror or Fresnel
lens devices) wil! often be found to offer economic advantages for large
installations at places where the necessary skilled maintenance staff are available.
10
A number of such systems have been built in recent years, including a 6.5 MWp
central generating plant built by Arco Solar at Carrisa Plains, Cahfornia USA, which
uses double-mirror concentrating collectors.
2.3 Systems and applications
Market categories
There are currently three main market categories for photovoltaic systems. Firstly,
there is the large and growing consumer market, for calculators and other small
electronic devices, cooling fans, battery chargers, lights and other small PV
systems. Sales in this market are largely dependent on good design, effective
marketing and reasonable prices. Secondly, there is the market for professional
systems, such as generators for telecommunication links, cathodic protection,
navigation lights, military equipment, etc. These systems normally have to be
justified on the basis of life-cycle castings using conventional economic criteria,
although environmental considerations can often be important. Thirdly, there is the
very large potential market for systems which primarily have a social benefit, such
as the provision of electricity for remote houses, water supply pumps for villages,
emergency telephone links, etc. These systems are generally expensive, but in
places where diesel generators or grid extension would be impracticable, the
photovoltaic solution can provide important social benefits to the community.
In the longer term, if the very low-cost targets can be achieved, a fourth market
category is expected to open up, namely that of grid-connected systems providing
electrical power to buildings of ah types or serving as central generators.
Figure 2.3 Selection of PV Modules
Stand-alone systems
Most photovoltaic manufacturers now offer a wide range of standard systems, for
battery charging, water pumping, street lighting, domestic lighting, refrigeration,
electric fencing, alarm and security equipment, remote monitoring, beacons and
other navigational aids; the list is constantly growing as other applications are
being found. Although some further improvement and demonstration of these
11
systems is continuing, in most cases they can be considered as developed products.
Commercial sales are growing steadily to private and public customers, who find
that photovoltaic systems provide the most economic or convenient solution to
their needs.
For developing countries, the main applications of interest for rural
development are:
- water pumping for potable water supplies, livestacY: watering and irrigation
-vaccine storage in refrigerators and other medical applications
-lighting systems for domestic and commercial uses
- village electrification.
There are a number of other applications which also have their place in rural
development, such as radio telephone links, educational television, mechanized
milling of rice and other grains, electric cattle fencing and cathodic protection of
steel structures and pipelines.
Further information on the technical, economic and social/institutional
aspects of the above applications are given in subsequent chapters of this survey.
Grid-connected systems
There has been much discussion of the possibility that photovoltaics will
eventually become cheap enough to be economic for grid-connected applications.
At present (1988), with oil and coal prices depressed, this seems to be a remote
prospect, but in ;5helong term the position could change, particularly for countries
rich in solar energy but low in conventional fuels and unwilling (or unable) to
introduce nuclear technology. Many countries have made a strong political
commitment to encourage the use of renewable energy resources and some have
gone further by deciding not to build any new nuclear power plants (eg. Sweden).
Grid-connected systems are simpler and less expensive than stand-alone
systems, since they require little or no battery storage. The grid itself can serve as
‘storage’, with the photovoltaic plant supplying power to or drawing power from
the grid depending on the load and solar irradiance. There are a number of gridconnected photovoltaic systems in the USA.
2.4 Photovoltaic
manufacturers,
markets and prospects
The photovoltaic industry worldwide
Stimulated by the research and development programmes in the USA, Japan and
Europe over the last 10 years, a photovoltaic industry for terrestrial applications
has emerged in practically every industrialized country and in many developing
countries. As may be expected in a new area of technology, there have been a
number of failures and disappointments as well as some notable successes. Many
of the main photovoltaic manufacturers are subsidiaies of, or are majority owned
by, major oil companies, who see photovoltaics as a natural extension of their
energy interest, and a field which may become very large in the future. In the USA,
there are now some 10 to 15 companies well established in the manufacture and
marketing of photovoltaic modules and systems. In Europe, there are about 8 to 10
photovoltaic manufacturers with annual production at least 100 kWp, plus a similar
number of smaller organizations more concerned with marketing systems using
bought-in components, the so-called OEMs (original equipment manufacturers).
In Japan, at least six companies are active in photovoltaics, including several
of the main manufacturers of electrical and electronic goods, such as Kyocera,
Mitsubishi, Sanyo, Sharp, Toshiba and Fuji.
A number of developing countries are developing their own photovoltaic
12
industry, notably Brazil, China, India and Pakistan. Locally manufactured
photovoltaic modules are not significantly cheaper than similar products made in
industrialized countries, but because foreign products are usually subject to import
taxes, local manufacturers are protected and are thus able to secure the local
market, supplying photovoltaic systems for demonstration projects and
professional applications such as telecommunications and water supplies. In this
way, local manufacturers are able to build up experience in the design,
manufacture, operation and evaluation of photovoltaic systems, in readiness for
the day when new technologies can be introduced and become a major energy
resource for rural development.
Photovoltaic markets
Total sales of PV systems amounted to about 25 MWp in both 1984 and 1985,
indicating that the business is currently worth at least $800 million per annum. The
growth in sales over the last 10 years is illustrated in Figure 2.4. The very high
growth rate in the 1970shas now settled down to about 30 per cent per annum. The
approximate breakdown of the total market in 1984 is set out in Table 2.1.
PV Module
Production
MWP
76 77 78 79 80 81 82 83 $4 85
Figure 2.4 Photovoltaic
YEAR
Sales for Terrestrial
Applications
Technology and prices
Over the last ten years, the price of photovoltaic modules and systems has been
steadily falling in real terms. Module prices for both forms of crystalline silicon are
currently around $5 to $7/wp for large orders (FOB). Bearing in mind that the cells
13
Sales in MWp
Application
Telecommunications
5.0
C?thodic
2.0
protection
PV/Diesel hybrid
generators
1.0
Water pumps
Navigational
Off-grid
0.4
aids
residential
Grid-connected
0.4
-
USA
residential
Rural electrification
2.0
-
USA
(excluding
0.1
USA)
0.2
Refrigerators
0.2
Military
0.1
applications
Central generators
-
USA
6.3
Consumer
products
-
crystalline
Consumer
products
-
amorphous
Si
0.1
Si
7.0
Miscellaneous
0.2
TOTAL SALES IN 1984
Table 2.1 Estimated
25.0
Total Sales qf Photovoltaics
in 1984
account for about 60 per cent of the module price, some further price reductions,
possibly down to about $2~$3!Wp,are foreseen through the introduction of cheaper
silicon and larger, fully automated manufacturing plants. Much lower costs, even
down to $l/Wp or less, are potentially attainable with thin film cells. In view of the
large efforts being made world-wide to develop different thin film technologies, it is
probable that large-area thin-film cells will become ava,ilable by 1990 with much
improved efficiency and stability compared with current products. Some
researchers maintain that crystalline silicon cells could continue to be competitive
with thin film processes.
Market prospects are largely dependent on prices of photovoltaics in relation
to alternative energy sources, but other factors are important, such as government
incentives, availability of finance and the general perception of the technology held
by potential customers. Although it is not, possible to predict with precision what
the future market will be, Table 2.2 indicates what the future sales of photovoltaic
systems worldwide might be for two scenarios. The la>wprice scenario assumes
14
that large area thin film cells with adequate performance for power applications
start becoming available within the next two to three years; the high price scenario
is based on the assumption that the technical targets for thin fihn cells remain
elusive, leaving crystalline silicon as the dominant technology for power
applications. There will continue to be a growing market for consumer goods
powered by small area amorphous silicon cells, even if prices remain at today’s
levels.
For the low scenario, with thin film module prices falling to around $1.5/Wp,
total annual sales are projected to grow rapidly, from the current level of about 25
MWp to as high as 5000 MWp by AD2000, with continued expansion thereafter.
Most of the output would be in and for developing countries for rural electrification
and irrigation pumping, using stand--alone systems, but there would also be many
applications in industrialized countries for consumer systems, professional
systems and remote houses and villages. Grid-connected applications could begin
to become a significant market in some countries by the late 1990s.
For the high scenario, with crystalline silicon module prices falling to about
$3/Wp and thin film cells not able to compete for power applications, the total
market would grow much more slowly, possibly levelling out at about 200 MWp per
annum by AD2000. Most of the sales would be for consumer systems and professional
systems, with relatively little going to rural electrification, because of the high
capital costs involved. However in some countries, there would be good markets
among more wealthy private customers for powering isolated houses and for
consumer systems, particularly for the tourist and leisure markets. Systems
installed by national governments and public utilities would be mainly for
applications with high social value. Although the market would be relatively
limited, probably only a few megawatt per annum, the benefits to isolated
communities would be high.
Year
LOW PRICE SCENARIO
Modules Systems
Sales
MWplyr
wwp
$MP
HIGH PRICE SCENARIO
Sales
Modules
Systems
MWp/yr
$MJP
$MP
1984
6.5
12-22
25
6.5
12-22
25
1990
3.0
6-10
100
4.0
8-15
50
1995
2.0
3-7
700
3.0
6-10
100
2000
1.5
2.5-5
5000
3.0
5-9
200
Table 2.2 Projection
of PV Prices arzd Sales 1983-2000 (1985 US$)
Market development
Developing countries have always been considered as a very large potential market
but, due to financing problems, actual commercial sales in these countries are at
present very small. In fact, the greater part of the systems installed to date in
developing countries has been assisted by foreign governments and/or the
international aid agencies. Developing countries are rightly concerned to ensure
that at the right time photovoltaic technology is transferred to them, rather than
find themselves dependent yet again on an imported energy technology. In due
course, it is likely that most developing countries will have their own PV industry,
but this will take many years to establish, during which time there will be a need to
import systems for demonstration projects, professional applications and key
community applications.
A significant market in the short term may well be for solar refrigerators and
other consumer products for the wealthier sections of society, but in the medium
and long terms much larger markets can be expected to develop for professional
systems, particulary for telecommunications, village water supplies and generators
for police posts and health centres. If system costs can be brought down to about a
third of current levels which may well be possible within five to ten years, rural
electrification using photovoltaics will become a viable option in many situations,
with market potential reckoned in many hundreds of megawatts per annum.
Institutional
support
Established institutions already exist in the industrialized countries for coordinating and funding energy research, development and demonstration
activities, and for disseminating information on the performance, costs and
benefits of new technologies. In view of the importance of photovoltaics for
applications now and in the future, it is clear that continued institutional s~~pport
will be needed for this technology. The photovoltaic industry in each country has to
compete in world markets and for this there needs to be a sustained programme of
technical support, not only for research and development but also for the
formulation of appropriate norms and standards and dissemination of information.
The recent instability in the price of oil underlines the need for consistent long-term
support for new and renewable energy sources such as solar photovoltaics.
Until such time as photovoltaic systems become cheap enough to be
economically viable for grid-connect.ed applications, the markets in industrialized
countries wti remain relatively limited. Apart from the cost factor, there are no
major technical or institutional barriers to be overcome, although there is still a
need for information on photovoltaics to be disseminated to potential customers
and to government ofi?cials, many of whom remain unaware of the opportunities
offered by this relatively new and formerly rather exotic technology. Some
electricity utilities have adopted a rather negative attitude to photovoltaics in the
past; they now need to recognize that photovoltaic systems can complement more
traditional systems and, in particular, can help solve electricity supply problems in
remote areas.
In the developing world, the opportunities for PV are much greater than in the
developed world but so are the obstacles to be overcome. A major need in many
countries is for an effective institutional base that can monitor, plan and regulate
developments in a technology as new and as promising as photovoltaics. The
experience available within the industrialized countries could do much to assist
developing countries build up the necessary institutional base. Private and public
interest need to be brought together to ensure that the following aspects are
properly covered:
(a) Technical development
-identification
of projects that can be properly justified with reference to all
relevant factors
-applied research to develop systems appropriate for local needs
-commercial manufacture and assembly of components and systems
16
- integration of photovoltaics into related fields (eg. building construction and
electronics)
- field installations for demonstration purposes, with monitoring and evaluation
(b) Regulatory aspects
.L development and implementation of appropriate norms and standards
-regulations for consumer protection and warranties
- independent testing and certification of components and systems
-planning of integrated projects with no disturbing side effects
(c) Incentives and finance
-tax credits, subsidies and grants
-public and private finance for manufacturers and customers
- co-ordination of aid-funded projects
(d) Training and information
-training of professionals in aJ.laspects of photovoltaics
-public information and advisory services
-liaison with other energy supply agencies.
17
3. WATER PUMPING
3.1 Introduction
Pumping techniques
Hand- and wind-powered pumps have a long history for lifting water in rural areas
for water supply and irrigation. Improved designs of hand pumps that are more
efficient and durable and easier to maintain are now being widely introduced.
There has also been renewed interest in wind pumps in recent years, with the
emphasis on lower cost designs suitable for local manufacture. Petrol and diesel
pumps are also being used in some areas, particularly for low-lift irrigation.
In rural areas that have been electrified by grid extension, electric pumps are
usually a reliable and relatively low cost option. However, for most rural areas, it
will be many years before this alternative is available.
Substantial efforts have been made in recent years to develop reliable and
cost-effective solar-powered pumping systems. A number of prototype solar
thermal systems have been developed, but none so far offer sufficient reliability,
ease of operation and maintenance and cost-effectiveness. Photovoltaic systems
on the other hand offer a number of attractive features and, after several years of
development, are now readily available in various standard configurations, as
shown in Figure 3.1.
A solar photovoltaic (PV) water pumping system consists of the following
main components: the PV array, with support structure, wiring and electrical
controls; the electric motor; the pump; and the delivery system, including pipework
and storage. These components have to be designed to operate together to
maximize the overall efficiency of the system (or, rather, to optimize the costeffectiveness of the system). An electrical controller is sometimes incorporated to
improve the electrical performance of the system. Energy storage in the form of
batteries is rarely used, as it is generally cheaper and simpler to store the water to
cover periods of low solar input or high demand.
The advantages and disadvantages of the various pumping techniques are
compared in Table 3.1.The main problem with PV pumps has been their high initial
cost, but with cheaper PV modules coming onto the market and with improved
system designs incorporating volume-produced pumpsets this does not constitute
such a barrier.
Water supply
An example of a solar photovoltaic borehole pump used for village water supply is
shown in Figure 3.2.Water supply requirements do not vary much month by month.
It is important to provide sufficient storage to cover periods of cloudy weather,
when the output from the PV pump will be low. A covered tank at or near ground
level, connected to a number of automatic shut-off supply taps on a.concrete or
stone pad would be a typical arrangement.
Whenever possible, it is safer to take water Lztended for human consumption
from closed boreholes or protected wells. If a surface water source, such as a lake
or a stream, has to be used, it is usually possible to construct some form of filter
19
I
ltlr
rump
Shaft
~IIIVC
I
Cablehm
PVArrayf
I ;
I k--?
Type
bl
Submerged
PumpCylinder
Portable
PVArray\
Electric
Type
(d)
Figure 3.1 Standard
[email protected]
20
of PV Pumps
Type
M
PVArmy
Pumping technique
Main advantage
Main disadvantages
Hand pumps
Low cost
Simple technology
Easy maintenance
Diesel and
gasoline pumps
Low capital cost
Can be portable
Extensive experience
Easy to install
Easy to use
Wind pumps
Moderate capital cost
Suitable for local
manufacture
Easy to maintain
Non-polluting
Need no fuel
Long life
Extensive experience
Low maintenance
Non-polluting
Need no fuel
Reliable
Long life
System is modular
Low flow
Absorbs time and
energy that could be
used more productively elsewhere.
Often involves
uneconomic
use of
expensive boreholes
Maintenance
often
inadequate,
reducing life.
Fuel often expensive
and supply
unreliable.
Noise, dirt and fume
problems. Unreliable
if not maintained
Very sensitive to
wind speed, with
periods of low
output.
Needs open terrain
Not easy to install
Solar PV pumps
Table 3.1 Comparison
of Pumping
High initial cost
Low output in cloudy
weather
Techniques
when building the pump sump. Complete solar-powered water treatment plants are
now available, as discussed in Chapter 7 of this survey.
Water intended for livestock is usually pumped from a borehole and stored in a
raised tank so that the cattle drinking troughs may be gravity fed through ball
valves. The PV array needs to be well-protected to prevent damage by hvestock.
Irrigation
PV pumps are well suited to irrigation applications. They produce the most water
when the solar radiatjlun is greatest and hence when the crop water demand is
highest. Because PV pumps deliver water over a period of about 10 hours each day,
it is important to plan carefully the distribution of the water to avoid losses by
evaporation and infiltration. The irrigation technique will need to be adapted to
take best advantage of ,the available water. For example, instead of one large pump,
it may be better to deploy several small pumps at different places in a large field, or
in several separate fielda. Alternatively, it may be feasible to store water for
discharge at a higher flow rate over a shorter time.
21
/
PRWWConditioning
Level
Figure 3.2 Tgpical PV Water Pumping
Sgstem
The pumping of relatively large volumes of water for flood irrigation for rice is
unlikely to be cost-effective, whereas a fruit farmer may well find that the relatively
small volumes required for a trickle system could be supplied very economically.
The underlying principle is that the cost of water used must be less than the value of
the extra crop gained through the irrigation.
A typical PV irrigation pumping system is shown in Figure 3.3.
Field experience
A substantial volume of field experience is now available relating to solar pumps.
