Renewable Energy 61 (2014) 43e50

Contents lists available at SciVerse ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Low head pico hydro turbine selection using a multi-criteria analysis

S.J. Williamson a, *, B.H. Stark a, J.D. Booker b
a
b

Department of Electrical & Electronic Engineering, University of Bristol, University Walk, Bristol, BS8 1TR, UK
Department of Mechanical Engineering, University of Bristol, Bristol, UK

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 30 November 2011
Accepted 11 June 2012
Available online 7 July 2012

Turbine types suit specic ranges of head, ow rate and shaft speed and are usually categorised by
specic speed. In the pico range, under 5 kW, the requirements are often different to that of larger scale
turbines and qualitative requirements become more inuential in selection. Pico hydro turbines can be
applied beyond these conventional application domains, for example at reduced heads, by using nontraditional components such as low speed generators. This paper describes a method to select which
turbine architecture is most appropriate for a low-head pico hydro specication using quantitative and
qualitative analyses of 13 turbine system architectures found in the literature. Quantitative and qualitative selection criteria are determined from the particular requirements of the end user. The individual
scores from this analysis are weighted based on the perceived relative importance of each of the criteria
against the original specication and selects a turbine variant based on the total weighted score. This
methodology is applied to an example of a remote site, low head and variable ow requirement, leading
to the selection of a propeller turbine variant or single-jet Turgo turbine for this specication.
2012 Elsevier Ltd. All rights reserved.

Keywords:
Pico hydro
Turbine selection
Low head
Application range
Turgo

1. Introduction
There is a distinct link between poverty and access to modern
energy [1]. With electricity, people are able to improve their
productivity through the better use of their time, and so their
income, which allows them to raise themselves out of poverty.
Where electricity is not available many people use kerosene lamps
for lighting. These lamps have health issues, pose signicant safety
risks, provide only a dim and inefcient light source, and can take
a signicant portion of the monthly income for a family [2]. In
urban areas the percentage of the population with access to electricity is high, due to the low cost of connecting them to the grid. In
rural locations, however, access is limited due to the high cost of
extending grids to low density population centres [3]. Pico hydropower is able to provide rural electrication where grid extension is
too costly and consumers have low incomes [4,5]. In a World Bank
report it was shown that pico hydropower represented the
cheapest opportunity for off-grid generation under 5 kW in 2005
and was projected to be at least 25% cheaper than the nearest
alternative still after ten years [6].
The typical turbine solutions for these pico hydropower systems
are either pumps as turbines (PAT) [7], locally made pelton wheels

* Corresponding author. Tel.: 44 (0)117 954 5499.

E-mail address: sam.williamson@bristol.ac.uk (S.J. Williamson).
0960-1481/$ e see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.renene.2012.06.020

[8], mass manufactured propeller systems [9] or home made

systems normally based on impulse turbines [10]. Selecting hydro
turbines is traditionally based on the specic speed of the turbine,
a pseudo non-dimensional parameter that includes head, output
power and output shaft speed [11,12]. From this, the commonly
used application domain for turbines is used to aid selection, as in
Fig. 1, which has been compiled from [13e15]. There are alternatives such as the nomogram presented in [16], but still based on the
same principle. This leads to the choice of Pelton and Turgo
turbines at high heads, crossow and radial (Francis) turbines at
mid heads and propeller turbines and waterwheels at low heads.
This is also reected in the commercially available turbines for
these heads. As can be seen in Fig. 1, the pico range, under 5 kW
generation, appears to be sparsely covered by reported application
domains. There are several commercially available pico hydro
products at high, mid and low head, and these tend to follow the
topology of the larger scale turbines. The low head sector of the
market is dominated by propeller type turbines as shown by the
three commercially available systems that are superimposed onto
Fig. 1, images of which can be seen in Fig. 2. All of these turbines are
propeller based in either closed ume, Fig. 2a, or open ume
design, Fig. 2b and c, and are recommended to have draft tubes
installed to increase the turbine efciency.
The use of traditional 4- or 6-pole generators and direct
generatoregrid interfaces, as described in most literature such as
[20] restricts the application domains for pico hydro turbines. For

44

S.J. Williamson et al. / Renewable Energy 61 (2014) 43e50

Nomenclature
D
H
Hg
Hl1
Hl2
Hl3
Hl4
Hl5
Hl6
P
Q
T
g
hl6
n
r

Dvw
h
r
u

Diameter of Archimedes Screw (m)