Approximately 2000 units made up of the different configurations shown in Figure
3.1 have been supplied worldwide. A comprehensive study of solar pumps,
involving field and laboratory testing of component and complete systems, was
completed in 1983by consultants for the UNDP and World Bank (Ref. 3.1). A survey
of solar pumping field performance was recently been carried out by consultants
for the World Bank (Ref. 3.2). There is also nearly ten years of experience available
with Mali Aqua Viva, who have sponsored some 60 PV pumps, mainly for village
water supply (Ref. 3.3).
3.2 Principal
case study -
Mali
Over the past ten years, more than 80 PV pumps have been installed in Mali, mainly
under the auspices of the charitable organization Mali Aqua Viva (MAV) with
financial support from various aid agencies. Other systems have been installed by
various organizations, in particular four systems for which good data exist installed
by LESO (Laboratoire de 1’Energie Solaire), Bamako.
Information on the PV pumping experience in Mali has been obtained from
personal knowledge supplemented by published data (References 3.3, 3.4, 3.5).
22
Water Level
FkutingMator/PmpSet
Figure 3.3 Tgpical PV Water Pumping
Sgstem for Irrigation
Most of the systems installed have been for village water supplies and are wellappreciated by the users. Arrangements for technical support have been
established, so that. on-going advice can be given and faults corrected. The MAV
approach to all their water supply improvement projects is to involve the local
people from the beginning, to ensure their full understanding and commitment. In
the case of solar pumps, the villagers are expected to build as much of the local
infrastructure as possible (eg. storage tanks, access, foundations) and this means
that a significant proportion of the total capital cost (up to 25 per cent) is met from
local sources. The motivation generated by this initial involvement has proved to be
a key factor in the successful implementation of most of the MAV projects.
Technical mpects
There are many practical difficulties in measuring accurately the performance of
solar pumps in the field. The first attempt to do this in a systematic way was part of
the programme of work included in the UNDP/World Bank solar pumping project,
carried out between 1979 and 1983. Four PV pumps were installed and tested in
Mali as part of this project and the results are presented in Reference 3.6.
More recently, LESO have been systematically monitoring the performance of
a number of PV pumps and keeping detailed records of faults and other incidents
that affect the system reliability. Staff from IXSO have also recently visited over 30
PV pump sites in Mali, making notes on the technical condition of the system, the
financial arrangements, operation and maintenance, and the social and
institutional aspects (Ref. 3.4).
The experience in Mali is mainly with low and medium head systems
incorporating various types of centrifugal pump. There is at least one surfacemounted positive displacement pump, but no submerged reciprocating pump (jack
23
pumps, Type C in Figure 3.1). There have been few problems with the PV arrays,
although discolouration of cells and corrosion of module frames has been
observed. Most technical problems have been associated with the pumps, motors
and control systems. In some cases, the wel.l itself was not vertical or not deep
enough, which lead to major pump problems. In other cases the yield of the well
was insufficient to match the pump capacity, leading to the pump running dry with
consequent failure.
Originally, the majority of the PV pumping systems installed in Mali had
submerged pumps and surface mounted DC motors (Type B in F’igure 3.1). More
recently the systems have been of the AC submersed pumpset type (Type A in
Figure 3.1), a change which has increased reliability and eliminated a major
maintenance cost, namely shaft repair. In the multi-stage vertical turbine pumps,
vibration in the connecting drive shaft has caused at least one broken shaft and
many bearing and seal failures. Directly-coupled motor/pump units, either down
the borehole for medium and high head applications or floating units for low head
are preferred.
There have been a few problems with motors, principally due to overheating
caused by overloading. The main electrical failures have been with the electronic
control systems associated with some systems.
The availability (proportion of time operation) of the systems installed prior to
1984 has been generahy found to be between 70 to 85 per cent, due largely to the
delays involved in getting faults repaired. The availability of the more recent
systems is expected to remain much higher, at around 96 per cent.
Where it has been possible to measure the actual output of water in relation to
solar energy input, it has often been found that manufacturers’ claims have not
been met. However, this result is not uncommon for many commercial systems,
including conventional diesel-powered pumps. It underlines the need for clear
specifications at the tender stage and for independently-witnessed performance
tests before shipment to the site.
The independently measured performance of four solar pumps in Mali is
shown in Figure 3.4 (from Reference 3.7).
Economic aspects
The organizations installing PV pumps in Mali have found that capital costs have
been steadiIy falhng over recent years. A 1300 Wp system purchased in 1979 cost
$35/Wp, but now a similar system would cost about $1l/Wp. To this must be added a
further $6-$lOnVp for shipping and installation.
Most of the PV pumps in Mali are administered by a village co-operative, which
in addition to organizing the local support for the initial installation, also arranges
for water charges to be levied on users. Some well-organized cooperatives are able
to raise enough money in this way to pay for on-going costs and for further
development, such as a second pump. In other places, the organization is not so
efficient or the circumstances may not be favourable, leading to insufficient local
funds to pay for maintenance and any further development.
A recent study of solar system economics in Mali found that PV pumps could
be competitive with diesel pumps especially for low lift applications (Ref. 3.5). An
example of a typical economic analysis for a 1400 Wp low lift system is shown in
Table 3.2. Based on a 16 year period and 10 per cent discount rate, the annual
levelized cost comes to $3905. If 100 per cent of the water output is useful, the
average unit cost is $0.07/m3.Even if it were only practicable to charge for 25 per
cent of the water produced by the system, at say the equivalent in local currency of
$0.25/m3,the system would still be able to compete with a diesel pump.
24
! I;
Ii
ou
‘1
Size:
5816 W
Suia
irraduhua
(MJlm’lJay)
Sulrr irrulirtion
(MJlrp’IJtiy)
solar
irradiation
(MJlm’lda~)
SOIU irrdiution
(MJlm’l3uy)
Figure 3.4 Performance
of Four Solar Pumps in Mali
Social and institutional
aspects
The PV pumping systems have generally been well-received in Mali, particularly at
the sites where there haz been a good history of local involvement with the
installation from the planning stage. Demand is usually greater than supply and this
is the main reason for user dissatisfaction. There have also been problems with
pipework, storage tanks and water distribution arrangements, which some users
have perceived as being failures of the PV system.
Because of organizations such as MAV and LESO, maintenance and technical
support for PV pumps is probably better in Mali than anywhere else in the
developing world. Nevertheless, there have been problems in repairing faults, due
__5
25--
___._
.____._
I*-.___C_-i_*--...--------_----_
---
to poor communications. In many cases, the overseas supplier of a system has not
responded promptly to request to repair faults or to supply spare parts, which is
disappointing. Efforts are now in hand at LESO to tram engineers and technicians
in PV system technology and to carry out trouble-shooting, repairs and routine
maintenance.
The level of technical support required may be judged from the cost of
maintenance. According to figures given by MAV, it has cost on average about $540
per pump per year to maintain some 30 PV pumping systems. With the experience
that has been gained and the introduction of improved system configurations, the
avtrage cost of maintenance is expected in future to come down to around $330 per
pump per year. LESO has estimated the cost of maintaining the submerged AC
motor/pump systems (Type A in F’igure 3.1) to be $250 per pump per year.
BASIC DATA
1400 wp
160 mVday at 5m head
361 W
340 days/year
54 400 m3
System size
Capacity
Hydraulic power
Use (assumed availability)
Annual output
COST ANALYSIS
15 years
10%
Period of analysis
Discount rate
Capital costs:
- equipment CIF
- installation
- total
Replacement costs, 2 pumps
Total Present Worth of system
$16 052
$ 8 250
$24 302
f 3500
$27 802
Recurrent costs:
- maintenance
- Present Worth
$ 250 per year
$ 1902
Total Present Worth of Life
Cycle Costs
Annual Levelized Cost
Units Water Cost:
- 100% use
50% use
25% use
Table 3.2 Typical Cost Analysis
-....
$0.07 per m3
$0.12 per m3
$0.29 per m3
of PV Pumping Sgstem in Mali
.r
26
3.3 Other water pumping projects
UNDP/World Bank project
Reference has already been made to the comprehensive evaluation of solar
pumping technology and economics carried out by consultants for the UNDP/
World Bank over the period 1979-33. The project also resulted in the production
of a solar pumping handbook covering the technology and economics and giving
recommendations on procurement of appropriate equipment (Ref. 3.8).
In addition to providing much practical advice on all technical aspects of PV
pumping systems, the study also presented general guidance on the economic
prospects. Based on current costs for installed systems in the range $1523/Wp, it
was found that for irrigation applications, solar pumps could be cost competitive
with diesel pumps for small-scale, low-lift applications (eg. 30m3/day through 5m
head) in places where the average solar input was at least 15 MJ/mz per day (4.2
kWh/mz per day). For water supply applications, solar pumps could be competitive
with diesel pumps for relatively low-flow, medium head applications (eg. 25m3/day
through 10m head).
India
India has been manufacturing PV cells and systems using practically 100 per cent
local components since 1979. A number of solar pumps have been installed at
various villages in India for water supply and irrigation applications. In many cases,
the technical aspects have been completely overshadowed by the social and
institutional problems encountered.
In one case, a PV pumping system was chosen because of a constant history of
breakdowns and fuel supply problems for an existing diesel pump (Ref. 3.9). The
water depth was 15m, too deep for animal-powered devices. Construction of the
well and installation of the pump was subjected to many delays and a solar
management committee came into being as the most acceptable means of ensuring
the equitable distribution of the water produced by the pump.
Botswana
The results of a study are reported in References 3.10 and 3.11 on the economics of
PV pumps in comparison with diesel pumps for rural water supplies in Botswana..
The studies compared a 6 kW diesel engine powering a progressing cavity pump
with a PV-powered electric motor driving the same pump. The study found that a
significant technical consideration was the borehole yield characteristics. In a
typical application, a maximum flow of 2 to 6 m3/hour will occur with the PV system
around noon on clear days. Yield tests indicated that this would exceed the rate of
water inflow into the well, resulting in the well level falling too low.
Another relevant issue concerns the use of existing pumps and wells to
minimize the cost of establishing a PV system. This requires good communication
between the field personnel and the equipment suppliers to ensure a properly
designed system. Economic analysis showed that the PV option would be costcompetitive with the diesel option at current prices.
Two PV pumps have been operating since 1981 as part of the Desert Development
Demonstration and Training Project at Sadat City (Ref. 3.12). A 10 kWp array
coupled to an inverter supplies AC power at 220 V, 50 Hz to the headquarters
buildings and to an AC submersible pump set in a deep tubewell (43m head). A
d7
separate 3 kW system provides DC power to a surface-mounted motor driving a
submerged progressing cavity pump. These is also a surface-mounted booster
pump for irrigation.
The PV arrays have operated reliably with average daily conversion efficiency
reported as 7.22 per cent. The AC submersible pump has also performed well. The
progressing cavity pump has run reliably since mid-1934. Prior to that, excessive
vibration of the drive shaft reduced the performance of the pump and resulted in a
number of failures. The problem was solved by installing additional guide bearings
to the shaft, since when the pump has operated reliably. The AC system includes
batteries which need regular maintenance.
3.4 Conclusions
Technical aspects
PV pumping technology has improved significantly over recent years, with the
emphasis on better matching of system components, increased reliability and
reduced maintenance requirements. The type and size of system needs to be
chosen carefully on the basis of a systems approach to the problem, taking into
account all relevant factors, including operation and maintenance implications.
(See Reference 3.8 for design and procurement recommendations).
PV arrays based on the well-proven crystalline silicon cell technology have
generally proved to be the most reliable component of a pumping system. With the
advent of lower-cost thin film cell technologies, it will be necessary to monitor
array performance carefully to ensure adequate provision is made for long-term
degradation.
The introduction of brushless DC motors by some manufacturers for surface
mounted or floating pumps has eliminated the need for brush replacement. For
submerged motors, water-filled AC induction motors are proving to be much more
reliable than sealed DC motors. This type of borehole system is now preferred to
turbine pumps with surface motors and long vertical drive shafts. The variable
frequency DC-to-AC inverters required for AC systems provide a low-zest means of
matching the PV array output to the motor load and they appear to perform reliabily
in the field.
Centrifugal pumps can be well-matched to PV arrays. Problems have been
reported with surface-mounted centrifugal pumps, due to the need to maintain
prime. A self-priming tank on the suction side has proved to be more reliable than a
foot valve. Centrifugal pumps should not be used for suction lifts more than 5 to 6 m
and wherever possible a floating or fully submersed unit is to be preferred.
Positive displacement pumps have a water output that is practically
independent of head and directly proportional to speed. For PV-powered systems,
problems have been experienced due to the cyclical nature of the load on the motor
and +he high frictional forces, particularly at start up. At high heads this type of
pump can be more efficient than a centrifugal pump, since the frictional forces are
relatively small compared with the hydrostatic forces. Positive displacement
pumps are however usually very rugged and reliable, provided the overall system
has been well-designed in the first place to suit the conditions obtaining at the site.
A common cause for pump and/or motor failme has been overloading due to
sediments in the water or tight shaft bearings. Dry running due to loss of prime
(surface pumps) or falling water level in the well is another common cause of
failure. Increasingly manufacturers are providing low water level and/or high
temperature protection for the motor.
The use of tracking PV arrays, maximum power point trackers and batteries
28
may offer advantages in theory, but experience has shown that, at remote sites
where maintenance is difficult to provide, the extra complexity introduced is
counter-productive.
Pump performance is heavily dependent on the assumptions made at the
design stage regarding solar and water resource characteristics. Careful account
has to be taken of the variations in solar input to the array, the static water level in
the well and the water demand. Failure to do this has resulted in many systems
being undersized so that they fail to meet the demand, or excessively oversized,
with associated additional capital cost. However, these problems should diminish
as experience builds up and a larger data base applicable to each country becomes
available to system designers and suppliers.
Economic aspects
The F.O.B. prices of PV pumping systems have been steadily falling from about $30/
Wp in 1978 to as low as $lO/Wp in 1986.To this has to be added the cost of shipping
and installation. The unit water costs expressed per volume-head products ($/m>
may be calculated on a life cycle cost analysis for different assumptions regarding
demand, solar insolation and fuel costs. This has been done in Reference 3.13 the
results of which we presented in Figure 3.5. The data assumed in the analysis is
given in Table 3.3. It can be seen that solar pumps are typically competitive up to
1000 m4/day demand (eg. 40 m3/day pumped 25m. This approximates to 1400 Wp
array power. Reference 3.2 also concluded that PV pumps are competitive up to
approximately 1000Wp (and for regions of very high diesel operating cost up to 2.5
kWP).
Water for irrigation is characterized by a large variation in demand from month
to month. Hence a solar pump sized to meet the peak demand is under-utilized in
other months. This adversely affects the economics/unit water costs and it is for
this reason that solar pumps are more competitive for rural water supply.
Social and institutional
aspects
PV pumping systems, being a new technology, need continuing institutional
support to enable them to be successfully integrated into the rural communities
that stand to benefit.
There are three main areas where institutional support is particularly needed:
-at the planning and procurement stage
-for administering the operation
-maintenance and spare parts.
At the planning stage, it is important to involve the local community from the
outset and encourage them to organize a management committee. The local costs
should be raised locally, either in cash or in direct labour. Clearly experienced
technical advice will be needed for the design and procurement of suitable
equipment.
The local organization must then be assisted to organize appropriate
arrangements for distributing the water and levying charges. It is helpful to appoint
a keeper or operator to watch over the system and he will need to be given basic
training in routine maintenance and simple trouble-shooting.
With good design and the installation of the latest types of system, system
reliability should be good. However, there will inevitably be faults arising from time
to time which cannot be fixed by the users. Established arrangements need to exist
for calling in technical support and for the procurement of any necessary spare
parts. Good communications between the site and the source of support are of
course very desirable, but it has to be recognized that in many rural areas the
necessary infrastructure simply is not available.
29
[email protected] UNIT CUXS @FUEL(a6T = $0.25/liter
e RlEt axn = Sl.SO/liter
0 t
10
I
20
I
30
1
I
I
40
50
80
WATER DIMAND (xWda.v}
HEAD = 25 METERS
I
1
I
I
70
80
00
100
Figure 3.5 Solar Pump and Diesel Pump Cost Comparison
PV PUMP SYSTEM CHARACTERISTICS
Depth of Water Supply, Head
Annual Average Daily Water Demand
Annual Max Daily Water Demand
Insolation
PV Array Peak Power
PV Pumping System Capital Cost
PV Pumpin
System Availability
PV Array Li7e
Pump Life
Normal Discount Rate
Inflation Rate
PV NPV Unit Costs
25m
20 m3/day
30 mYday
5.0 kWh/m*/day
1.11 kWp
$‘fg$b
2010 yrs
5.0 yrs
10.0 %
5.0 %
$0.14/m3
DIESEL PUMP CHARACTERIS’TlCS
Depth of Water Sup ly, Head
Annual Average Dai Py Water Demand
Annual Max Daily Water Demand
Diesel Generator Power Rating
Average Load Factor
Diesel Fuel Cost
Diesel Gen-Set Pump Capital Cost
Diesel Pump Availability
Diesel Gen-Set Life
Pump Life
Nominal Discount Rate
Inflation Rate
Diesel NPV Unit Costs
25m
20 m3/day
30 m3/day
3.0 kW
45.4 %
$0.75/litre
?g;y
s:o yors
5.0 yrs
10.0%
5.0%
$0.26 /m3
Table 8.3 Data Used for Solar Pumping Cost Comparison
30
Chapter 3 - References
3.1 ‘Small-scale solar-powered pumping systems: the technology, its economics and
advancement’. Main report prepared by Sir William Halcrow and Partners in association with I.T.
Power Ltd, for UNDP Project GLO/80/963 executed by the World Bank, June 1983.
3.2 IT Power, ‘Solar-Powered Pumping Systems: Their Performance, and Economics’. Report to
World Bank (1986).
3.3 Mali Aqua Viva Report No. 7, Activities from October 1979 to October 1981.
3.4 Private communication Tom Laboratoire de Energie Solaire, Bamako, Mali, Sept 1986.
3.6 T.R. Miles Jr, ‘Economic Evaluation of Renewable Energy Techologies at LESO’ for US-AID and
Energy/Development International, Dee 1986.
3.6 ‘Small-scale solar-powered irrigation pumping systems - Phase 1 Project Report’ by Sir
William Halcrow and Partners in association with Intermediate Technology Development Group
Ltd, for UNDP/World Bank, July 1981.
3.7 ‘Small-scale solar-powered irrigation pumping systems - Technical and Economic Review’
by Sir William Halcrow and Partners in association with Intermediate Technology Development
Group Ltd, for UNDPMTorld Bank, Sept 1981.
3.8 J.P. Kenna and W.B.Gillet, Solar WaterPun~ping-A HandbookIT Publications, London, 1986.
1986.
3.9 P. Amado and D. Blamont, ‘Implementation of a solar pump in a remote village in India;
economical and socioeconomic consequences - three years of working experience’, i%oc. of
Third lntemational Conference on Energy for Rural and Island Communities, Inverness, UK, Sept
1983.
3.10 R. McGown and A. Burrill, ‘Current Developments in Photovoltaic Irrigation in the Developing
World’, A.R.D. Inc., 1986.
3.11 D. R. Darley, ‘PV vs Diesel: A Grounded Economic Study of Water Pumping Options from
Botswana’, Massachussetts Institute of Technology, USA, 1984.
3.12 IT Power Inc, Proceedings of the Photovoltaics Information, Symposium and Workshops,
Nairobi and Chiang Mai, April 1986. (Sponsored by the UNDP Energy Office).
31
4. PHOTOVOLTAIC
REFRIGERATORS
HEALTH CARE
FOR RURAL
4.1 Introduction
The need for PV refkigerators
In many developing countries, living conditions for the majority of the rural
population are poor and there is widespread malnutrition combined with a high
incidence of disease.Infant mortality is particularly high in the rural areas,where in
some countries, as many as one third of the children die before the age of two. Much
of the disease could be eliminated or controlled through mass immunization
the practical problems involved are formidable. Most countries are however
making large efforts to improve the quality of rural health care, including expansion
of their immunization programmes.
Vaccines require refrigeration during transportation and storage to remain
effective. It is important to maintain the vaccine ‘cold chain’ from the place of
manufacture right through to the point of use. This imposes a major logistical
problem because generally there are no reliable electricity supplies to operate
conventional electric refrigerators in the rural areas where the clinics and health
centres are located. Kerosene and bottled gas (LPG) powered refrigerators are
available but their performance in many cases is not adequate and there are often
problems in ensuring regular fuel supplies.
Solar photovoltaic refrigerators have the potential for better performance,
lower running costs, greater reliability and longer working life than kerosene or
LPG refrigerators, or diesel generators powering electric refrigerators.
Recognizing this potential, the World Health Organisation (WHO), the Centre for
Disease Control (CDC), the US Agency for International Development (US-AID),
the European Community (EC) and other agencies have installed and evaluated
many PV refrigerators throughout the developing world. At least 800 PV medical
refrigerators have been installed over the last five years, mainly for testing and/or
demonstration purposes. The stage has now been reached where a number of
designshave been approved by the WHO, opening the way to wider implementation
of this technology.
The use of PV refrigerators instead of kerosene or LPG units offers the
following benefits;
i) Elimination of fuel supply costs and delivery problems
ii) Reduced vaccine losses through improved refrigerator reliability, with
associated reduced anxiety among medical personnel
iii) Reduced maintenance workload for technicians and medical personnel, with
associated cost and time savings
iv) Overall cost savings for the vaccine cold chain equipment.
v) A more effective and sustainable immunization programme, leading to reduced
incidence of disease.
The technology
Five alternative methods of solar powered refrigeration were surveyed by the WHO
during 1980 (Ref. 4.1). These were: photovoltaic/vapour compression,
33
photovoltaic./Peltier effect, solid absorption/zeolite, solid absorption/calcium
chloride, and liquid absorption/ammonia. Photovoltaic systems are the only type
commercially available for vaccine storage (with the exception of one solid
absorption ice-making plant manufactured in Denmark). Of the photovoltaic
systems, several manufacturers offer vapour compression systems in suitable
forms for use in the vaccine cold chains.
A schematic diagram of a photovoltaic/vapour compression refrigerator is
given in Figure 4.1. A PV array charges a battery via a charge regulator, to ensure
that the battery is not overcharged. The battery powers a DC motor which is
coupled directly to the compressor. The motor/compressor is usually
manufactured as a hermetically sealed unit. The motor is of the electronically
commutated brushless type. A second regulator is employed to ensure that the
motor/compressor operates only within its rated power range and to prevent overdischarge of the battery. Freon refrigerant is used in the cooling cycle in the normal
way, ie. the cooling effect is achieved by the heat absorbed by the refrigerant as it
evaporates in the evaporator. A thermostat switches the motor/compressor unit on
and off as required. Some models have two compressors and thermostats, one each
for the refrigerator compartment and the freezer compartment.
The insulation is normally of the expanded polyurethane type and double the
usual thickness to reduce heat gain and thereby reduce the energy consumption
and increase the time the refrigerators can maintain safe temperatures with no
power. Most units are top opening, to reduce loss of cold air and often have a
secondary hinged or removable cover under the main door. Figure 4.2 shows a
typical example of a PV vaccine refrigerator.
Condcnsor
Evoporotor
RefrigemtorSub-System
Photovoltsic Array
,
Chorging //
RegulotorJ
I
\
Thermostat1
flattery /
Compressor
Regulator
Motor
Figure 4.1 Schematic 4f Photovoltaic
ReFigerator
System
WHO specification
The vaccine capacities of solar refrigerators available or being developed vary
widely, from 3.6 to 200 litres. The need for solar refrigerators is greatest at the
peripheral health centres serving populations of 20,000 to 100,000.The quantity of
packed vaccine needed to immunize fully 150 infants and their mothers is
approximately 4 litres. There is however no general agreement yet on the best size
34
Figure 4.2 Example of a Photovoltaic
Vaccine Refrigerator
for a PV vaccine refrigerator. Opinions differ on the quantity and volume of other
biological products (eg. blood) which might be stored in the health centre
refrigerator and many people believe that a larger cabinet will have a wider market.
It is also important for the system to have the capacity to freeze ice-packs which are
used when transporting vaccine from the health centre for immunization in the
field. The ice production capacity is a significant load on the system and has a major
influence on the PV array size and hence system cost,
In 1981, the WHO issued an outline specification for PV refrigerators which
laid down minimum requirements covering vaccine capacity, ice-making
performance, refrigerator performance, hold-over time, battery maintena: ?
interval, etc. The basic requirements have been modified in the light of field
experience and the current WHO specification (Ref. 4.2) provides as follows:
35
1. The design of the system will be sized to enable continuous operation of the
refrigerator and freezer (loaded and including icepack freezing) during the lowest
periods of insolation in the year. If other loads, such as lighting, are included in the
system, they should operate from a separate battery set, not from the battery set
which supplies the refrigerator.
2. The design of the system will permit a minimum of five days continuous
operation when the battery set is fully charged and the photovolatic array is
disconnected. During this time the internal temperature of the refrigerator will
remain within the range of +0 to +8”C when the constant external temperature is
a minimum of +32”C.
3. Refrigerator/Freezer: In continuous ambient temperatures of 2O”C, 32°C and
43”C, the internal temperature of the refrigerator, when stabilized and fully loaded
with empty vaccine vials, wilI not exceed the range +O”C to +8”C. This range will.
be maintained when, in an ambient temperature of +22”C, the maximum
recommended load of icepacks containing water at +22”C is placed in the freezer
and frozen solid without adjustment of the thermostat. The recommended load of
icepacks will freeze in less than 24 hours and will weigh at least 2 kg without the
material of the pack.
4. Photovoltaic Array: Modules will meet the latest applicable specifications laid
down by the Jet Propulsion Laboratory (USA) and the Joint Research Centre, Ispra
(Italy). Array structures will be designed to withstand wind loads of +200 kg/m2
and wilI be provided with fixings for either ground or roof mounting. Appropriate
photovoltaic-type sealed connectors incorporating proper strain relief will be
provided for the array cable. Lightning protection devices will be provided.
6. Battery Set: The battery set will be sealed or low waterloss or non-liquid
electrolyte deep discharge type (minimum 1000 cycles to 50 per cent discharge).
Automotive batteries are specifically unacceptable for this application. The
batteries will be housed within the refrigerator freezer cabinet, or in a cabinet
separate from the refrigerator, but in either case lockable. No dry cell batteries
shall be used to power instruments and controls.
6. Voltage Regulator: A voltage regulator will be provided, which meets the charge/
temperature requirements of the selected battery and which cuts off the load when
the battery has reached a state of charge which can be repeated to a minimum of
1000 cycles. Lightning protection will be provided. The load should be
automatically reconnected when the system voltage recovers.
7. Instrumentation: A LED alarm will be installed to warn that power to the
compressor has been cut by the regulator. An expanded scale voltmeter or a LED
alarm will be installed to warn the user when the battery charge is in an unusually
low state of charge to give adequate advance warning. The warning light of the
minimum voltage limit should be clearly labelled ‘DO NOT FREEZE ICEPACKS’ in
the appropriate local language. If an external reading thermometer is provided for
the refrigerator, it should be marked clearly in green between 0°C and +8”C.
A thermostat or a defrost switch should be provided but no other power
switches should be installed. Circuit breakers of cartridge fuse holders will be
fitted with a polythene bag holding 10 spare fuses and special attention will be
given to corrosion of fuse mountings.
36
8. Individual sea-crating of the components of each system should be provided
whether or not containers are used to transport the systems. No package should be
heavier than can be handled by hand in the country. Labels bearing handling
instructions should be printed also in the appropriate local language.
9. Essential spare parts which may be needed during the first three years’
operation should be assembled as a kit in appropriate quantities for central and
regional storage in the country. A minimum list is as follows:
QuantityperlOsystems
2
2
2
1
1
3
3
Item
1. Photovoltaic modules
2. Regulator components (sets)
3. Battery sets
4. Array cables
5. Compressor, complete
6. Spare compressor regulator cards
7. Thermostat
10. Manuals will be provided for the installation and use of each system.
The WHO guidelines also require the supplier to provide a warranty for the
replacement of any component which fails due to defective design, materials or
workmanship. The minimum period of the warranty is required to be 10 years for
the PV array, five years for the batteries and two years for the remaining
components. The system supplier is also required to provide technical support for
maintenance and repair operations in the country concerned for a period of at least
two years. The supplier has to train an engineer to assist with installation of each
system and also train users and repair technicians in each area.
Commercially available equipment
At least 20 companies now supply PV refrigerators for vaccine storage. To assist
health authorities choose appropriate equipment, the WHO Expanded Programme
on Immunization (EPI) publishes Product Information Sheets (Ref. 4.3). Inclusion
of a product on these sheets in effect means that based on the information and test
results available with WHO-EPI, the product is approved for use in the vaccine ‘cold
chain’.
The PV refrigerators currently approved by the WHO-EPI for use in the vaccine
‘cold chain’ are listed in Table 4.1. It should be noted that there are many 12 V DCpowered refrigerators available, intended primarily for leisure applications in
boats and caravans. These have been designed for low capital cost, without
consideration of energy consumption or internal temperature variation. Although
such systems can readily be adapted for PV powering, they are not suitable for
vaccine storage.
Field experience
The most significant work on testing and evaluating PV refrigerators in the field has
been carried out under the WHO-EPI programme, involving field testing of over 50
systems for 12 suppliers in some 30 countries. Other work has been carried out by
UNDP, UNICEF, European Development Fund, AFME (France), GTZ
(W.Germany), ODA (UK), Oxfam and other aid agencies.
3’7
System Supplier
Refrigera:onr;/:reezer
.
Net Vaccine
Capacity (litres)
Refrigerator Freezer
/‘-EG-Telefunken (W Germany)
BP Solar (UK)
FNMA (Zaire)
Leroy Somer (France)
Polar Products (USA)
Solarex (USA)
Solarex (USA)
Solavolt (USA)
Solavolt (USA)
Table 4.1 PV Reigerators
Polar Products RR2
LEC EV5750
FNMA 75
Leroy Somer R50 + IF5
Polar Products RR2
Polar Products RR2
Marvel RTD4
Marvel RTD4
Polar Products RR2
Approved bg WHO-EPIfor
33
1:
16
33
33
10
2
Vaccine Storage
Not all these projects have been consistently monitored, but two major
projects are in progress which should provide substantial operating data in due
course. Both these projects are funded by the European Development F’und, one
involving the installation of 100 systems in Zaire and the other 20 systems in the
South Pacific. A summary of experiences with photovoltaic refrigerators for
medical use is given in Reference 4.4.
4.2 Principal case study - India
Background
Five PV refrigerators for vaccine storage and other medical uses in India have been
selected for detailed monitoring, at the following locations;
l Dappar Subsidiary Health Centre (SIIC), Pa&la District, Punjab.
a Adalaj Public Health Centre (PHC), Gandhinagar District, Gyjarat.
0 Tirupparankundram PHC, Madurai District, Tamil Nadu.
l Solar PHC, Bangalore District, Karnataka.
l Balesar PHC, Jodhpur District, Rajasthan.
The PV refrigerators at these five sites were installed in August 1984.All five
systems are similar and were supplied by AEG-Telefunken. They incorporate
refrigerators/freezer units supplied by Electrolux. The PV array consists of 12AEGTelefimken type PQ 10/20/Omodules giving a total rated power of 230 Wp at 12 V.
The regulator/control unit includes a shunt regulator to prevent overcharging and a
load-disconnect switch to protect against over-discharging of the battery. The
battery consists of two Varta 6 V units, with a total capacity of 150 Ah at 12 V. The
refrigerator is an Electrolux type RCW 42 DC unit. The daily energy demand was
estimated by the manufacturer to be between 670 and 860 Wh/day, depending on
the site and climatic conditions.
Assessment of design
System performance assessments for each of the five systems in India were made
for WHO. The assessments are based on the local climatic data (solar insolation
and temperatures for each month) and calculated cooling load data. A 20°C rise
above ambient was assumed for the normal operating cell temperature (NOCT)
38
and an array power temperature coefficient was taken as 0.005. A day/night
demand ratio of 0.67 and a battery turn-around (energy out to energy in) of 0.78
were assumed.
The analysis of the Tirupparankundram system indicated that there are two
months (November and December) when the system Loss of Energy Probability
(LOEP) or Loss of Load Probability (LOLP) is greater than 10 per cent. The system
could be upgraded to give a LOEP or LOLPof 1 per cent or less by increasing the
array size by 9 per cent (one module) and the battery size by a factor of 2.
The analysis for the Solar PHC installation indicates that there are four months
(July, August, November and December) when the system of LOEP/LOLP is greater
than 10 per cent. The Adalaj PHC site provides a severe test of the PV refrigerator
system because of the higher ambient temperatures experience there, which
results in lower array output. The system LOEP/LOLP is projected to be greater
than 10 per cent in four months (July, August, December and January).
The Balesar PHC system is projected by the supplier to have a higher demand
than the other sites. As a consequence, the LCEP/LOLP is projected to be in excess
of 10 per cent in four months (August, November, December and January). The
Dappar SHC installation appears to be the severest test of the system design
because of the low isolation in the winter months and the projected high demand.
Field experience
It is too soon for the results of field monitoring of the five systems to be available.
However, preliminary comments on the performance have been given to WHO.The
Tirupparankundrum PHC system was reported to be functioning well several
months after installation.
The main concern is with the batteries, which were considered to have too
small a capacity. Also the batteries supplied are not ‘low maintenance’, requiring
frequent topping-up by site personnel. Another concern was that the site personnel
could relatively easily alter the temperature setting of the thermostat, which could
result in freezing of the vaccines.
Bhoorbaral system
A PV refrigerator supplied by Solar Power Corporation (USA) was installed at the
end of 1981at Bhoorbaral near Delhi. The system has a 355 Wp array and a 630 Ah
battery and incorporates an Adler-Barbour refrigerator. The system is reported to
be performing well.
4.3 Other PV refkigeration
projects
NASA Lewis Research Center
The NASA Lewis Research Center (LeRC), US-AID and the Center for Disease
Control (CDC) signed an agreement to develop and test PV refrigerators for
vaccine storage in 1979.Refrigerator systems were first laboratory tested under
supervision by LeRC. Thirty-five systems are now being evaluated in 25 countries.
‘Ike&y-nine are stand-alone systems and 6 are refrigerator/freezer units instaIled
in larger PV systems in US-AID development projects. A summary of the field trial
programme is given in Table. 4.2.
From October 1981 to July 1984, the refrigerator/freezers in the NASA trials
accumulated almost 500 system-months of operation. They are reported to have
operated correctly, maintaining internal temperatures within the specified range
for a total of 84 per cent of the time for the SPC system and 83 per cent for the SVI
Polar Products and Marvel systems (Ref. 4.5). Although this is not an acceptable
39
level of availability for vaccine storage, it is considerably better than achieved by
many kerosene refrigerators.
A full record was kept of component failures during the trials. Most faulty
equipment was discovered during commissioning tests. The most damaging of
problems, and the most difficult to diagnose, concerned the power cable connector
of the SPC system, which became electrically disconnected whilst appearing to be
connected.