Head (m)
Gross Head (m)
Head Loss due to Penstock (m)
Head Loss due to Draft Tube (m)
Head Loss due to Kinetic Energy of Outow (m)
Head Loss due to Water Entering Waterwheel (m)
Head Loss due to Swirl of Water Exiting Waterwheel
(m)
Head Loss due to Friction of the Flow on the Water
Bed for Undershot Waterwheel (m)
Power (W)
Volume Flow Rate (m3/s)
Torque (Nm)
Gravitational constant (m/s2)
Specic Head Loss due to Friction of the Flow on the
Water Bed for Undershot Waterwheel
Geometrical factor for Archimedes Screw
Average Impact Radius (m)
Change in Whirl Velocity (m/s)
System Efciency
Water Density (kg/m3)
Rotational Speed (rad/s)

example, the specic speed as dened in [16] for a turbine specication to produce 1 kW at between 1 and 3.5 m head with a 4- or
6-pole directly connected generator is 251e1800, suggesting
a radial or axial ow reaction type turbine topology. However,
introducing technologies such as low speed generators or inverter
based grid interfaces generally extends a turbines application
domain, leading to other viable turbine solutions. Replacing the
traditional 4- or 6-pole generator with a low speed generator
operating at 200 rpm such as [21], the specic speed is now
50e240. This greatly expands the choice of turbines available to
include Crossow, Turgo and multiple-jet Pelton turbines. In
addition to these selection criteria, the requirements on a pico
hydro turbine tend to be different to those of a larger scale turbine;
pico hydro generators require off-the-shelf solution as a unique
product for each site would be too expensive for the target users.
They may be located in remote locations several hours walk from

the nearest road and have no skilled labour locally to operate and
maintain the system. The application domain selection method for
turbines does not take these more qualitative factors into account.
The methodology proposed in this paper is used to select a pico
hydro turbine for a low head specication using both quantitative
and qualitative criteria.
2. Turbine selection methodology
In the design of complex systems, the evaluation and selection
of candidate solutions can be facilitated using a Pugh Matrix [22,23]
in the early stages of a project. This approach can be subjective and
still necessitates a great deal of design analysis and detailed design
work in order to create an optimal solution. The traditional
approach to concept design has been the subject of a great deal of
adaptation and improvement recently through reviewing its
application on a number of electro-mechanical machine research
projects. The motivation is to demonstrate that you can get closer to
the nal design sooner by placing more effort on a knowledge rich
and systematic approach, thereby reduce design iterations and
mitigating costly design changes; helping deliver novel and efcient solutions more rapidly [24]. Specically, concept selection
may be enhanced by the inclusion of quantitative performance
metrics predicted using simple physics-based models, alongside
the more traditional qualitative criteria derived from the specication. These performance metrics provide a measureable and
tangible way of guiding and tracking overall design performance
and evolution against targets or benchmarks without committing
a large proportion of development costs in prototype testing, and
provide crucial decision-making information. The use of a multicriteria selection methodology has been discussed and developed
to identify the appropriate renewable energy sources for a site
[25,26], however neither of these go into the more detailed selection or design of the renewable source.
A ow chart of the methodology derived for pico hydro turbine
selection is shown below in Fig. 3. Each block in this chart is discussed in more detail in Sections 2.1e2.8.
2.1. Specication
Each turbine system will have a set of requirements and specication, which is developed from discussions with the project
stakeholders. This will include either site conditions, such as head
and ow rate, or output power requirements. There will be environmental requirements, for example the site may be in an inaccessible location, be subject to extremes in temperature or have to
comply with shery regulations. The turbine may be able to have
regular maintenance checks from an onsite operator, or it may be
required to be operated remotely and therefore should require
minimal maintenance and have a high reliability.
2.2. Selection criteria
Using the specication and derived requirements, a set of selection
criteria can then be developed. Table 1 shows some of the possible
selection criteria that could be used, and divides them up into
quantitative and qualitative criteria. The assignments are not denitive, as the some criteria may be either quantitative or qualitative.
2.3. Quantitative analysis

Fig. 1. Typical turbine application range chart adapted from data in [13e15] populated
with three commercially available low head pico hydropower systems: a. Nepal Hydro
and Electric Ltd. PT1-Mk2 [17], b. ECO-Axial ZD [18] c. Powerpal MHG-1000LH [19].