A number of systems experienced times when the internal temperature went
outside specified limits either as a result of component failure or from improper
use or incorrect system sizing. Specific reasons cited were:
- defective components (temperature controllers, thermostatically controlled air
doors, voltage regulators)
-incorrect setting of thermostat
- excessive amounts of warm material being placed in the refrigerator/freezer
-shadowing of the array as a result of poor siting or not clearing trees.
Sufficient performance data for system optimization studies could not be
collected because of the failure of many monitoring instruments. The two most
important instruments, the pyranometers for measuring solar energy input and the
ampere-hour meters for system energy consumption, were particularly failure
prone. Enough data was collected to indicate that in many casesthe percentage run
time and hence energy consumption was higher than anticipated.
Some user reactions to the systems installed by NASA-LeRC have been very
favourable. For example, the following comment is taken from the World Bank preinvestment study mission to The Gambia (Ref. 4.6):
The mission was told that the two solar refrigerators had performed quite
satisfactorily since their installation, except for one occasion when the unit at Kaur
had to be defrosted and disconnected for about three days. This was done on
instructions from NASA-LeRC in order to fully recharge the batteries, after the data
had indicated that the batteries were not staying at an adequate voltage. There are
two possible reasons for inadequate v&age: first, there ma.3 have been only a
minor fault either in the charging circuit or in one of the batteries, since all the
primary components appeared to be functioning correctly; second, there may have
been some abuse of the system, eg. through overloading the retigerator by the
operators with food and drinks in addition to the normal vaccines and ice-packs.
The former fault could have been corrected easily if the attendant were properly
trained, as is planned through the present project. The latter fault could have been
prevented by equipping the refrigerators with a padlock, as indeed the kerosene
refrigerators are currently equipped. The experience with solar refrigerators in The
Gambia can therefore be considered successful.
In Ecuador, the users were so impressed by the 100per cent reliability of their
systems that the officials now plan to make the health post with the PV refrigerator
a central storage facility for aJl vaccine in that area (Ref. 4.6).
World Health Brganisation
To prepare for field trials, the WHO organized laboratory tests on 12 types of
refrigerators suitable for PV powering at the Consumers Association in the UK
during 1981-83.The objectives of the tests were:
(i) To determine independently the most suitable refrigerator models
(if) To identify possible component or system failure modes prior to expending
resources on implementing field trials.
Of the 12 models tested, 10 passed the laboratory testing, a number after being
referred back to the manufacturer for modification.
40
SYSTEM SUPPLIER
(Refrigerator Model)
COUNTRY
POWER
SYSTEM
cn
z
E
4 9
kg r
3
6% I=
fat! G
AFRICA
Burkina Faso
Gambia
Ivory Coast
1
s
ylw
23
525
1
264
*
1
1
231
264
SC
264
200
221
200
160
630
630
630
630
630 4
630
420
525
420
315
9/82 AID
3183 AID
lo/82
AID
9182 AID
2/83 AID
9182 AID
l/84
AID
IO/82 CDC
l/84
AID
l/84
AID
160
330
240
240
*
420
525
315
420
630
6/84
1O/83
2184
2/84
I/83
330
297
264
200
630
630
630
420
IO/81 CDC
4/82 AID
5182 CDC
II/83
AID
b
1
1
1
1
1
1
1
1
*Part of larger photovoltaic
00
nz
264
1
1
ASIA
India
Indonesia
Maldives
Thailand
e
E
630
630
315
525
630
630
630
1
1
NEAR EAST
Jordan
Morocco
Tunisia
$ii3
z+
Ei
*
363
200
330
264
*
264
1
Haiti
Honduras
Peru
St. Vincents/
Grenadines
5
2/84 AID
I/83 CDC
2/83 CDC
2184 AID
5/83 AID
IO/84 AID
2184 AID
2/83 AID
2/83 AID
5/83 AID
9182 CDC
,
8182 AID
1
Guatemala
%uyana
ul
315
630
1
AMERICAS
Colombia
Dominican
Republic
Ecuador
P
y2
200
297
2
2
Kenya
Liberia
Mali
Zaire
Zimbabwe
AGENCY
1
2
1
AID
AID
AID
AID
AID
installations.
Table 4.2 WHO NASA-EeRC - CDC ‘- US-AID Field Trial Programme
41
Following the laboratory testing, the WHO has been co-ordinating the
implementation of the field trials programme carried out directly and by a number
of other organizations. The WHO-EPI office in Geneva, Switzerland, acts as a
clearing house for enquiries 2nd the collection of field trial data. It also arranges for
the analysis of the data z;ld publication of reports. Further laboratory testing of
refrigerators is continuing with Wil[O sponsorship at testing facilities at Cali,
Colombia.
The primary objective of zhe field trials is to evaluate the performance and
reliability of PV refrigerators when used in the EPI under widely varying climatic
conditions, The first installations in the WHO-EPI programme started in 1983 and
are being conducted in co-operation with other agencies and the manufacturers.
The field trials in progress are summarized in Table 4.3. The current status of the
programme is presented in Reference 4.7.
The principal interest is in the recent instahations of BP Solar/Let systems
undergoing trials in Kenya, Tanzania and Ghana, the AEG/Electrolux systems in
India and the Solar Force/Leroy Somer system in the Yemen Republic.
The WHO-EPI field trials have not been underway as long as the NASA-LeRC
trials and hence the data collected so far is very limited. Initial indications are that
some problems have been experienced. For example, systems instalki in the
Philippines were found to have an undersized PV array. The BP Solar/Let system
was reported to be successfully meeting expectations during visits in July and
November 1934 (Ref. 4.8). The BP Solar/Let systems are instrumented with
comprehensive data logging equipment”
+
Trust Territory of the Pacific Islands
In 1981Motorola (USA) was awarded a contract to supply 24 PV refrigerators (and
also lighting sytems) to dispensaries and schools in the Trust Territory of the
Pacific Islands. Motorola assigned the contract to Solavolt. The dispensary systems
consisted of a 200 Wp array, a battery and a WSR-12refrigerator. The installations
started in March 1982 when training was also given.
The systems were easy to install but during the initial training seminars,
defects with the WSR refrigerators emerged. Two units both had defective
thermostat cards and repeated problems with blown fuses were experienced. Ten
of the 24 systems were found to be out of action one year later, of which eight were
repairable.
The main problem with this project was the poor refrigerator reliability. The
WSR unit is no longer commercially available.
Zaire
The world’s largest programme to use PV refrigerators (100 systems) and also
lighting systems (760 installations) is currently underway in Zaire (Ref. 4.9).
Systems are being installed in clinics and dispensaries throughout the country. The
programme is the responsibility of the Departement de la Sante Publique and is
being financed by the European Community (European Development Fund). The
systems are supplied by Solarforce (France), working with a local company,
FNMA, as sub-contractor.
The FNMA PV refrigerator is manufactured in Zaire using imported Danfoss
compressors and Delco batteries. The refrigerator compartment has 66 litres
capacity and the freezer compartment has 20 litres capacity. The refrigerator has
been tested and has been demonstrated to be satisfactory. Such local manufacture
reduces foreign currency requirements, reduces supply lines and builds up local
42
SYSTEM SUPPLIER
(REFRIGERATOR MODEL)
COUNTRY
_ _
daa
KB
KY
aa
a!
AFRICA
Ghana
Kenya
2
Tanzania
NEAR EAST
Yemen Arab
Republic
Eit
WEST PACIFIC
Philippines
1
1
1
Vanuatu
Solomons
POWER
SYSTEM
:
1
160
252
7l84
230
150
7184
8184
296
400
660
11183
180
:z
288
420
840
!Ei
8/83
2l84
Table 4.3 WHO-EPI Field Trials
knowledge of the technology, which in turn improves operation, maintenance and
troubleshooting.
A complete system with 120 Wp PV array sells for the equivalent of about
$4000, inclusive of ail taxes and duties.
The users of the PV refrigerators and lighting systems have expressed very
favourable views. It has been estimated that in Zaire kerosene refrigerators in
practice work on average only 60 per cent of the time and at best 70 per cent in the
well-managed health centres. Although initially some problems with the PV
systems were reported, PV systems are generally considered by the users to be
more ieliable than kerosene refrigerators or diesel generators. The FNMA solar
refrigerator is shown in Figure 4.3.
SouthPacific
The European Development Fund is funding the supply of 23 PV refrigerators for
various sites in the South Pacific (10 in Tuvalu, IO in Papua New Guinea and 3 in the
Solomon Islands). The units are being supplied by a Belgian company. The project
is being executed by the South Pacific Bureau for Economic Co- operation (SPEC)
and installation started in late 1986. It is worth noting that these systems will
experience a wide range in climatic conditions. For example, the worst month solar
input is only about 2.0 kWh/m2per day in the highland areas of PNG compared with
over 4 kWh/m2 per day at the coastal sites (Ref.4.10).
French Polynesia
Since 1978, the French non-profit organization GIE-Soler has been implementing
the installation of more than 280 kWp of small PV systems in French Polynesia,
43
including reportedly some 300 refrigerators for medical and residential
applications. The project is supported by AFME and the European Commission. It
is reported that users are paying from 75 to 80 per cent of system costs and that the
project is stimulating further sales (Ref. 4.11).
suahu
The Sudan Renewable Energy Project (SREP) funded by US-AID and the Special
Energy Project funded by GTZ (W Germany) have both provided PV vaccine
refrigerators for Sudan. One of the f’lrst three units supplied by GTZ failed prior to
installation in the field. Laboratory tests conducted by GTZ showed that the
vaccine compartment could fall below zero under certain conditions. A further four
units (AEGiElectrolux) have now been installed in northern Sudan.
Figure 4.3 The FNlMA Solar [email protected]
[email protected]&med
in Zaire
A PV refrigerator supplied by US-AID under the SREP programme is
reported to be providing good service at a refugee camp.
Mali
An evaluation has recently been carried out of all known PV refrigerators installed
in Mali under various projects (Ref. 4.12). These total 13:3Arco/WSR, 1 ArcoIPolar
Products, 1 BP SolatYLec, 5 F’rigesol, 1 Solavolt/Polar Products, 2 SPC/Adler
44
Barbour. In addition, two Polar Products systems are awaiting instahation.
Statistically significant performance data was available for nine installations.
Analysis showed that the percentage time the systems were fully operational varied
from only 27 per cent (SPC/Adler Barbour) to 79 per cent (Frigesol). Average time
between failures for different system types ranged from 4 months to 20 months.
These results exclude the recently installed systems. Some of the Frigesol systems
and the BP Solar system have been 100per cent reliable in the short time they have
been operating.
Senegal
A 670 Wp medical power system was installed at Mt. Rolland in the Thei district of
Senegal in 1982. The PV system provides lighting, refrigeration, improved
ventilation with the use of fans and high quality power for laboratory instruments.
Overall it has made a significant improvement in health service effectiveness. The
dispensary covers a population of 10,000,providing 100-150consultations per day.
The system costs the equivalent of about $20,000,a high price since the system was
experimental in nature. The system is also oversized (430 Wp would have been
sufficient) due to an overestimate of the load.
Non-Government Organizations
A number of NGOs have installed PV refrigerators in recent years. These include
Oxfam (10 F’NMA units in Zaire and BP Solar/Let systems in Uganda) and the
International Committee of the Red Cross (ICRC) who have installed several
systems in refugee camps. The Save the Children Fund purchased a BP Solar/Let
system which was tested at LESO in Mali before being installed in Douentza. Large
orders from German and British based NGOs for solar refrigerators have recently
been announced.
4.4 Conclusions
Technical aspects
Although some 800 to 1000PV refrigerators have been installed to date, experience
has shown that the technology has only recently matured. Laboratory testing by
WHO and NASA-LeRC has prevented totally inadequate equipment being sent into
the field, but some reliability problems are still being experienced. In particular, the
sizing of the PV arrays and/or the batteries have been found to be inadequate for
actual conditions, in particular in regions of high ambient temperature and poor
insolation levels (eg. Philippines, parts of India). Average availability has been
round 80 to 85 per cent for systems installed from 1981 to 1983. A summary of
observed temperature control is given in Table 4.4.
Systems that have been installed more recently, particularly those from
suppliers with previous experience, are being found to be more reliable. Some
models are now showing 90 to 100 per cent in-service time, with certain
installations operating with 100 per cent reliability for more than two years. Some
suppliers have withdrawn from the market (eg. SPC, WSR and Adler Barbour).
The problems of system sizmg and load prediction remains a cause for
concern. A recent evaluation of tenders for the supply of 23 PV refrigerators for
islands in the South Pacific demonstrated that some tenderers proposed PV array
sizes and/or battery capacities that would be grossly inadequate. Fortunately, many
of the systems tendered were correctly designed. An easily applied method for
prospective purchasers to check system sizing would be a significant help.
The ice-making capability of most PV refrigerators commercially available is
45
less than 2 kg/day. Some users have expressed the opinion that this is inadequate
for many vaccine cold chains. Battery maintenance has been a common problem
with many systems. The possibility of developing a battery-less refrigerator making
use of soft-start compressor motors and thermal storage instead of electrical
storage should be given more attention as a development project.
The field trials have highlighted the need for a number of relatively minor
improvements. These include the provision of door locks on some models and the
relocation and/or redesign of some thermostats to reduce the possibility of
unnecessary adjjustments.
Economic aspects
Very little work has been undertaken on assessing the financial benefits of solar
refrigerators using actual field data, but it is important to ensure that investment in
PV refrigerators constitutes a sound use of development funds. The WHO-EPI is not
an ‘economic activity’ in the normal sense and so it is not possible to carry out a
cost-benefit analysis. The only meaningful analysis involved the comparison of
costs of the various options and their likely influence on the achievement of the
immunization programme objectives. It is important to note in this regard that the
SYSTEM
Per cent times within temperature
bands
No. Days
Data
Correct,
0” to 8°C
High, >8”C
Low, CO”
1
89.92
10.08
0.00
295
2
86.22
0.00
13.50
711
3
71.72
0.17
28.11
471
4
82.25
11.45
6.29
3057
5
79.11
7.03
13.86
462
6
73.86
0.00
1.59
440
7
62.56
1.37
36.07
219
Source: WHO.
Table 4.4 Observed Temperature
Seven &stems (to June 1985)
Control during
46
WHO Field Trials on
fixed costs for any immunization programme are generally large compared with the
direct costs of vaccine refrigeration.
Kerosene refrigerators used in the vaccine cold chain have an initial capital
cost of only $300 to $800, considerably less than for PV refrigerators. With
transportation and installation, this may rise to $1500 installed compared with
about $6000 for an installed PV system. The operation and maintenance costs of
kerosene refrigerators are high however and their reliability is low, sometimes
resulting in an availability of only about 50 per cent.
The results of a comparative cost analysis relating to an actual immunization
programme in The Gambia are presented in Reference 4.6.based on data collected
in 1984 and 1985.
It was concluded that the overhead cost per dose is reduced by $0.06to $0.07
by using a PV refrigerator, due to the greater reliability. The refrigerator cost per
dose is small compared with the overhead cost and is not significantly different
between kerosene and PV. The overall cost per dose is cheaper for the PV
refrigerator even where the PV capital cost is high.
It is important to note that periods when vaccinations cannot take place result
in incompleted and hence ineffective courses of vaccinations. This effect is difficult
to quantify but clearly favours the refrigerator with the higher reliability.
A life cycle costing comparative analysis is given in Reference 4.4 which also
concluded that the poor reliability of kerosene refrigerators makes photovoltaic
refrigeration more economic in comparison. The results of this analysis are
presented in Figure 4.4. The data assumed is given in Table 4.5.
Unit Cost Of Refrigeration
$ Per Litrc-Month
I
QPM
n Recumnt
I
KEROSENE
Figure 4.4 Unit Cost Comparison
Rt$bigerators
for
47
Kerosene and Solar Powered
Social and institutional aspects
Based on the reported experience of PV refrigeration projects to date, there is no
doubt that the systems are widely acceptable to the users. The main need is to
ensure a considered approach is taken to project implementation, in particular
with respect to:
-project design (selection of systems and sites)
- selection of local implementing agencies
-user training (operation and maintenance and trouble shooting)
-technical support centres serving each region using PV refrigerators
- communications with technical support centre.
The recommendations of the WI-IO-EPI with regard to procurement of
approved equipment and the contractual arrangements for warranties and training
of operating and maintenance staff should be followed by alI authorities wishing to
install PV refrigerators for the vaccine cold chain.
WHO-EPI have now commissioned the preparation of installation, user and
repair technician handbooks for photovoltaic refrigerators for use in the vaccine
(cold chain. In addition regional technician training courses are planed by WHOEPI. Both of these initiatives should assist with the successful introduction of solar
refrigeration into vaccine cold chains.
There is a continuing need to gather data on system performance and
therefore efforts should be made to provide the necessary instrumentation and
organization required to monitor the systems in the field. The information, both
quantitive and qualitative, should be passed on to the EPI co-ordination office in
Geneva The following minimum information should be recorded on a daily basis:
-maximum and minimum internal temperatures of the refrigerator and freezer
- am’bient air temperature
- solar irradiation
-system use (kg of ice removed, vaccine removed)
-details of breakdowns or component faults.
For system optimization studies, the electrical energy delivered by the PV
array and the energy consumed by the motor/compressor unit is also required. The
preferred method of data collection is to use data loggers supplemented by a log
book or pro-forma sheets for noting details of system use and reliability.