Basic uid ow equations are used to derive simple performance

characteristics about the turbine option for the quantitative analysis. The performance variables are turbine power P, overall turbine
system efciency h, ow rate Q, and gross head Hg, two of which

S.J. Williamson et al. / Renewable Energy 61 (2014) 43e50

45

used to analyse the impulse turbines [11]. The head loss in the
penstock Hl1 and the loss in the nozzle Cv reduces the inlet jet
velocity. Using velocity triangles, the change in whirl velocity Dvw is
used to calculate the rate of change in momentum of the water,
generating a force at the blade. The force is concentrated at the jet
impact point, radius r, assuming the ow enters and exits at the
same radius, which causes a torque T on the wheel. The power is the
product of the rotational speed at maximum power u by the torque:

P T u Q rDvw r u

Fig. 2. a. Nepal Hydro and Electric Ltd. PT series low head turbine b. ECO-Axial ZD [18]
c. Powerpals MHG-1000LH [19].

will need to be dened, leaving two unknowns. These variables are

combined using Eq. (1) [11] for a general turbine system:

P hrgQHg

(1)

The two unknowns are solved using a second equation, which is

derived from further analysis of the turbine torque generation
mechanism. This analysis may result in the power available at
a specic site and the turbine system efciency or the efciency and
ow rate required to produce a specied power depending on the
variables dened.
There are 13 turbine types commonly used which are divided
into 4 categories: Impulse, Reaction, Archimedes Screw, and
Waterwheel. These are shown in Fig. 4. Each category has
a different torque generation mechanism and is analysed in
a different manner, as summarised in the following sections. The
performance modelling used here is simplied, neglecting uid
mechanic non-linearities, and assumes linear geometric scaling. For
some turbines considered, impractical geometries may be generated at the extremes of their head range. The results from the
quantitative analysis are then normalised against the maximum
value for that criteria to give a criteria score for the turbine.
2.3.1. Impulse type e pelton and turgo (single- and multiple-jet),
crossow
For the impulse turbines, as illustrated in Fig. 4aec, velocity
triangles of the water jet impacting with the blade of the turbine are

(2)

2.3.2. Reaction type e axial, radial

For the reaction turbine designs, as shown in Fig. 4deg, the
gross head Hg in Eq. (1) is reduced by head losses in the penstock
Hl1, the draft tube Hl2, and from the outow of kinetic energy from
the system Hl3. These head losses are functions of the speed of the
water passing through the component and a factor dependent on
the geometry of the component [11]. The power generated is then
dependent on the net head and the estimated turbine hydraulic
efciency ht, which is different to the overall turbine system efciency h in Eq. (1):

P ht rgQ Hg  Hl1  Hl2  Hl3

(3)

2.3.3. Archimedes screw

The Archimedes screw, as shown in Fig. 4h, operates on
a hydrostatic pressure difference across the blades [27]. The overall
turbine system efciency h is a function of the geometry of the
screw n, the diameter D, and the ow passing through it Q:

2n 1
2n 2


1

0:01125D2
Q


(4)

2.3.4. Waterwheels e overshot, breastshot, undershot

The analysis for the waterwheels is based on the losses in the
system, as described in [28]. For the Overshot and Breastshot
Waterwheels, Fig. 4i and j it is assumed that the losses are only the
kinetic energy loss from the water entering the wheel Hl4, and the
swirl in the water on the exit of the wheel Hl5, and the overall
turbine system efciency is then:

Hg  Hl4  Hl5
Hg


(5)

For an Undershot Waterwheel, Fig. 4k, two efciency losses are

considered. Firstly the inlet efciency hth represents the non-ideal
ow entrance due to xed wheel geometry, secondly the friction
on the water bed hl6 ( Hl6/Hg), which models friction loss as
a function of inlet water velocity. The overall turbine system efciency is then:

h hth  hl6

(6)

Table 1
Range of selection criteria.
Quantitative criteria

Qualitative criteria

Rated ow/head efciency

Environmental e regulatory,
weather, location
Required civil works
Portability
Maintainability and serviceability
Reliability

Part ow/head efciency

Cost
Turbine rotational speed
Power for given site or
required site conditions
Size of system
Fig. 3. Turbine selection methodology.