Refrigerator type:
Photovoltaic
Kerosene
Low Typical High
case
case
Parameter
100
4500
1500
150
15
95
Net vaccine capacity (l&es)
Initial Capital Cost ($)
Cif and installation ($)
Fuel costs ($/day)
Maintenance costs ($/year)
Life time (years)
Availability (% time in service)
100
300
500
0.25
50
10
80
100
500
800
0.70
100
5
50
100
1000
1000
3.00
150
2
20
Table 4.5 Data Assumed in Unit Cost Comparison Presented in Figure 4.4
48
Chapter 4 - References
4.1 ‘Solar Refrigerators for Vaccine Storage and Ice Making’, World Health Organisation, EPI/
CCIS/Sl.S, 1981.
4.2 ‘Purchasing Guidelines and Product Information Sheets’, World Health Organisation, EPI/
CCKY86.4, 1986.
4.3 ‘The Cold Chain Product Information Sheets’, World Health Organisation, SUPDIR 55 AlT 5,
1986.
4.4 A. Derrick and J.M. Durand, ‘Photovoltaic Refrigerators for Rural Health Care - Experiences
and Conclusions.’ Proc. of the UKISES Conference Solar Energv for Developing Countries Power for Villages, London, May 1986.
4.5 A F Ratqjczak, ‘Photovoltaic-powered Vaccine Refrigerator/Freezer Systems; Field Test
Results’, NASA-LeRC (1986).
4.6 ‘Pre-Investment Report on Solar Photovoltaic Applications in the Health and
Telecommunication Sectors, The Gambia’, UNDPWorld Bank, March 1986.
4.7 ‘Solar-Powered Refrigerators for Vaccine Storage and Icepack Freezing: Status Summary June
1985’, World Health Organisation, EPl.KCXY85.4, 1986.
4.8 IT Power Inc., ‘Photovoltaic Powered Refrigerators’, evaluation report for Meridian
Corporation, ref 86153/04, February 1986.
4.9 B McNelis and J M Durand, ‘Photovoltaic Refrigerators and Lighting System for Zaire’, %. of
6th EC Photovoltaic Solar Energy Coherence, London, April 1986.
4.10 IT Power Ltd. ‘Report on the Evaluation of Tenders for Solar-Powered Refrigerators’, for South
Pacific Bureau for Economic Co-operation, April 1985.
4.11 P Jourde, ‘French Polynesia Already in Solar Age’, Proc. of 6th EC Photovoltaic Solar Energy
Conference, London, April, 1985.
4.12 TJ. Hart, ‘Technical Assessment of Solar Refrigerators in Mali’, Draft Report, January 1986.
49
5. LIGHTING
5.1 Introduction
Alternative lighting techniques
Lighting is a steadily growing need in ihe rural areas of developing countries, not
*x;ising but also because more people want to be
only because the population is ~J\C
active in the evening. S&xzl chi‘idren need to study and there are new work and
leasure opportunities fo: [email protected] An important need is for lighting for small
commercial enterprises in ?.hestreets, such as food stalls, shops and recreational
activities. In addition 1x3these residential and commercial needs, there is an
associated need for lighting for streets and public open spaces.
In areas where there is no electricity supply, lighting for domestic and
commercial applications is usually provided by kerosene lamps or candles. In
general, lighting from these sources is of poor quality, expensive and a fire hazard.
The best light using kerosene comes from a pressure device (Coleman type), but
these are expensive. The more common wick devices (hurricane lamps) produce
less than 15 per cent of the light of a 20W fluorescent tube. Due to the high price of
kerosene in remote areas, a household may have to spend the equivalent of over
$200 a year to operate two kerosene lamps.
Photovoltaic lighting systems would be an attractive alternative to kerosene
lamps and candles throughout the areas where it is likely to be many years before
regular electricity supplies become available. The key considerations are
comparative quality, reliability and cost.
Technical requirements
PV lighting systems have become readily available over the last five years, with
manufacturers offering two basic types of unit, one for area lighting, the other for
domestic applications.
Area lighting units may be used for street lighting, public open spaces and
security lighting. These systems consist of PV array; battery; simple voltage
regulator; timing and/or photosensitive switch controls; and an efficient
fluorescent or low-pressure sodium or mercury vapour lamp. Several
manufacturers offer complete self-contained units including poles with mountings
for the lamp and the PV modules and a weather-proof container for the battery and
controls. As the poles represent a significant proportion of the total cost, some
manufacturers supply only the PV array, lamp, battery and controls, to allow the
purchaser to provide the pole from local sources.
Domestic lighting units typically require only one or two PV modules for
charging a battery which supplies from one to four fluorescent tubes, from 20W to
40W depending on the application. Some systems are portable, with a lantern unit
incorporating a rechargeable battery. Larger systems can be obtained, capable of
supplying other end uses such as refrigerators, radios and televisions, but it is more
appropriate to consider these systems in the next chapter under the general
heading of rural electrification.
Fluorescent lamps are commonly used for both area lighting and domestic
51
lighting systems. Fluorescent (or gas vapour) lamps offer high efficiency, long life
and high reliability. They require a ‘ballast’ and a ‘starter’ which give a high
frequency impulse for starting, followed by much lower power and frequency for
normal running. Standard AC fluorescent units may be converted for DC powering
(and therefore suitable for PV systems) by changing the ballast and starter
components, a relatively simple task.
Field experience
Several thousand PV lighting systems are in use in developing countries.
Experience is particularly extensive throughout the South Pacific and more
specifically in Papua New Guinea, F‘iji and French Polynesia.
There are privately-funded schemes in some countries to enable the benefits
of PV lighting systems to be accessible to relatively poor people. In the Dominican
Republic, for example, a US-based organization distributes PV lighting systems to
villagers in the northern part of the country, with loan finance repaid over two to
five years.
6.2 Principal
case study -
South Pacific
Papua New Guinea
A number of PV systems have been installed in Papua New Guinea (PNG) for
communications, lighting, water pumping and medical refrigeration. The total
installed capacity in 1982 was about 50 kWp, of which over half was for
telecommunications systems (Ref. 5.1). The potential for domestic lighting in the
villages over the next 10 years was estimated at about 500,000single module units
of 35 Wp each, a total of 17.5 MWp.
Field experience with the early types of lighting systems (prior to 1980) was
not satisfactory, as the battery charge controllers were found to be complex and
unreliable. Since then, fully tropicalized charge controllers have been supplied and
these have proved reliable. In terms of light quality, a 20W fluorescent lamp was
found to give an illumination of 100lux at a distance lm below the lamp, whereas a
kerosene pressure lamp was found to give only 12 lux at a similar distance. There is
general agreement that PV lighting systems are now technically superior to
alternative forms of lighting in rural areas. The PNG government has now decided,
on grounds of safety and reliability, to use PV lighting systems for official patrol
posts in the villages.
Regarding economics, reference 5.2presents the results of a survey which was
conducted among 30 village houses to assessthe cost of kerosene-fuelled lighting
systems for comparison with PV systems.A typical household was found to use one
hurricane lamp and one pressurised lamp. Annual costs were found to be the
equivalent of about $230 (1981 prices). The Present Worth at 10 per cent discount
rate of five years expenditure is thus about $970. The comparison PV system,
comprising a 35 Wp module, battery, regulator and two 20W fluorescent lamps,
could be installed in 1981for about $775,with the module guaranteed for five years
and the battery for two years.
Although the PV system was cheaper than the kerosene lamps over a period of
five years, it was recognised that the capital cost would be a formidable barrier to
the potential user. In view of the improved technical quality of the lighting and the
benefits to the national economy, the study concluded that lending institutions in
PNG should be encouraged to provide finance to customers wishing to purchase
PV lighting systems.
Fiji
Most Fijian villages, consisting typically of 10 to 50 homes, are located at too great a
52
distance from the exsisting national grid to be connected at acceptable cost. For
many years, the mi Public Works Department has been installing small diesel
generating plants (7.5 to 15 kVA) in villages to provide electricity mainly for lighting
purposes. To date, however, less than 100villages out of more than 2000 in Fiji have
received such installations, due to limited technical and financial resources.
In view of the potential advantages of PV lighting systems, a programme was
initiated in 1982 by the Fiji Department of Energy to test and demonstrate PV
domestic lighting systems (Ref. 5.3). A specification was drawn up for a system
comprising two 15Wfluorescent lamps, a 35 Wp PV module, a battery charging unit
for four Ni-Cad ‘D’ cells for powering a portable light or radio and a gel-type main
battery. The system is a total package including all components and instructions for
field installation and use.
After international tendering procedures, a total of 100 systems were ordered
from South Pacific Solar (USA). These were delivered in early 1983 and installed in
three villages. Although each system was partially subsidized, all users had to make
a contribution to capital costs and pay a regular sum for maintenance and
improvements. The systems have been widely welcomed by the users.
An economic study was also carried out in 1983 comparing the costs of
kerosene lighting with diesel generator-based systems and PV systems. Fuel costs
for wick and pressure lamps (kerosene or benzine) for a typical household were
found to range from the equivalent of $120 per year to over $180 per year,
depending on the remoteness of the site and the associated fuel transport costs.
Fuel costs for urban users were considerably less, typically $90 per year. A life of
five to seven years is reasonable for pressure lamps with a replacement cost of
about $40 each. Wick lamps have an initial cost of from $5 to $15 and have an
indefinite life unless broken.
Diesel generating sets were found to cost typically $2000 per year for fuel and
maintenance in a village of 30 houses. The initial cost was $550. Installation would
cost a further $50.Assuming a battery life of four years, the replacement costs over
10 years would be $220.Annual maintenance was assumed to cost about $15 per
system, to cover an annual visit by a qualified technician to a group of 30 systems.
Based on the above costing assumptions, the Present Worth of all costs over a
10year period for providing lights for a village of 30 homes for the three methods is
as follows:
- Kerosene and benzine lamps $28,000
- Diesel-generator system
$26,000
-PV systems
$25,000
The study concluded that for remote villages where lighting cannot be
provided by grid-delivered electricity, photovoltaics are marginally better
economically and should be examined carefully in the light of their advantages over
diesel electrification and kerosene or benzine lamps.
Initial results from the demonstration projects in Fiji have been favourable,
with a high level of user acceptance and a reasonable prospect for economic
success.
French Polynesia
In 1983,the French Polynesian fm of SOL E.R. received a rush order to supply
electric power systems to 200 homes on the remote island of Faaite. Within a week,
this locally-run renewable energy firm had assembled and despatched by boat 200
PV modules, 40 regulators, 60 batteries, 350 fluorescent lamps, 5 refrigerators, 5
washing machines, 3 water pumps, and other electrical equipment. This equipment
was installed over the next four weeks (Ref. 5.4).
53
This report of the rapid electication
of an island is apparently typical of
many similar projects undertaken in the South Pacific. Most of the equipment, apart
from the solar cells which are imported from Prance, is designed and built in Tahiti.
The major electrification programme for the remote islands started in 1982,after a
three-year assessment and testing phase. By mid-1983, over 550 houses had been
equipped with PV systems, usually with from 2 to 20 modules mounted on the roof,
at a cost ranging from $700 to $7000 (average $1500) per house. Most of the houses
thus equipped were traditional Polynesian thatched-roof homes.
The purchasers of these PV systems receive a 25 per cent capital grant from the
government. The balance is raised by the householders themselves, either from
their own resources or with the help of a commerical loan. The systems are
reported to be cost-effective in comparison with diesel generators, which are
particularly expensive to install, operate and maintain on remote islands.
An important feature of the PV electrification programme in F’rench Polynesia
has been the training of technicians in system design, construction, installation and
maintenance. In addition to many demonstration projects, a comprehensive
programme has also been undertaken by the government to inform the public in
general of the potential offered by PV and other renewable energy systems
(Ref. 5.5).
6.3 Other lighting projects
Mali
A classroom in Mali was equipped with a PV-powered lighting system in 1980 to
enable it to be used for evening classes. Although five years later the system is
reported to have operated well, requiring little maintenance, the high initial cost
was seen as a major barrier to widespread replication, in view of the limited funds
available for the development of rural education.
United Arab Emirates
PV-powered street lights have been installed in a number of countries in the Middle
East, including a project in Dubai involving 21 street lights and a high-mast flood
light for a traffic circle. The equipment was supplied by Mobil Solar Energy
Corporation (USA) in 1983. Each street light consists of a 20W fluorescent tube,
two 35 Wp PV modules, battery with regulator and controls. The high-mast system
consists of eight 400W high-pressure sodium vapour (‘BPS) lamps powered by a 15
kWP PV array and battery bank, with inverter and control system.
Five commercially available fluorescent tubes were tested at the design stage
and large differences in ilhunination efficiency were observed. The most efficient
tube and ballast combination was chosen for the actual installation. The street
lights give a good level of illumination and are reported to be operating reliably.
There were initial problems with the BPS lamp due to the inverter used which
was overcome by replacing it with one of a different type.
China
Three hundred and fifty million people in China live in vilIages without an
electricity supply. Many are nomadic herdsmen living in Inner Mongolia and other
remote provinces using candles, kerosene and the oil derived from sheep for
lighting. The costs of transporting kerosene are very high and extending the grid
would be out of the question in many cases. PV lighting systems however offer
particular advantages and experience of recent years indicates that such systems
would be technically and econorr,zally feasible, In 1985,it was reported that there
were over 2000 PV lighting systems of various types being used in China (Ref. 5.6).
54
Dominican Republic
The US-basedorganization Enersol Associates Inc. has been helping villagers in the
remote parts of the Dominican Republic install PV lighting systems since the fist
was installed in April 1984.A total of 15 systems had been installed by the end of
1985 and a rural solar co-operative formed to facilitate financing and education
through information exchange. A hardware store and service centre has been
opened to provide a continuing back-up service to users. Peace Corps personnel
have helped run training workshops (Ref. 5.7).
With assistance from US-AID and other sources, Enersol Associates have
established a revolving fund to finance the co-operative’s activities. Local users pay
for their PV systems over a period of two to five years with reasonable interest
rates.
The calculated cost of electricity provided by the Gternative technologies
available for remote areas are as follows: (Ref. 5.5).
- PV system
$l.OO/kWh
-Car battery
$B.SO/kWh excluding costs of transporting the
battery from the r,earest place where it can be
recharged.
-Kerosene wick lamp
$3.5Oikwh
- Kerosene pressure lamp $14.OO/kWh
- Dry cell battery
$30~$35/kwh
Zaire
An area where lighting can make a substantial impact is in rural health clinics and
medical centres. In Zaire the Departement de la Sante Publique is installing 860
lighting systems as well as 100 photovoltaic vaccine refrigerators. The project is
funded by the European Development F’und.
5.4 Conclusions
Techuicai aspects
PV lighting systems covering a wide range of sixes and types are widely available as
standard products. The components required for a typical domestic lighting system
are listed in Table 6.1. Such a system would provide up to 200 Wh/day of useful
energy for lighting given a solar input of 6.0 kWh/mz per day. The 20W lamp could be
used for 3 to 5 hours every night for general activities and the 7W lamp could be
used for 8 to 12 hours for security. The battery provides about three days storage.
Many of the smaller systems for domestic use are portable, which makes them
particularly suitable for use in place of kerosene lamps. The introduction of longlife rechargeable Ni-Cad batteries is an interesting development in this regard.
The battery charge controllers used for some early designs of PV lighting
systems were found to be unreliable, but now fully tropicahzed units are supplied
which have proved very reliable in practice. The reliability and efficiency of the
ballasts used in commercial fluorescent lamp units have been found to be variable.
Careful selection of this component is therefore essential, particularly as it
accounts for up to 75 per cent of the cost of the lamp. The lifetime of the DC ballasts
used for some 60 DC fluorescent fixtures tested as part of the NASA-LeRC PV
medical refrigerator programme was found to be less than five years. Further
experience of the lifetime of these components under field conditions is needed.
56
Life
bars)
Specification
15
250-300
12V/105Ah
4
50-80
lamps
2
50-70
5
5.0-200
1 PV module Battery Fluorescent
Price
(44
Battery charging
40 Wp
controller
Total
Table 5.1 Tgpical PV Lighting
400-500
Sgstem for Domestic Use
PV-powered street lights and security lights are also available from several
manufacturers. These units generally use low-pressure sodium vapour or highpressure mercury vapour lamps. Some problems connected with the need to adapt
standard AC units for DC operation have been experienced and there is need for
further development of suitable ballasts.
Economic aspects
Lighting is not a directly economic activity and therefore a costbenefit analysis is
not posssible for this application. A number of cost comparisons for alternative
lighting methods, including the projects detailed earlier, indicate that PV lighting
systems offer the cheapest solution for lighting in villages where grid electricity is
not available.
It should be noted that the light output of a kerosene pressure lamp is about
200 lumens. That for a kerosene hurricane lamp is about 80-100 lumens, whereas
the light from a 2OWfluorescent tube is about 1000 lumens. Also the life of the PV
modules (the most expensive part of the PV system) is at least 16 years, much
longer than the life of kerosene lamps.
Provided suitable means are available to finance the initial cost with
repayments over say 6 years, PV systems should be widely attractive on both cost
and performance grounds to potential users in the villages.
Social and institutional aspects
For the successful introduction of PV lighting sytems, the potential users need first
to be convinced of their technical performance and reliability. This requires
demonstration systems to be available, possibly as mobile units to be taken from
district to district.
There are then two major institutional requirements:
a) The provision of finance, probably through the provision of a subsidy and lowcost loan, repayable over two to five years;
b) The provision of technical support, in particular for the supply of spare
fluorescent tubes and balk& and batteries.
56
In many cases, it would be preferable for the implementing organization to
establish local cooperatives, who can arrange for the administration of funds and
the provision of technical supports. Training for key personnel would be needed.
The local co-operative would need to be able to refer major problems to a central
resource centre.
Although in the short and medium terms most countries would need to import
the PV modules, most other components could be locally manufactured and
assembled, thereby greatly reducing the foreign exchange requirements whilst at
the same time building up local technical skills.
Chapter 5 - References
6.1 G.H. Kinnell, ‘Solar Photovoltaic Systems in the Development of Papua New Guinea’, Proc. of
the Fourth EC Photovoltaic Solar Energy Conference, Stresa, May 1982.
6.2 K. Maleva, ‘Feasibility Assessement for Photovolaic Cells Replacing Kerosene Lighting in
Papuan Villages’, Report No. 7/81, Energy Planning Unit, Department of Minerals and Energy,
Konedobu, Papua New Guinea.
5.3 HA. Wade, ‘The Use of Photovoltaic Systems for Rural Lighting- an Economic Analysis of the
Alternatives’, Proc. of the Solar World Congress, Perth, August 1983.
5.4 D.O. Hall, ‘Electrifying an Island a Month’, Earthscan Feature, London, 1984.
5.6 ‘Renewable Energy in French Polynesia’, by SOL E.R. and CEA-Ger, Pappeete, French
Polynesia, 1986.
6.6 Zhu Gangi, ‘Recent Experience on Solar Energy Utilisation in China’, Proc. of UK-ISES
Conference C42, Energy for Development - Where zux?the Solutions?Reading, UK. Dee 1985.
5.7 ‘Newsletter No. 4’, Enersol Associates Inc. Somerville, USA, Jan 1986.
57
6.
RURAL ELECTRIFICATION
6.1 Introduction
Conventional approaches
There are two main techniques used for rural electrification at present: either the
main electricity grid is extended to cover the selected area, or a diesel generating
station is established to serve a smalI network not connected to the grid. Both
approaches have associated economic and technical problems. Extending the grid
over long distances is expensive and, in the initial years at least, the loads are often
small, resulting in load factors and stability problems for the system as a whole.
Maintaining long transmission lines over difficult terrain presents difficulties for
the utility.
Diesel generators require regular supplies of fuel, which often presents
problems for remote areas, particularly at certain seasons of the year when roads
are practically impassable. Moreover, any national fuel shortages are likely to
impinge more severely on the rural areas. In addition to the provision of fuel, it is
often the case that the operating utility finds difficulty in retaining competent
operating and maintenance staff, since as a parastatal organization they are unable
to offer competitive salaries. Even given the necessary staff, it may be practically
impossible to obtain the necessary spare parts to keep the engines running.
Rural electrification is regarded as a development priority in most developing
countries, for social and economic reasons. Although vast sums are expended each
year on rural electrification projects, it wiIl nevertheless be many years before
villages that are a long way from the main electricity grid lines or from the nearest
all-weather road will benefit from a reasonably reliable and affordable electricity
supplyThere is no doubt that new and renewable energy sources would be preferred
for rural electrification, if systems were available at acceptable costs and with
proven reliability. In some locations, wind generators or micro-hydro systems may
be feasible, but in general photovoltaic systems would be favoured if the costs were
right, since they involve no mechanical moving parts and require only simple
maintenance.
Design and load estimation
For many households in developing countries, the main use for electricity would be
for lighting. Many thousands of small PV systems dedicated for lighting
applications are already in use worldwide, as discussed in the previous chapter. For
many households, however, electricity would be useful for many other
applications. Some of these could also be covered by small dedicated stand-alone
systems, such as sytems for water pumping or rice milling as discussed elsewhere
in this survey, but in general it would be preferable for rural electrification schemes
to be non-specific, with the electrical power available for any small-scale
application.
The great majority of users would be domestic households. Based on data
presented in References 6.1 and 6.2,for design purposes the peak demand and daily
59
energy consumption for a typical household would be as shown in Table 6.1.
Although in general it may take several years to build up to the power and energy
levels shown, as most households could not afford all the appliances involved at
once, it is prudent to plan a rural electrification scheme for the situation likely to
arise within a few years.
There will of course be a number of connections which exceed this load
estimate, as well as many connections which are much less. There are differences
between countries, depending on standard of living and social patterns (eg. number
of persons living in one household). In addition to estimating domestic load, it is
necessary to calculate specific loads for commercial, institutional and industrial
applications.
Household versus central systems
To date, most PV demonstration projects for rural electrification have consisted of
either small packaged systems for lighting or other specific end uses, or larger
centralized projects serving a whole village. For a demonstration project, it is
administratively easier to install a single central unit to serve a whole village. There
is no doubt that the larger the system, the greater the publicity. There are however a
number of important advantages to be gained by installing individual household
systems rather than a central PV plant. The main reasons for preferring household
systems (from Reference 6.1) are as follows:
(a) The PV arrays can be roof-mounted, out of reach of livestock and people and
not taking up valuable land;
(b) A distribution system is not required to take the power to each house, an
expensive item if the houses are widely spaced;
(c) No metering system is needed, thereby avoiding the associated admiistrative
costs for meter readers and bilhng computations;
(d) With no distribution system, the problems associated with unauthorized
connections and theft of electricity are avoided;
(e) A centralized system may soon prove unreliable, due to its complexity and to
possible overloading aggravated by unauthorized connections; failure of the
system affects everyone, whereas failme of a household system affects only
one consumer.
Although a centralized system may appear a little cheaper in initial cost, even
after ahowing for the cost of the distribution system, this is heavily outweighed by
the reduced operational problems and administrative costs associated with
separate household systems.
System features
If PV generators are used for rural electrification, whether through household
systems or centralized plants, the DC output would normally need to be converted
to AC at the national standard voltage and frequency. Although most of the loads
such as lighting and domestic appliances could be designed or adapted for DC
operation, this would be unpopular with consumers, who would wish to have the
freedom to buy cheaper standard AC products from the normal suppliers in the
markets. DC equipment would introduce too many complications. A typical PV
system for rural electrification by means of a central plant is shown [email protected]
in Figure 6.1. The schematic arrangement for a household system is similar, as
shown in Figure 6.2.
Central plants would need to have battery storage for about three to five days
supply, depending on the Loss of Load Probability (LOLP) level considered to be
60
Power
(WI
Load
Duration
(hrslday)
Energy
(Whlday)
Lights3 x 20 =
60
5
300
Fans 2 x 60 =
120
6
720
60
4
240
200
1
200
TV
iron
Miscellaneous
60
1.5
500
Total
Table 6.1
9’0
1550
Typical Design Electrical
Load for Rural Household
appropriate. The battery storage for household system could perhaps be less, say
two to three days supply, since energy management procedures could more readily
be introduced by the user when an alarm on the system indicates that battery
charge is low.
An energy meter would not be required for a household system, since it would
be most appropriate to charge the consumer a constant monthly or quarterly rent
for the system. Although the consumer would need to understand the load
limitations of his system, there would be no particular advantage in trying to
economise on the use of electricity, at least all the while the battery state-of-charge
was within its normal operating range.
Commercial equipment
There are no standard systems for central PV plants, although a number of
demonstration plants have been built and all the major PV manufacturers are able
to offer a design service for such plants. Several of the PV pilot and demonstration
plants sponsored by the European Community are for rural electrification (eg.
Aghia Roumeli, Crete, Greece, b0 kWp and Kaw, French Guyana, 35 kWp). In
addition to the central plants, there is also an EC-supported demonstration project
currently in progress which provides for the installation of small stand-alone PV
generators for some 40 houses in the south of France and in Corsica involving three
standardized designs: 400 Wp, 800 Wp and 1200 Wp.
A number of village electrification projects involving photovoltaic systems
have been undertaken in developing countries. Two villages in Indonesia have been
equipped with central PV plants and there are similar schemes in Tunisia, French
Guyana and Senegal. There are also a number of villages on Greek islands in the
Mediterranean equipped with central generators.
In addition to these and other central plants, the alternative disaggregated
approach using stand-alone systems for household and specific end uses has been
followed in bench Polynesia, Gabon and elsewhere.
61
PV Array
wrrrrmlrr
I
TI
-mI
Battery
Storage
Figure 6.1 Schematic of PV Central Plant for Village [email protected]
6.2 Principal case study - Indonesia
Background
A joint Indonesian/West German project called ‘Solar Village Indonesia’ was
initiated in 1979. The project included a number of collaborative activities to
develop systems for the direct and indirect use of solar energy in a tropical
environment, with the long-term objective of improving living conditions and
agricultural output in the rural areas (Ref. 6.3).
Four field test sites were selected for different activities within the overall
programme. The village of Picon in West Java has 350 inhabitants (60 families) and
was provided with two main systems:
- two fxed-bed gasifiers and a village biogas plant which together produce gas to
operate engines which drive electrical generators (60 kW)
- a 5.5 kWp PV generator producing power for 10 irrigation pumps, carpentry
workshop, rice processing and village lighting (see Figure 6.3).
The village of Citius on the northern coast of Java has a population of 780
people, the majority fishermen. The ice needed for fish conservation was brought
daily from 20 km away. A 25 kWp PV system was installed for the following
applications:
- a vapour compression ice-making plant for producing 500 kg of crushed ice per
day
62
3.
-=-Q...
‘\
SUN
‘\
. -
PHOTOVOLTAIC ARRAY
BATTER1:ES
REGULATOR
4
EMERGENCY
LIGHTS
I
4
TELEVISION
REFRIGERATOR
Figure 6.2 Schematic of PV System for Household Electr#Zcation
(Source: Solar Energie Technik)
- A Reverse Osmosis desalination plant providing 5 m3of drinking water per day
using brackish water from a well
- a comprehensive data-recording system.
In addition, two PV-powered navigation buoys were installed to mark the
harbour entrance and a PV-powered television set provided for use by the villagers.
Puspiptek is an Indonesian government research station occupying a 350
hectare site situated near Jakarta. In Puspiptek, five solar thermal pilot plants are
installed for research and development purposes.
63
Figure 6.8 5.5 kWp PV Array at Picon, Indonesia
The fourth field test centre is on the remote island of Sumba south-east of Java.
Many villages on this island do not have any water at all, so that the inhabitants in
some caseshave to walk up to 15 km to the nearest well. The aim of the activity on
Sumba is to show that photovoltaic pumping systems offer a realistic solution to
water supply problems encountered in remote, arid areas. Three PV pumping
systems have been installed as part of the project: a 5.76 kWp DC system at Gollu
Watu pumping water from an underground river 28m deep for a population of 6000
throughout the year; a 3.65 kWp AC system at Pemuda pumping water from a 60m
deep well for a population of about 2000in the dry season; and another 3.65kWp AC
system at Wee Muu providing drinking water from a 37m deep borehole for about
2000 throughout the year.
Experience
The ‘Solar Village Indonesia’ project (Ref. 6.4) is being carried out in three phases,
as follows:
Phase I - Definition phase, completed September 1980
Phase II - Implementation phase, completed 1983
Phase III - Performance testing, operation and maintenance, evaluation of
systems, from 1983.
A joint working group has been established to undertake the tasks included in
Phase III. The group consists of a test and evaluation team, an operation and
maintenance team, experts for special problems and tasks, and an expert from W.
Germany permanently on site.
PhaseIII is still in progress, but it is envisaged that when the evaluation studies
have been completed it will be possible to replicate similar systems at other sites in
Indonesia. Future systems will incorporate such improvements as have been found
desirable and will be manufactured as far as possible in Indonesia.
Preliminary results from the 5.5 kWp PV plant at Picon village have indicated
that overall system performance and efficiency has been strongly influenced by the
poor partial load performance of the inverter originally installed. A new inverter
with improved performance has now been incorporated.
64
6.3 Other rural electrification
Tunisia
projects
A solar village project involving PV, wind and solar heating systems has been
operating in Tunisia since 1983(Ref. 6.4).The main photovoltaic generator consists
of a 29 kWp array with battery bank and inverter to deliver power at 22OV,50 Hz, to
the vil.lage. There is also a 1.4 kWp stand-alone system serving a farm and two 1.4
kWp drip irrigation systems.The solar village was developed with the assistance of
the NASA-Lewis Research Center (USA). Operation and evaluation is now the
responsibility of the Societe Tunisienne de 1’Electricite et du Gaz (STEG).
French Guyana
A 35 kWp PV plant to serve the isolated village of Kaw in French Guyana was
installed in 1982 as part of the European Community photovoltaic pilot plant
programme (Ref. 6.5). The PV system replaced two diesel generator sets which had
proved unreliable and expensive to operate. The system consists of 528 PV
modules, each rated at 66 Wp, arranged in 44 panels of 12modules each. The battery
bank provides 410 kWh storage capacity to ensure continuous power supply
throughout the year to the village. A high efficiency inverter is used, giving three
phase and neutral output at 38OV,50 Hz.
The system is operated and maintained by the French electricity utility, EDF.
Special training for the EDF technicians responsible was provided by the main
contractor, Seri Renault (France) and the inverter supplier, Jeumont Schneider
(France). The operation has been practically trouble free since commissioning in
January 1983, fuhilhng all expectations.
Senegal
A PV/wind hybrid generating system started operating at the village of Naiga Wolof,
Senegalin February 1983.The PV portion is rated at 6.5 kWp and the wind generator
is rated at 4.5 kW. Initially the system provided power only for public lighting. Then,
in April 1984, a water pump was added, followed by two refrigerators and
additional lights. The system is eventually intended to supply residential lighting, a
communal television, carpentry equipment, a sewing machine, a grain mill and an
ice maker.
The PV portion of the system is reported to have operated reliably without the
need for a full-time system operator.
French Polynesia
As reported in the previous chapter, there is a major rural electrification
programme for remote areas of French Polynesia.
The programme is supported by the French Atomic Energy Commission
(CEA), the French Agency for Energy Management (AFME) and the Government of
French Polynesia. The majority of the PV systems installed are relatively small,
designed to be stand-alone systems for individual households. In addition to the
many small systems dedicated for lighting, systems are also being instaIled to
provide for other end uses. A typical household system provides enough power for
three 13W fluorescent lights, an 80W television set, a fan and a small refrigerator.
The cost of such a system is about $2000,including taxes, but a government subsidy
is available covering 50 per cent of the module price. Finance is available for
purchasers to enable them to spread the cost over a five year period.
66
6.4 Conclusions
Technical aspects
There is limited experience available to date on the performance of larger
centralized PV plants for village electrification. Some systems have worked well,
others have experienced problems with inverters and other components. No
standardized designs exist as yet and care must be taken at the design stage to
ensure that only components with proven reliability are used in systems that are to
be installed at remote sites.
There is however a considerable volume of experience with smaller household
systems, with many hundreds of installations in French Polynesia and elsewhere.
Standardized designs have been developed and performance is generally reported
to be very satisfactory.
Economic aspects
There is little economic data on the cost of installing, operating and maintaining
centralized PF7 plants for village electrification. There is however extensive
information on the economics of household systems. The current (1986) installed
cost of a small (less than 500 Wp) stand-alone PV generator (including batteries and
inverter) is at least $BO/Wpif bought as a one-off item. If however standardized
systems are bought in large quantities and integrated with the building, it is possible
to get installed system costs down to around $15/Wp. Based on the low price
scenario for PV modules and system presented by Starr and Hacker in Reference
6.6, a cost projection for standard 360 Wp PV systems suitable for stand-alone
household AC generators can be constructed, as shown in Table 6.2 (Ref. 6.2). The
price per peak Watt is shown as falling from a current level of $16 to $6 by the year
2000.
Component
1985 1990
Modules
1800
1080
810
540
Battery (299 Ah/48V)
1600
1200
800
440
Inverter/Regulator
1000
600
300
250
300
300
300
300
700
----
470
330
270
5400
3650
2540
1800
15
10
7
5
(360 Wp)
Building-related
Supplier’s
costs, transport,
installation,
mark-up
Total instaised system price
Price per peak Watt
etc
1995 2000
Note: all costs in 1985 US Dollars
Table 6.2 Cost Projection for Household PV Systems morn Reference 6.2)
66
It is now relatively straightforward to calculate the average unit cost of
electricity produced by the PV generator for the range of system costs shown in
Table 6.3.Annual maintenance expenses are assumed to be 1.5per cent of the initial
capital cost. Administrative expenses are assumed to be $10 per system per year,
given that there are a large number of systems within each administrative area. The
average electricity cost is based on a 20-year period of analysis and a 10 per cent
discount rate. The average solar input is assumed to be about 6 kWh/m* per day.
Each system is assumed to give a useful energy output of 540 kWh/year. The
resulting average energy costs are as follows:
6
Installed system cost ($/Wp)
15
10
7
0.90
0.64
Average energy cost ($/kWh)
1.34
0.46
These unit energy costs may then be compared with the real (unsubsidised)
cost of electricity supplied by alternative means, such as grid extension or by diesel
generators. Grid extension costs depend on two main factors; the distance the
feeder line has to be extended to serve a new area; and the number of connections
served by the new feeder. Based on typical cost data, the unit energy costs for grid
extension schemes are as set out in Table 6.3 for three values of feeder line length
and five values of the number of connections per feeder.
The unit energy costs for the PV and grid extension approaches are compared
in Figure 6.5.Grid extension is the cheaper alternative for the near future except for
the occasions when it is necessary to provide electricity to a limited number of
connections involving long feeder lines. Indeed, for providing power to a few
isolated houses, it is likely that a PV system will be cheaper than the grid if the grid
is more than a few hundred metres away.