Ease of manufacture
Design modularity

46

S.J. Williamson et al. / Renewable Energy 61 (2014) 43e50

Fig. 4. Turbine types a. Pelton/Turgo single-jet b. Pelton/Turgo multiple-jet c. Crossow d. Propeller turbine e no draft tube e. Propeller turbine e draft tube. f. Radial ow turbine e
no draft tube g. Radial ow turbine e draft tube h. Archimedes screw. i. Overshot waterwheel j. Breastshot waterwheel k. Undershot waterwheel.

2.4. Qualitative analysis

The qualitative analysis uses clearly dened criteria to score the
qualitative aspects of the turbine selection, as in [23]. Each criterion
is given an unambiguous denition and a dened scoring system
between 1 (poor) and 5 (excellent), typical extremes used in this
concept evaluation approach. Each turbine is then scored against
this scoring system as a part of a team discussion. The scoring
system which was developed as part of the research methodology
and used in Section 3 can be found in Appendix A as an example. A
scoring system is dened for each of the different selection criteria.
If within a criterion there are several different aspects, then sub
criteria are used to fully dene all the different aspects. These are
combined to form a single score for the criterion, through either an
arithmetic or weighted mean. The qualitative scores are then normalised against the maximum value.
2.5. Criteria weighting
A weighting for each of the criteria is decided by project
stakeholders, as recommended by [23]. These can be derived from
the specication and selection criteria, with contributions from the
project stakeholders once again. The sum total of the weightings
must equal one. Typically the quantitative criteria are prioritised
over the qualitative criteria as performance is an over-riding

consideration [29]. A set of example weightings are shown in

Table 4.
2.6. Candidate turbine choice
Using the results from the quantitative and qualitative analyses,
the normalised scores are then combined using the criteria
weightings to give a weighted average score for each turbine type.
This will rank the turbines and give the different options to choose
between.
2.7. Top level requirements
The top level requirements are based on opinion and real-life
non-assessable issues. Examples of these are management direction, the availability of materials and components or the
manufacturing capability of a company.
2.8. Final turbine selection
The candidate turbine choices are assessed with top level
requirements and solution chosen. The nal chosen solution may
not be top ranked candidate, as this may not t the top level
requirements.

S.J. Williamson et al. / Renewable Energy 61 (2014) 43e50

3. Example application of selection methodology

The village of Bhanbhane in central Nepal has several low head
sites for turbines. The heads at the sites vary from 0.5 m to 3.5 m,
the ow available at the sites also varies depending on the season
and the location in the river. There are two different rivers that
would supply the sites. The villagers would like to install the same
turbine in all of the locations, making savings in bulk buying and
allowing them to stockpile spares. At each turbine site, they do not
require more than 1 kW of electric power, so allowing for inefciencies in the system, such as drive and generator losses, the
turbine should produce 1.3 kW of mechanical power. Bhanbhane is
a rural village, and the sites lie several kilometres from the nearest
road, so the units need to be portable, ideally able to be carried by
villagers. Also, with the rural location and the distance from a road,
cement is expensive and difcult to obtain, therefore the civil
works should be minimised. The villagers intend to carry out onsite maintenance and servicing themselves, so the unit should be
simple to maintain and have a modular design allowing faulty
modules to be easily identied and replaced. Any faulty modules
that cannot be repaired onsite are to be returned to the manufacturer or a service centre for repair. It is assumed that the generator
output is 50 Hz, with a direct interface to the distribution system.
From this, the specication for the turbine is derived (Table 2), and
the selection criteria developed (Table 3).
In this example, power and head are the known variables, with
the efciency and ow rate required from the quantitative analysis.
The volume of the unit is estimated as a function of the ow rate
using existing designs and rules, for example [30]. The portability
and power are dealt with by deriving a power density metric (rated
power/volume) which provides the volume required to generate
1.3 kW. This volume includes the penstock, turbine and casing, as it
is assumed that all the other equipment required for the site, such
as the generator and control system, will be the same for each
design. The power density and rated ow efciency are thus treated
by the quantitative analysis, whilst the part ow/part head efciency, civil works, modularity and maintenance will be analysed
qualitatively. The scoring regimes for the qualitatively analysed
criteria can be found in Tables A1eA4 in Appendix A. The quantitative analysis designs a set of turbines for heads from 0.5 m to
3.5 m in 0.5 m steps. Power density is chosen to be the most
important criterion. This is due to the restricted access to the site,
a similar access route can be seen in Fig. 5, which is via a narrow,
steep path, and the equipment is required to be carried by the local
people. Therefore, if the unit is too large and unwieldy then the
villagers will not be able to implement it in their chosen sites. The
maintenance and serviceability and modularity in the design are
considered less important, which is reected in the weightings,
shown in Table 4.