Diesel generation costs are summarised in Table 6.4 based on data published
in Reference 6.4. The unit energy cost for a system supplying 540 kWh/year (about
1.5kWh/day) ranges from $l.OO/kWhfor the low cost caseto over $2.50/kWh for the
high-cost case. Larger diesel generators serving whole villages typicahy give costs
from $O.GO/kWhto $l.SO/kWh, given reasonable maintenance and no major
interruptions in the supply of fuel and spare parts.
Prom this data, it is clear that PV systems are cost-competitive today with
small diesel generator systems. They can also be competitive with larger diesel
generator systems in places where fuel supplies and maintenance present major
difficulties. PV systems will become increasingly competitive with grid extension
schemes as the cost of PV modules and other components continues to fall with
improved technology and larger volume production.
Feeder length
km
No. of connections per feeder
2000
100
200
500
1000
Unit energy cost
IO
1 .oo
0.67
0.47
0.40
0.37
in $/kWh
20
1.69
0.97
0.61
0.47
0.40
30
2.35
1.34
0.74
0.54
0.44
Table 6.8 Unit Energy Cost for Grid Extension
67
(J’rom Reference 6.2)
PV system cost
$lS/Wp
1.404
Jz
3
k
a
*
s
C
.C
2
1.20'
---
Solar
l.OO-
PV
Grid Extension
$10/W
0.8007/HP
0.601 -
15lWP
*z 0.409 .C
2
; 0.201'
w
20
F’igure 6.4 Electricitg
30 Length.of
Feeder Line
Unit Cost - Grid Extension
Size of diesel (kW)
in km
and PV Sgstems
Lowcost
Highcost
2.5
2.5
7
7
Capital cost ($)
1000
3000
Fuel cost ($/Iitre)
0.40
0.80
20
15
200
400
0.5 kWh/day
2.30
6.00
1 .O kWhlday
4.50
3.30
2.0 kWh/day
0.80
2.10
Life of system (years)
Overall efficiency
Operating
(%)
0 & M ($/year)
Unit energy cost in $/kWh for:
Table 6.4 Diesel Generator
Costs
Social and institutional aspects
Whatever method is adopted for rural electrification, costs are bound to be high.
Many developing countries have established rural electrification boards (REBs) to
undertake the necessary planning, implementation and operation of rural
electrification schemes. The REBs normally require substantial external funding,
since it is widely recognised that rural electrification projects can rarely be self68
financing. The selling price of electricity to rural consumers has to be kept at a low
level, comparable to that obtaining in urban areas, since (a) most rural consumers
are very poor and (b) any significant disparities would generate strong political and
social pressures. The cash flow for a typical REB with an expanding programme is
thus bound to be poor, particularly as it often takes several years for the loads and
associated revenues to build up.
Rural electrification schemes therefore have to be economically appraised
and justified on broader grounds than simply costs and revenue from the electricity
system alone. The value of the social benefits that accrue to the community as a
whole, through the raising of living standards, improvements to land and labour
productivity and the generation of new employment opportunities, all need to be
assessed.Thus a rural electrification scheme may be economically justified, even
though the electricity has to be supplied at a loss by the utility. This normally
requires a substantial financial subsidy to be made to the REB by the government.
The true cost of supplying electricity to rural areas may often be more than $l.OO/
kWh, whereas the price to consumers may only be $O.OWkWhor less.
Sometimes the full value of the subsidies involved is hidden in the accounts of
the electricity supply utility, especially in countries where there is no separate REB.
This situation can arise when rural electrification costs are not distinguished from
overall costs and a uniform tariff is applied throughout the country, for both urban
and rural areas.
Thus, when comparing the costs of alternative techniques for rural
electrification, whether grid extension, isolated diesels or photovoltaics, it is
important to ensure that similar assumptions are made regarding the value of
subsidies that are made available.
If it is decided to proceed with a rural electrification scheme based on PV
systems, it is vital to ensure that the users of the new technology are supported by
adequate arrangements for technical assistance and the supply of spare parts. This
will call for good information programmes to help users understand how to get the
best out of their systems and to identify faults. Technicians will need to be trained
to install systems, instruct users in operation and routine maintenance, and carry
out more extensive maintenance operations when necessary.
Chapter 6 - References
6.1 M.R. Staff, ‘Photovoltaic Prospects for Rural Electrification’, hoc. of ZN7TEMOL 86 World
Solar Energy Conference, Montreal 1985.
6.2 M.R. Staff, ‘Rural Electrification - Solar versus Grid Extensions - Updating the Economics’,
Proc. of UK-ZSES Conference, Solar Energy for Developing Countries - Power for Vihges,
Reading, May 1986.
6.3 S. Riphat, ‘Solar Photovoltaic Development in Indonesia’, paper presented at Regional Expert
Seminar on Solar Photovoltaic Technology, 19-14 June 1985.
6.4 Solar Village Indonesia, brochure published by TUV Rheinland Institute for Energy
Technology, W. Germany, and BPPT Badan Pengkajian Dan Penerapan Teknologi, Indonesia.
6.5 ‘Photovoltaic Rural Electrification in French Guyana’, paper by Seri Renault Ingbnierie
resented at European Community Pilot Projects Contractors’ Meeting, Nimes, Prance, 13114
x pril1983.
69
7. OTHER APPLICATIONS
7.1. Agricultural applications
Photovoltaic systems can be used for a number of applications related to
agriculture. Water pumping for irrigation is the major use, as discussed in Chapter 3
of this report. Other applications include agricultural product processing (eg. grain
milling), milking machinery, cattle fencing, refrigerated storage of perishables and
ice product.ion for fish preservation.
All these systems are similar in that they employ a PV array to charge a battery
which then supplies the end use, either direct for DC applications or through an
inverter for AC applications. Because of the high cost of the PV array and batteries.
it is necessary to optimize the design of the system as a whole. It is not normally
advisable to use standard AC or DC appliances, since these, although relatively
inexpensive, often have poor efficiency.
Most PV manufacturers offer standard systems for battery charging, but
specialized advice is generally needed for the selection of appropriate appliances,
such as DC motors for grain mills, or inverters for AC systems, Several
manufacturers offer PV-powered cattle fencing systems, consisting of a small PV
module for charging a battery through a regulator and a standard high voltage pulse
generator. These systems are beginning to fmd wide acceptance in areas where it is
difficult to arrange for cattle fence batteries to be recharged regularly.
There are not many examples of dedicated systems for agricultural product
processing, as usually the relatively small electric motors are powered from a
larger system supplying a number of end-uses. A PV-powered grain mill has been
operating successfully for 6 years in Tangaye, Burkina Faso. It forms part of a 3.6
kWp PV system which also includes a water pump. The grain mill has been modified
twice since the time of original installation in 1979to obtain the required degree of
fineness and consistency, but the overall availability has been reported as over 90
per cent.
Although no installations for milking machines are known in developing
countries, a large PV generator for a milking parlour has been built as a research
project in Ireland. The 65 kWp grid-connected system built on Fota Island in 1982
provides electricity for milking machines and milk processing equipment on a large
dairy farm. The performance is being closely monitored by the University of Cork
and the reliability is reported to be very high (Ref. 7.1).
Another research project in Europe is the PV-powered refrigerated cold store
for agricultural produce built on Giglio Island, Italy, which started operation in
1984.A 45 kWp PV array provides power for an ozonizer (for water disinfection)
and a cold store of about 275 m3.A particular feature of the compressor used in the
refrigeration plant is the control system, which selects the number of cylinders to
be loaded to match the power available from the PV array (Ref. 7.2).
These large PV systems for agricultural applications are still at the
development phase. The smaller systems, for powering grain mills, cattle fencing,
irrigation pumps, and so on, are much simpler in concept and are technically
developed. They are most likely to be cost-effective in rural areas of developing
71
countries where the following conditions obtain:
(a) There is no electricity supply from the grid and the costs and practical
difficulties of running diesel engines are high
(b) The solar input is reasonably good throughout the year, with an average of at
least 4.5 kWh/m’ per day
(c) There is a need for the application for the major part of the year, as the high
capital cost requires a high utilisation factor to achieve low unit costs.
Two important factors affecting the success of any attempt to introduce new
technologies into the rural sectors are (i) the institutional arrangements for
technical training and support and (ii) the local management of the system. It is
significant that the success of the PV grain mill in Tangaye is largely attributed to
the way existing practrces were adapted to form a co-operative to manage the
installation. This gave a sense of communal ownership, encouraging interest and
concern for the success of the project.
7.2 Water treatment
systems
Although not yet commercially available, several PV manufacturers are developing
complete PV-powered water treatment systems. In some systems, the water is first
filtered and then given a prophylactic chlorine dose, usually in the form of sodium
hypochlorite, before being transferred to storage and thence to consumers. In view
of the difficulties in ensuring the supply of hypochlorite and its correct use when
available, some system designers are concentrating on chemical-free water
treatment processes, which involve slow sand filtration for sterilization. This
approach removes all harmful bacteria, but care has to be taken to avoid
subsequent contamination.
In these two approaches, the PV power is used only for pumping the water,
fust from the source, such as a borehole or river, and then through the various
filtration stages to the final storage reservoir. Another type of water teatment
system being developed incorporates a PV-powered UV light for sterilization.
After the necessary development and testing has been completed, these
complete water treatment packages using PV power are likely to be of particular
interest for water supply projects in remote areas where the existing water sources
are known to be heavily polluted. Two British-designed systems are known to be
operating in Nigeria.
7.3 Telecommunications
Telecommunication systems in developing countries have traditionally been
powered by grid electricity or stand-alone diesel generators. Battery banks are
usually provided for security of supply in the event of power interruptions.
Problems of unreliable supply, variable quality (voltage spikes, low voltage) and
high cost of operation and maintenance cause constant problems for the system
operators. The quality of communications frequently suffers as a result.
Photovoltaic generators are particularly suited for telecommunication
systems, since they can provide the relatively small amounts of power required at
remote transmission/reception sites reliably and with little or no maintenance. PV
generators are widely used for this application and many hundreds of systems are
operating worldwide, including in places where the average solar input is as little as
2.5 kWh/m2 per day. In fact, photovoltaic systems have had more commercial
success for telecommunications applications than any other remote power
application.
72
Figure 7.1 PV Sgstem powering a UHF network in South America
(Source: BP Solar)
There are three main types of telecommunication systems which can be PVpowered:
(i) Two-way radios, including radio-telephones
(ii) Radio and television secondary (infill) transmitters
(iii) Telephone systems, including exchanges, repeater stations
and satellite ground stations (see Figure 7.1).
In addition television sets can have a dedicated PV power supply and
educational TV is an application which has found wide use in certain West African
countries.
73
In each case, the PV system is primarily required for battery charging. For the
larger systems, a stand-by diesel generator may be provided, with controls for
automatic start-up if the battery voltage falls to a preset low level. The hybrid
arrangement is optimized for the least cost configuration.
PV systems are likely to be found cost-effective for sites where grid electricity
is not available for loads up to about 2.5 kWh/day for most locations. In some
circumstances, PV/diesel hybrid systems may be found cost effective for loads up
to 10 kWh/day.
7.4 Cathodic
protection
Another application where PV systems have been found substantial commercial
markets is for cathodic protection of steel pipelines and other steel structures.
Cathodic protection by the impressed current method involves maWaining the
steel structure at a negative potential with respect to the surrounding soil or
atmosphere. PV systems are particularly well suited for this application, since they
provide the necessary DC power without the need for transmission lines,
transformers and rectifiers as required for grid-powered systems. An example is
shown in Figure 7.2.
7.5 Unusual applications
Photovoltaic systems should be considered wherever there is a requirement of
small amount of power in remote or inaccessible locations. For example, aircraft
warning lights ontall structures or on hilltops, or navigation lights marking out the
channel into a harbour.
The eleven large steel lattice towers carrying a 230 kV double circuit
transmission line 15km long across the Jumuna River in Bangladesh are each fitted
.
*
‘, .
Figure 7.2 PV-powered Cathodic Protection
Dhabi
74
,.*,
Sgstem for Pipeline in Abu
with five aircraft warning lights powered by a 700 Wp PV array (Ref 7.3). Battery
storage is sufficient for 10 days supply. Other methods of supplying the lights
considered but rejected were diesel generators (access problems for fuel supplies),
induction from the phase conductors (no power during times of outage), earth
wires from the local network (unreliable).
There are now many navigation buoys and lighthouses supplied by PV
generators. The cost savings can be substantial, as servicing the batteries in
navigation buoys of the diesel generators associated with lighthouses is expensive
and sometimes hazardous. There are five lighthouse systems ranging from 2.6 kWp
to 18.2 kWp operating or under construction in the Mediterranean (Greece, Italy,
France) plus several hundred PV-powered navigation lights for harbour entrances
and buoys along the southern coast of Prance. There are similar developments in
many other countries.
PV systems can also provide power conveniently and economically for remote
metering installations, such as river gauging stations, meteorological stations and
groundwater-level monitoring systems. The records can either be transmitted by
radio to the control centre or stored on tape or disk for later collection, Some PV
installations in Prance use satellites (Argos or Meteosat systems) for data transfer.
Chapter
7-
References
7.1 S. McCarthy, G.T. Wrixon and A. Kovach, ‘Data Monitoring of the Photovoltaic Project’,
Presented at the first Working Session of the European Working Group on Photovoltaic Plant
Monitoring, ispra, Italy, 11-13 November 1985.
7.2 ‘Water Disinfection System and Cold Store’, Design Report presented by Pragma at
Contractors Meeting of the Commission of the European Communities Photovoltaic Pilot
Projects, Nimes, France 13/14 April 1983.
7.3 DA. Hughes and A.B. Wood, ‘Jamuna River 230 kV Crossing - Bangladesh: II. Design of
Transmission Line’, Proc. Znstn. Civil Engineers, Part 1, vol 76, pp 951-964, Nov 1984.
75
8, CONCLUSIONS
8.1
Summary of experience
General
Several thousand photovoltaic systems have been installed in developing countries
over the past ten years, the great majority over the past five years. The size of these
systems ranges from a few watts to over 30 kWp. for applications as diverse as
water pumping, vaccine refrigeration, domestic lighting, cattle fencing and
telecommunications. Many of the systems have been of an experimental nature, for
developing and demonstrating the technology, but increasingly photovoltaic
systems are being installed for sound commercial reasons, as being the most costeffective solution for particular applications.
The technical, economic, social and institutional factors associated with each
application need to be carefully considered before any general conclusion can be
made regarding the viability of photovoltaics. Obtaining reliable data on the
performance of PV systems is not easy, as there have been only a limited number of
published reports on actual field experience. It is also not easy to evaluate the
economic prospects, as many systems have not been operating long enough to yield
sufficient data on component lifetimes and replacement costs. Social and
institutional factors have a major influence on the success or failure of projects to
introduce new technologies into the rural sector and many valuable lessons have
been learned in the course of implementing PV projects in different countries.
Water pumping systems
About 2000 PV-powered water pumping systems have been installed worldwide, in
various configurations. Most of these are for water supplies for villages and
livestock watering, but some have been installed primarily for irrigation. The
performance of many systems has been disappointing due to a number of factors,
including:
- Use of unreliable and/or inefficient sub-system components (motors, pumps
and power conditioning equipment)
- Poor overall system design, resulting in poor matching between the components
in relation to the solar input and water level
- Use of inaccurate data regarding solar input and water resource conditions at
the design stage.
There is evidence that the manufacturers have learned from past experience
and that the latest types of pumping systems are considerably more reliable and
efficient than earlier models.
On the basis of reasonable assumptions regarding the life of motor/pump
units, the unit water cost from PV pumps is found to be competitive with the unit
water cost from diesel pumping systems in remote areas for applications where the
volume-head products (daily volume required in m3times the total pumping head in
m) is less than about 1000 m4. Wind pumps would probably be cheaper than PV
pumps if the mean wind speed in the periods of maximum water demand is at least
2.6 m/s. Even though the unit water cost for PV pumps may be cheaper than for
77
diesel pumps in certain circumstances, this does not necessarily imply that PV
pumps would be an economic solution for irrigation applications, since the
economic market price of the crop has to be considered. Nevertheless, PV pumps
combined with trickle irrigation or other low-water-use irrigation techniques, when
used for fruit and other high value crops, may be found to be economic. Finance in
the form of capital grants and low-cost loans will be needed to bring high capital
costAow running cost equipment within reach of the potential users.
PV water pumps have found wide social acceptance, particularly in villages
which previously had to pump water by hand. The full involvement of the end-users
from the planning stage onwards has been found to be the key to successful
implementation of PV pumping systems. In some cases, a village co-operative has
been formed to administer the local aspects, such as fund raising to meet local
expenses and the levying of water charges on users. A central organization, usually
government-backed, is essential however to support the users with finance and
technical advice. Training of the users is needed to enable them to obtain the best
results and to adapt local practices as appropriate. They also need to be trained in
routine maintenance and trouble-shooting. Simple faults can often be repaired by
unskihed operators, with the central support organization providing help for more
serious problems and obtaining spare parts.
Vaccine refkigerators
As a result of the co-ordinated development and field testing programmes
sponsored by WHO, PV vaccine refrigerators are now available which perform
reliably and are cost-effective for use in areas without grid electricity. The key
technical factor is the sizing of the PV array and the battery, which requires
knowledge of the climatic conditions and the loads likely to be experienced by the
system at the proposed site.
Detailed technical requirements for PV refrigerator/freezers intended for
vaccine storage as part of the ‘cold chain’ and other recommendations for
procurement are set out in guidelines issued by the WHO. At present, nine
commercial ,systems have been approved for use in the ‘cold chain’ and several
other systems are being evaluated.