47

Table 3
Selection criteria.
Efciency:
Power:
Portability:
Civil works:
Modularity:

Maintenance &
Serviceability:

Efciency of the unit at rated ow/head

and at part ow/part head
Power of the unit must be 1.3 kW
Minimised volume for easy transportation
Minimised civil works e concrete sparsely
available in site locations
Scope to incorporate modularity into the
design for line replaceable units and to
disassemble the unit for ease of portability
The ease of maintaining and servicing the unit,
especially with unskilled labour.

The turbines assessed represent the four main different

turbine types described in Section 2.2. Impulse and reaction
turbines require a penstock, which is assumed to have a 5% head
loss. The nozzles in the Pelton and Turgo turbines were assumed
to have coefcient of velocity for the nozzle of 0.98 [11]. The
multiple-jet turbines are assumed to have four jets which do not
interfere with each other. The extra jets give a volume penalty,
which is calculated as the volume of the extra pipe work, and
efciency penalty, assumed to be 10%. It is assumed that the
turbine is connected to a generator that can produce 50 Hz as long
as the rotational speed is between 200 and 3000 rpm (commercially available generators). If the rotational speed is less than or
greater than these limits, a gearbox is required to bring the speed
within these limits and so gearbox efciency is also taken into
account.
4. Results and discussion
For the quantitative analysis, the equations described in Section
2 are rearranged to calculate the efciency and ow rate required to
generate 1.3 kW, and from the ow rate the volume of each type
unit can be calculated. From this two equations can be formed using
the design inputs, one to derive the power density and the second
to calculate the efciency of the turbine. The outputs from the
analysis are shown in Fig. 6 for power density and Fig. 7 for the
efciency.
Each turbine is then assessed against the scoring regimes in
Appendix A and given a score between 1 and 5, which is normalised
against 5. All the criteria, quantitative and qualitative, are then
combined using a weighted average, the weightings detailed in
Table 4. The weighted average for each of the turbine designs is
shown in Fig. 8.
The breakdown of the scores can be seen from the analysis, to
see how the score builds up for each turbine type. This can be used
to see where each design has its strengths and weaknesses. An
example of the score breakdown is shown in Fig. 9. In this example,

Table 2
Turbine specication.
Power:
Head:
Portability:
Reliability:
Output frequency:
Maintenance:
Flow rate:
Modularity:
Civil works:

1.3 kW
Range from 0.5 to 3.5 m
Able to be transported to locations with
limited transport infrastructure
High reliability for low maintenance
operation
50 Hz output from generator
Maintenance and servicing carried out by
unskilled labour
Large variation across the seasons
Turbine in modules to allow for easy fault
identication and module replacement
Small civil works

Fig. 5. Accessibility issues at hydro sites in Nepal a. carrying transmission wire to

powerhouse b. access route to powerhouse.

48

S.J. Williamson et al. / Renewable Energy 61 (2014) 43e50

Table 4
Weightings used in example.
Criteria

Weighting

Power density
Full ow efciency
Part head/ow efciency
Civil works
Maintainability & serviceability
Scope for modularity

0.30
0.25
0.20
0.15
0.05
0.05

it is obvious to see that the single-jet Turgo is weak in the power

density (portability criteria) but is strong across all other criteria
compared to the reaction turbines.
Fig. 6 shows that the reaction turbines have a superior power
density. As the head increases, the power density of the impulse
turbines and waterwheels improves, whilst that of the Archimedes
Screw decreases. The conventional solution of adding multiple-jets
for Pelton and Turgo turbines at low heads is penalised in the
power density, as the extra pipe work required to feed the jets takes
up a large amount of volume. For the multiple-jet Pelton turbine, it
still provides a much better power density than the single-jet as it
requires a large diameter jet and so a very large turbine wheel. The
efciency of the single-jet impulse turbines is superior to all the
other turbines, as shown in Fig. 7. As the head increases then the
speed of the impulse turbine increases, so removing the need for
a gearbox, increasing the overall efciency of the turbine system
Fig. 7. Efciency variation over the head range 0.5e3.5 m, DT Draft tube.

further. Fig. 8 shows that a propeller turbine with a draft tube is the
most suitable solution between 0.5 and 1.5 m typical head, with the
single-jet Turgo turbine the best solution above 1.5 m typical head.
The propeller and radial turbines with draft tubes have a similar
weighted score to the single-jet Turgo turbine above 1.5 m head
and therefore are viable choices for the specication. The reaction
turbine result is expected, as many of the low head commercial
turbine systems available are propeller turbines with draft tubes.

Fig. 6. The output of the power density analysis over the head range 0.5e3.5m,
DT Draft Tube. a. full power density range b. zoomed in at low power density.

Fig. 8. The weighted scores for the 13 turbine choices over the head range 0.5e3.5 m,
DT Draft tube.

S.J. Williamson et al. / Renewable Energy 61 (2014) 43e50

49

project which will look at developing the network technology

requirements. Currently experimental work is being undertaken
which will be the subject of a future paper. This will present detailed
characterisations of efciency and torque speed curves over a range
of congurations validating this selection methodology.
Acknowledgements
This research is funded by Renishaw plc, and supported by
Engineers Without Borders (EWB), UK.
Appendix A. Qualitative selection criteria scoring regimes

Table A1
Civil works scoring regime.
Criteria e Civil works
Fig. 9. The weighted score at 3.5 m head for the 13 turbine choices with the contributions from each of the different selection criteria: 1 e Power density, 2 e Rated ow
efciency, 3 e Part ow efciency, 4 e Civil works, 5 e Maintainability and serviceability, 6 e Scope for modularity, DT Draft tube.

The surprising result in this analysis is the Turgo turbine which is

typically only used in medium to high heads e as indicated in the
application range graph in Fig. 1, and stated in several literature
resources such as [13]. However, as Harvey points out in [16], the
Pelton turbine, and therefore the Turgo turbine, can be used in low
heads if low speed and runner size do not pose problems. Although
the runner size is a consideration in this analysis it is shown that
the power density it is not too large. Using the calculations in the
previous section, the runner diameter for a single-jet Turgo turbine
generating 1.3 kW at 3.5 m head would be 435 mm, which would
therefore not pose a portability problem in rural areas.
At this point, with the choice of three main candidate turbines,
the top level requirements can be applied. For example, if off-theshelf, currently produced products are required then the
propeller turbine would the most appropriate. Alternatively, if
a manufacturer has more experience in building impulse turbines,
such as the high-head Pelton wheel, then the single-jet Turgo
solution would be appropriate.
5. Conclusions
In remote locations with low density population the cost of
extending the grid is very high. Kerosene lamps are common
where there is no electricity, which have many health risks as well
as increasing the possibility of re hazards. With the availability of
off-grid electrical power these lamps can be replaced by electric
bulbs, communication is possible through radios and mobile
phones and there are opportunity for income generation activities
such as grain processing plants and saw mills. If a location has
a river nearby, pico hydropower is a cheap way of providing this
off-grid electricity. There are many different options for a pico
hydropower system depending on the head and ow rate of the
water source.
This paper has presented a method of selecting low head pico
hydro turbines through a multi-criteria analysis, using the specication of the turbine to assess the turbine types through
quantitative and qualitative analyses. Using this method,
a propeller turbine with a draft tube or a single-jet Turgo turbine
has been shown to be the best solution for a given low head,
variable ow specication, depending on the nal top level
requirements.
This research is part of a project to develop a low head pico-hydro
off-grid network. The turbine selection is the rst phase of the

Denition e The amount of civil works that needs to be completed

to allow the scheme to run, broken down into intake, turbine
mounting, outow requirements and powerhouse. Powerhouse
not relevant as this will be the same for all locations.
Scoring criteria e Intake/penstock

Score

Open channel intake; no regulation mechanism; coarse trash rack;

simple/no support for penstock (if applicable)
Simple regulation mechanism; medium trash rack; some support
structure for penstock (if applicable)
Pressurised intake; complex regulation mechanism; ne trash
rack; large support structure for penstock (if applicable)

5
3
1

Scoring criteria e Turbine mounting

Score

Mounted on local material with minimal structure to support

the turbine unit
Mounted with a small amount of imported material and some
local material with a small structure to support the turbine
unit
Mounted with a large amount imported material with a large
structure required to support the turbine unit

5
3

Scoring criteria e Tailrace

Score

Water from turbine can exit directly to an open channel, no

requirement for outlet to be submerged; no tailrace structure
required
Simple tailrace structure required
Water outlet must be submerged; complex tailrace structure
required

3
1

Scoring criteria e Installation effort/skill

Score

No specialist tools required to install; unskilled labour able to

complete installation; short installation time
Some specialist tools required; simple skills required from labour;
reasonable installation time
Specialist tools required; highly skilled labour needed to install
devices; long installation time

5
3
1

Table A2
Part head/ow efciency scoring regime.
Criteria e Part head/ow efciency
Denition e How the efciency varies at part ow and
part head for conditions at various sites. This does
not include varying head during the year.
Scoring criteria e Part ow
Maintains maximum efciency e 10% for 80%
Maintains maximum efciency e 10% for 60%
Maintains maximum efciency e 10% for 40%
Maintains maximum efciency e 10% for 20%
Reaches maximum efciency during ow

Score
of
of
of
of

ow
ow
ow
ow

range
range
range
range

5
4
3
2
1

Scoring criteria e Part head

Score

No change/increase in efciency with change in head

Small decrease in efciency, no practical limits

5
4
(continued on next page)

50

S.J. Williamson et al. / Renewable Energy 61 (2014) 43e50

References

Table A2 (continued )
Criteria e Part head/ow efciency
Medium decrease in efciency, no practical limits
Large decrease in efciency, no practical limits or small
decrease in efciency with practical effects
Large decrease in efciency with practical limits

3
2
1

Scoring criteria e Adaptability to part ow/head

Score

None required
Simple automatic control system
Complex automatic control system
Manual intervention required
None possible

5
4
3
2
1

Table A3
Maintainability and serviceability scoring regime.
Maintainability & serviceability
Denition e The ability to include features in the design
or already incorporated that aid a long service life and
ease of maintenance/servicing.
Scoring criteria

Score

Incorporates all the desired design features for a highly

maintainable and serviceable unit
Incorporates some of the desired design features for a
highly maintainable and serviceable unit
Incorporates none of the desired design features for a
highly maintainable and serviceable unit

5
3
1

Maintainability and serviceability desired design features

Accessibility without having to remove other components
No specialist tools required for maintenance/servicing
Easy fault identication
Interchangeability of units
Minimal component layers, failing components located
in upper layers
Minimal connections between assemblies
Standard components and fasteners
Number of modules/components minimised
Proven designs
Isolating elements for safety in maintenance

Table A4
Scope for modularity scoring regime.
Scope for modularity
Denition e Modules that allow the system to be broken into
carryable/shipable units and allow line replaceable for
easy servicing and fault identication, with the ability to
interchange identical modules.
Scoring criteria e Modular portability

Score

Able to disassemble unit into manageable, man-carryable

components; simple and quick assembly/disassembly
Some components large and unable to be carried by single
man, most easy to carry; assembly/disassembly
reasonably simple and quick
Unable to disassemble unit into manageable, man-carryable
components; long and laborious assembly/disassembly

5
3

Scoring criteria e Modular unit

Score

Few, but standard, interfaces; few system elements;

simple coupling mechanisms between elements;
simple element architecture orientation
Few non-standard interfaces, some standard interfaces;
manageable architecture; some non-standard
coupling between system elements
Many non-standard interfaces; many separate system
elements; complex coupling mechanisms between
system elements; unusual element architecture
orientation

[1] Saghir J. Energy and poverty: myths, links, and policy issues. Energy and
mining sector board e energy working notes no. 4. World Bank. p. 1e24.
Available from: http://siteresources.worldbank.org/INTENERGY/Resources/
EnergyWorkingNotes_4.pdf; 2005 [accessed 29.11.11].
[2] Pode R. Solution to enhance the acceptability of solar-powered LED
lighting technology. Renewable and Sustainable Energy Reviews 2010;14:
1096e103.
[3] Byrne J, Zhou A, Shen B, Hughes K. Evaluating the potential of small-scale
renewable energy options to meet rural livelihoods needs: a GIS- and lifecycle cost-based assessment of Western Chinas options. Energy Policy 2007;
35:4391e401.
[4] Anyi M, Kirke B, Ali S. Remote community electrication in Sarawak, Malaysia.
Renewable Energy 2010;35:1609e13.
[5] Maher P, Smith NPA, Williams AA. Assessment of pico hydro as an
option for off-grid electrication in Kenya. Renewable Energy 2003;28:
1357e69.
[6] ESMAP. Technical and economic assessment of off-grid. ESMAP technical
paper 121/07. Mini-grid and Grid Electrication Technologies. Available from:
http://www.esmap.org/esmap/sites/esmap.org/les/Technical and Economic
Assessment of Off-grid, Minigrid and Grid Electrication Technologies_
Report 12107.pdf; December 2007 [accessed 29.11.11].
[7] Williams A. Pumps as turbines: a users guide. 2nd ed. London: ITDG
Publishing; 2003.
[8] Williams AA, Simpson R. Pico hydro e reducing technical risks for rural
electrication. Renewable Energy 2009;34:1986e91.
[9] Rijsenbeek W. Pico hydro systems in Vietnam, rural energy supply models
report. Available from: http://resum.ises.org/documents/PicoHydroVietnam.
pdf; 2001 [accessed 12.07.2011].
[10] Redeld S. Five gallon bucket hydroelectric generator build manual e version
1, appropriate infrastructure development group. Available from: http://
www.aidg.net/index.php?optioncom_remository&Itemid34&funcleinfo
&parentfolder&lecatid79 [accessed 29.11.11].
[11] Massey B. Mechanics of uids. 7th ed. Cheltenham: Stanley Thornes Ltd; 1998.
[12] Alexander KV, Giddens EP. Microhydro: cost-effective, modular systems for
low heads. Renewable Energy 2008;33:1379e91.
[13] Paish O. Micro-hydropower: status and prospects, proceedings of the institution of mechanical engineers e part A. Journal of Power and Energy 2002;
216:31e40.
[14] Muller G, Kauppert K. Old watermills e Britains new source of energy?
Proceedings of the ICE: Civil Engineering 2002;150:178e86.
[15] European Small Hydropower Association. A Laymans guidebook on
how to develop a small hydro site. 2nd ed.. Available from: http://
www.microhydropower.net/download/layman2.pdf;
1998
[accessed
29.11.11].
[16] Harvey A. Micro-hydro design manual. 1st ed. London: ITDG Publishing; 1993.
[17] Nepal Hydro and Electric Ltd., Butwal, Nepal, http://www.nhe.com.np
[accessed 29.11.11].
[18] Sunco Green Energy Ltd., Shandong, China, http://www.micro-hydro-power.
com [accessed 29.11.11].
[19] 2V Microsystems Ltd, Bedfordshire, UK, http://www.powerpal.co.uk [accessed
29.11.2011].
[20] N.P.A. Smith, Induction generators for stand-alone micro-hydro systems,
Proceedings of International Conference on Power Electronics, Drives and
Energy Systems for Industrial Growth, 2 (1996) 669e673.
[21] Seoyung Tech, Gumi, Korea. http://www.evsmotor.co.kr/eng/main.php
[accessed 29.11.11].
[22] Otto K, Wood K. Product design. 1st ed. New Jersey: Prentice Hall; 2001.
[23] Pugh S. Total design. 1st ed. Wokingham: Addison-Wesley; 1991.
[24] Constantinou P, Aird CJ, Mellor PH, Smith DJ, Booker JD, Truman CE,
et al. An energy supply unit for an autonomous remote sensor monitoring stored nuclear waste. Sensors and Actuators A: Physical 2011;166:
52e65.
[25] Polatidis H, Haralambopoulos DA, Munda G, Vreeker R. Selecting an appropriate multi-criteria decision analysis technique for renewable energy planning. Energy Sources: Part B 2006;1:181e93.
[26] Datta A, Ray A, Bhattacharya G, Saha H. Green energy sources (GES) selection
based on multi-criteria decision analysis (MCDA). International Journal of
Energy Sector Management 2011;5:271e86.
[27] Muller G, Senior J. Simplied theory of Archimedean screws. Journal of
Hydraulic Research 2009;47:666e9.
[28] Lea FC. Hydraulics for engineers and engineering students. 6th ed. London:
Edward Arnold; 1945.
[29] Andersson P. Early design phases and their role in designing for quality.
Journal of Engineering Design 1994;5:283e98.
[30] Kempes engineers yearbook. Tonbridge: CMP Information Limited; 2001.