The average annual cost of a PV refrigerator is comparable to that for a
kerosene refrigerator. PV refrigerators are more cost-effective, however, since they
are more reliable, which results in a lower cost per dose. Since the refrigerator cost
per dose is relatively small compared with the immunization programme overhead
cost per dose, the use of the more efficient refrigerator is well justified.
Provided the PV refrigerator is properly designed, installed and maintained, it
should give a reliability of at least 90 per cent (ie. only 10per cent of vaccines lost),
compared with reliabilities as low as 50 per cent for kerosene refrigerators.
Immunization programmes can be seriously disrupted by vaccine losses, with
courses of immunization incomplete, medical staff frustrated and loss of
momentum of the programme as a whole.
The users of PV vaccine refrigerators need to be trained in the correct
operation and maintenance of the system. As for PV water pumps, there needs to be
a central organization available for users to obtain technical support, advice and
spare parts.
Lighting
Small PV systems for domestic lighting are widely available and several thousands
have been installed, particularly in the South Pacific, French Polynesia and China.
They are simple to operate and reliable, now that earlier problems with battery
78
charge regulators have been solved. PV-powered fluorescent tubes provide a much
higher quality of light compared with candles or kerosene wick or pressure lamps.
The efficiency of the DC ballasts used for fluorescent tube lamps is a key technical
factor in the design of systems.
Larger PV lighting systems are also available for street lighting and security
lighting. Standard systems using fluorescent tubes are available from most PV
manufacturers. Again the key technical factor has been the efficiency of the DC
ballasts. Specially designed AC systems for high-mast lighting have also been
demonstrated. The key factor here is the performance of the inverter.
Based on analysis over five years with 10 per cent interest rate, PV lighting
systems are found to be cost-competitive with kerosene lamps in areas where the
cost of kerosene is $0.75/litre or more. Thus, given suitable fmance, users would be
able to save money and improve the quality of domestic lighting by installing PV
lighting. Several organizations are operating successful financing schemes which
provide a capital grant plus a loan for the balance repaid over five years at 9 or 10
per cent interest.
It is essential that in addition to technical advice and the provision of fmance,
users of PV lights have access to spare parts, particularly replacement fluorescent
tubes and batteries. Some form of local co-operative, backed-up by a central
organization, is likely to be the best approach in most cases.
Rural electrikcation
Rural electrification differs from lighting in that it is a more systematic approach to
providing an electricity supply to a village or district. The electricity is then
available for a wide range of domestic, commercial and agro-industrial
applications. A number of relatively large central PV demonstration plants have
been built to electrify a complete village, but there are a number of technical and
institutional problems with this approach. Central systems of this type are
vulnerable to failure due to component faults or over-loading and thus need a high
level of supervision.
A more viable approach to rural electrification is to equip each household with
its own stand-alone system. A number of standard sizes would be required to suit
the size of the household and the nature of the loads likely to be imposed. Larger
systems would be needed for commercial, institutional and industrial premises.
The electricity produced for general applications, such as lighting, fans,
refrigerators, radios and televisions, needs to be AC, at the national voltage and
frequency. This means that the PV systems have to incorporate inverters, which
have not proved particularly efficient or reliable in the past. However, several
manufacturers are now introducing new inverter types which offer much improved
performance. At least one type switches itself into stand-by mode when no loads
are connected, thereby greatly reducing the internal system losses. Initial field
experience with these inverters is encouraging.
PV systems for rural electrification (as opposed to simple lighting applications
using DC output) are too expensive at current costs for large-scale implementation.
In certain circumstances, where the cost and other difficulties involved with diesel
generators are too great, PV systems can be cheaper, but users would not in general
be able to afford to pay the true cost, even if long-term finance were available.
Substantial subsidies, comparable to the subsidies already provided for rural
electrification schemes by grid extension or diesel generation, will help bring costs
within the reach of consumers. However, PV system costs need to come down to
about 50 per cent of present levels for this technology to be an acceptable
alternative to conventional techniques.
79
Any scheme for rural electrification using PV systems will need to be planned
to provide for public information and other forms of institutional support. This
would probably be best organized within the existing rural electrification board, to
cover all technical, administrative and fmancial aspects.
Agricultural applications
Stand-alone PV systems can be used for a number of agricultural applications
which require mechanical power. Besides water pumping for irrigation, grain
milling is probably the most important agricultural use. Although not many PVpowered grain milling systems have been built, there is no reason to doubt the
technical feasibility of this application. One system has been operating in Burkina
Faso with high reliability for some six years. The type of mill has to be selected to
suit the type of gram and the degree of fineness required. The mill may then be
driven by a DC motor supplied by a battery charged by a PV array. Care must be
taken to ensure that all components are correctly matched to suit the solar input as
it varies throughout the year, and the load imposed. Such applications are likely to
be economic in comparison with diesel-powered systems in remote areas,provided
the demand is reasonably uniform throughout the year.
Other PV applications for agriculture include milking machinery, refrigerated
storage for perishable produce and ice production. Experience with these
applications is too limited for any general conclusions to be made. One other
application has however found wide acceptability technically and economically,
and that is PV-powered cattle fencing. These electric fences can be used not only to
keep domestic livestock within required limits, but also to keep wild animals out of
areas where they could damage crops.
Water treatment
In addition to PV pumps for water supplies, PV-powered complete water treatment
plants are being developed for use in remote areas. The emphasis is on reliability
and minimum requirements for chemicals and other consumables. Several
approaches have been demonstrated, iucluding systems involving slow sand
ftitration, filtration followed by a prophylactic dose of sodium hypochlorite and
filtration followed by UV sterilization. Experience is too limited for any geikeral
conclusions to be made regarding the best approach. Costs are likely to be
competitive with any alternatives involving diesel generators.
Telecommunications
PV systems for telecommunications are finding wide acceptance throughout the
industrialized and developing world, since they offer a reliable and cost-effective
means of providing relatively small amounts of power in remote locations.
Essentially, the PV array charges a battery which suppiies conventional
telecommunications equipment. Hybrid systems, involving back-up diesel
generators, are often found to be the most cost-effective solution for larger systems
(ie. daily demand greater than about 2.6 kWh).
There are three types of telecommunication systems which can be PV
powered:
- Two-way radios and radiotelephones
-Radio and television secondary (infill) transmitters
- Telephone systems, including exchanges, repeater stations and satellite ground
stations.
In addition, television sets can have a dedicated PV power supply, an
80
application which has found wide use in certain West African countries for use in
village schools.
Cathodic protection
PV generators for cathodic protection systems used for pipelines and other steel
structures are commercially competitive in areas remote from the electricity grid.
There are many examples of such systems used for oil and gas pipelines in the
Middle East. This is an application where the DC output from the PV array can be
directly used.
Other applications
PV systems can be used wherever there is a requirement for small amounts of
power in a remote location. Onemanufacturer in Europe has recently developed a
mobile orthopaedic clinic powered by a PV generator to enable a full range of
equipment to be used in areas which have no electricity supply. Hazard warning
lights on tall structures, navigation lights at harbour entrances and lighthouses are
some examples of suitable applications for photovoltaics. Remote metering
stations, such as river gauging stations and meteorological recording stations, can
also be PV powered economically. New applications are constantly being found as
the potential of the new technology becomes more widely appreciated.
8.2
The technology
In general, the photovoltaic modules and arrays have performed reliably with very
few reports of failure or significant degradation. The experience to date has been
solely with crystalline silicon solar cells, both mono-crystalline and semicrystalline. This is because most manufacturers manufacture modules which meet
the qualification and performance requirements laid down by the Jet Propulsion
Laboratory (USA) or the CEC’s Joint Research Centre (Italy). The performance of
the newer types of thin film amorphous silicon solar cells, which are beginning to
become available for power applications, has yet to be evaluated.
Many early PV systems (ie. prior to about 1982) suffered problems with power
conditioning and control systems, such as voltage regulators and maximum power
point trackers. This was due to a number of factors, including inadequate weather
protection, over-complex design and not being sufficiently robust, both physically
and electrically. There is growing evidence that the power conditioning and control
equipment now being supplied for use with PV systems is much more reliable, but
further development to improve the efficiency of some devices, particularly
inverters, is required.
For systems that incorporate batteries, it is vital that the correct type of battery
is chosen and that the sizing calculation takes full account of actual operating
conditions. Automotive batteries are in general not suitable, as they have limited
life when subjected to many charge/recharge cycles, A number of battery
manufacturers have developed special batteries for PV applications which offer
long life and low internal losses and require little or no maintenance.
The end-use devices, such as motor/pump units and refrigerators, have
generally had to be specially developed for PV applications. Many problems
developed with the earlier systems, often due not so much to faults in the overall
concept, but to mistakes in the matching of components, faulty operation and
inadequate quality control. New products with better performance, reliability and
durability have become available in recent years for all applications of main
interest in developing countries.
81
There has been extensive work to test and evaluate PV pumping systems and
PV vaccine refrigerators. There is a continuing need to update the results of
previous projects and to keep potential customers informed of the state-of-the-art.
Similar testing and evaluation programmes are needed for other PV systems, such
as lighting systems, electrification systems suitable for households and
institutional buildings (health centres, police posts, schools, etc), and systems for
agricultural product processing such as grain mills.
8.3
The economics
It is not possible to generalize about the economic viability of PV systems. Each
application has to be considered on its merits, taking into account local conditions
and the cost of ahernatives. Although PV systems have high initial cost, they require
no fuel and little maintenance and should last many years. In many remote areas,
diesel generators, the main alternative to PV generators, would be impracticable
due to fuel supply costs and uncertainties, plus the problems associated with
maintenance and the supply of spare parts. Even if diesel generators appear to be
cheaper on a life-cycle cost comparison, it might be preferable to go for a PV system
because of the operational advantages.
For some applications, PV systems are widely found to be competitive with the
alternatives. For example, PV refrigerators for vaccines offer a lower cost per dose
than kerosene refrigerators and enable a more effective immunization programme
to be mounted. PV systems for domestic lighting are also competitive with
kerosene lamps and candles and moreover give a much better light.
PV water pumps may be cost-competitive with diesel pumps for applications
where the flow is low (less than 50 mYday) and the head is low to medium (less than
20m). The analysis is sensitive to the cost of diesel fuel and the life-expectancy of
the diesel pump. PV pumps are more likely to be viable for village water supply or
livestock water supply applications, where the social benefits are high, than for
irrigation applications, where the extra value of the crop made possible by the
irrigation water has to exceed the cost of the pump.
PV systems for telecommunications and cathodic protection are often found
to be the cheapest option when small amounts of power (up to about 10 kWh day)
are required at sites remote from public electricity supplies. Depending on local
circumstances, PV systems may also be the cheapest alternative for hazard warning
lights, remote metering stations, navigation buoys and lighthouses.
8.4
Social and institutional
factors
Experience has shown that PV systems are generally widely welcomed by the
users, provided they have been involved at an early stage in the planning process
and have been given basic instruction in how to operate the system and carry out
routine maintenance. Problems arise when a system is set down in the field by a
research organization and the local people are expected to use it without any real
appreciation of what is involved and without proper arrangements for follow-up.
Unlike diesel systems or grid-supplied electricity, PV systems are very site-and
load-dependent. Therefore considerable experience and technical expertise are
needed at the design and procurement stage in order to ensure the system will fulfil
its expectations. A central organization within the country concerned needs to be
established with the necessary skills in-house to undertake the necessary design
tasks and write system specifications in preparation for tendering procedures. The
same central organization can arrange for user training and technical support in the
operation phase. The supply of spare parts, particularly where these have to be
82
imported, is an important function of this central organization.
There are four important institutional factors that contribute to the success of
a PV application:
(a) Genuine involveme& by the users from the planning stage onwards, with
appropriate arrangement for a co-operative or other method of administering
the local aspects (operation, maintenance, fund raising, user charges, etc)
(b) Technical expertise at the system design and procurement stage to ensure that
the system is compatible with local needs, solar resources and other technical
constraints
(c) Finance to meet the majority of the initial costs, after allowing for any
government subsidies considered appropriate, with repayments spread over at
least five years at a preferential interest rate
(d) Arrangements for user training and technical support after installation, to deal
with faults that cannot be fixed locally and the supply of spare parts.
Although most developing countries have already established a department to
undertake the necessary supporting activities as outlined above, many need
technical assistance and finance from outside to help them identify and implement
appropriate photovoltaic projects.
83
9. RECOMMENDATIONS
9.1 Identification
of appropriate
applications
Photovoltaic applications need to be evaluated taking into account all relevant
technical, economic, social and institutional factors. The results of this evaulation
then form the basis for subsequent decision making. The technical evaluation will
need to include consideration of the following:
@ operating performance of the system as a whole in relation to demand and solar
resource variations with time;
0 operating performance of each main component with a view to identifying
possible improvements;
0 reliability, availability and durability;
0 ease of operation, maintenance and repair.
The economic evaluation will need to take into account:
@ initial costs for procurement, transportation to site and installation, including
any local civil engineering costs, subdivided into foreign exchange and local
currency costs;
0 labour costs for operation and maintenance;
l expected costs of spare parts and replacement of components which wear out;
0 life cycle cost-benefit economic analysis, based on appropriate discount rate;
l unit cost comparisons with alternative energy sources;
0 financial analysis from user’s viewpoint, taking into account subsidies,loans,
interest rates and repayment periods.
The evaluation of social and institutional factors will need to take into account
the following:
0 the actual demand for the energy and/or product from the system;
@ the social and environmental context in which the system will operate;
0 the level of operator skills available and associated training needs;
@ the appropriate form of local organization needed to administer the system;
e ,the appropriate form of central organization needed to provide technical and
financial support.
Only when satisfactory answers to each part of the evaluation are obtained
should the project proceed to the next stage in the decision-making process.
9.2 Strategic
approach to development
A strategic approach to development is essential if the limited resources of finance
and technical skills are to be utilised in the most effective ways. There are many
competing demands on national resources and national policies have to be
developed which balance these demands in the manner considered most
appropriate in the national interest.
Photovoltaic systems can contribute to development in a number of sectors.
Village water supplies and lighting systems can significantly improve the standard
of living of the rural population, thereby raising morale and helping to counteract
the drift of young people to the towns. Irrigation pumps can increase agricultural
output, thereby providing more food for home consumption or for export, as well
as raising rural incomes. More reliable vaccine refrigerators can improve the
effectiveness of immunization programmes and thereby reduce infant mortality.
Improved and extended telecommunication systems can contribute significantly to
raising efficiency in all sectors of national life.
PV projects must be considered alongside other development projects and
ranked in order of the net benefits likely to accrue to the nation as a whole, taking
into account both economic and social benefits. The effects the project would have
on matters such as employment, food production, balance of payments and
development of national skills and self-reliance need to be assessed. In this
connection, the possibilities for local manufacture and assembly of as much of the
PV system as practicable need to be considered.
These issues are not easy and straightforward to evaluate. They involve many
factors that are hard to quantify, but it is important to appreciate that the viability of
PV projects should not be assessedsimply in terms of economics or major energy
substitution. For example, the energy requirements for telecommunications are
relatively small but the benefits that can result from installing reliable PV
generators in place of unreliable diesel generators can be of very great significance.
9.3 Staged development
Solar energy activities, in common with all work in the field of appropriate
technology, need to be planned with the ultimate objective of introducing suitable
systems into the community on a commercial basis, either with or without
government subsidies. The full development sequence should therefore be planned
along the following lines:
Stage1 -Research
to identify the basic physical principles, materials and
designs; most of this activity in relation to photovoltaics has been (and will
continue to be) carried out in the industrialized countries, as it involves high
technology equipment for making and testing solar cells and other components;
Stage2 -Laboratory-based development work to adapt system designs to suit
local materials and needs, to characterize performance under local conditions and
to identify how performance can be improved; such work is very useful for training
professional and technical personnel in the principles of solar engineering and test
methods;
Stage3 -Field
work on pilot plants, to demonstrate and test representative
systems under realistic field conditions, preferably when used by local people such
as villagers or farmers; at this stage, it is important to monitor the systems in detail
and to involve potential industrial companies who may be interested in subsequent
commercial development;
Stage4 -Full-scale
demonstration plants, with prototype commercial units
installed at a number of representative sites throughout the country, with the fuIl
involvement of the industrial interests and continued technical support from the
research institution responsible for the original development.
Stage6-
Commercialization, with local manufacture/assembly of systems and
associated marketing and follow-up activities. In addition to the commercial
suppliers of hardware, a separate independent organization is essential for several
86
years to promote schemes and to provide technical support to the users.
Usually financial support from the government is required for stages 1 and 2.
Costs for stages 3 and 4 are often shared between the government and commercial
interests, with possibly some contribution from the users. For Stage 5, government
subsidies and finance for low-interest loans are often needed to bring high initial
cost systems within the reach of potential users in rural areas.
Taking into account the experience now available from many countries, the
photovoltaic applications of main interest in developing countries are at the
following stages of development:
- water pumps
Stages 4 and 5
- vaccine refrigerators
Stages 3,4 and 5
- lighting systems
Stages 4 and 5
rural electrification
Stages 3 and 4
-(central plants):
rural electrification
-(household systems):
Stages 4 and 6
- agricultural systems
-grainmills
Stage 3
- cattle fences
Stage 5
- milking machines
Stage 3
- cold stores
Stage 3
- telecommmunications:
Stage 5
- cathodic protection:
Stage 6
- hazard warning lights:
Stage 5
- lighthouses:
Stage 4
Each developing country needs to assessits own needs and institute a staged
development for each PV application of interest, taking into account the status of
development reached in similar situations elsewhere. An essential requirement is
to build up the necessary institutional support with the skills and finance needed to
implement the programmz.
87
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement