Beruflich Dokumente
Kultur Dokumente
Energy
journal homepage: www.elsevier.com/locate/energy
a r t i c l e i n f o a b s t r a c t
Article history: Kathmandu University (KU) is one of the leading educational institutes in Nepal with a decade long
Received 20 January 2012 experience with R&D of micro-hydro turbines. With a technical collaboration with Norwegian University
Received in revised form of Science and Technology (NTNU), KU has also conducted several studies related to sediment erosion in
18 April 2012
hydro turbine.
Accepted 8 May 2012
Available online 9 June 2012
In the presented study a new program has been developed to create and optimize the design of Francis
runners. The program is also featured to compare erosion in runner blades for different design cases. The
nal design can be exported to Computational Fluid Dynamics (CFD) and Fluid Structure Interaction (FSI)
Keywords:
Sediment erosion
for further analysis. Parametric survey was carried out with this program to evaluate the relative effect of
Francis turbine each design parameter on sediment erosion. Several optimized designs were developed and analyzed
CFD with CFD tools to fulll the desired condition of minimum erosion and maximum efciency.
Design optimization This paper summaries the problem of sediment erosion of hydro turbines in South Asia region. Some of
the important achievements in R&D of hydro turbine and sediment erosion at KU are also presented. The
ndings of design optimization of Francis turbine for effective reduction in sediment erosion is discussed
in details.
2012 Elsevier Ltd. All rights reserved.
0360-5442/$ e see front matter 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2012.05.014
B.S. Thapa et al. / Energy 47 (2012) 62e69 63
Fig. 1. Sediment erosion in hydro turbines under basins of Himalaya: a. Damage in runner at Jhimruk PP after one year of operation [7]. b. Surface erosion Pelton runner at Khimti PP
[8]. c,d. Damage to guide vanes and cheek plates at Nathpa Jhakri PP [9].
Fig. 2. Research in miniature turbine laboratory at Kathmandu University: a. 800 W Propeller turbine designed and tested at Pico turbine test laboratory [10]. b. Flow visualization
in Pelton bucket at water power laboratory [11]. c. 130 kW Francis turbine ready for performance test.
of Pelton and Francis turbines and also provided professional in free air, hence this system could be considered to be similar to
trainings. Five different Pelton bucket proles designed at KU and Pelton turbine system.
manufactured locally were tested for impact and ow visualization Sediment samples were collected from several locations
(Fig. 2b) [11]. Francis turbine for 130 kW micro-hydro projects covering streambeds of some of the Nepalese rivers of different
(Fig. 2c) is under development phase at KU. river basins. Mineralogical analysis of samples was done to identify
Both miniatures laboratory are still in operation and are mostly quartz content and their shape in each river. The erosion tests were
utilized for students laboratory and academic research activities. carried out on turbine material 16Cr5Ni Martensitic Stainless steel
Recently a new and advanced turbine testing lab has been con- by 1 kg sand samples. The erosion rate obtained from the laboratory
structed at KU to promote its R&D activities in development of erosion test in same operating condition with different sand
hydro turbines [2].
70 30,0
60 25,0
volume)
50 20,0
40
30 15,0
20 10,0
10 5,0
0 0,0
Tamakoshi
Jhumruk
Dolal Ghat
Khimti khola
Madi Rever
Khimti
Kulekhani
Bagmati
Dhobi
Roshi
Phedi
Sunkoshi
Sapta koshi
Modi-2
Modi - 1
Tinau
Chitlang
Palung
Rapti Khola
Ganaha
Manahara
Dhad Khola
Gaur (
West seti
Arun khola
Aadhi khola
Fig. 4. Quartz content from mineralogical analysis and erosion rate from laboratory
erosion test [3].
samples between 425 and 300, 300e212 and less than 212 mm are
averaged and presented in Fig. 4 along with the corresponding
quartz contents in the sample [3].
Rotating Disc Apparatus (RDA) was developed at Kathmandu Fig. 7 shows the erosion pattern generated by combined effects of
University to study the sand erosion, cavitation and their combined sediment erosion and cavitation.
effect, as shown in Fig. 5. It consists of a rotating disc with four
cavitation inducers and driven by a 7.5 kW motor at 2880 rpm. The
4. Use of modern computational tools
motion of submerged body can be simulated up to 39 m/s velocity.
Test objective was to compare performance of High Velocity
KU has also been able to use the numerical tools and compu-
Oxygen Fuel (HVOF) coatings with stainless steel. This provided an
tational software for R&D of hydraulic turbines. Two important eld
opportunity for accelerated sand erosion testing for a comparison
of study employing the computational tools for R&D of hydraulic
of different materials. The disc is made up of Stainless Steel SS316
turbines at KU are characterization of sediment particles, and
and only half of this disc was coated with tungsten carbide (86% Co
design optimization of Francis turbine to minimize effects of sedi-
10% Cr 4%). The erosion damage in stainless steel and HVOF coating
ment erosion.
were compared and analyzed based on the area and pattern of
erosion [12]. Fig. 6 shows the test specimen with HVOF coating and
4.1. Characterization of sediment particles
Fig. 5. RDA at KU for cavitation and erosion study. Fig. 7. Erosion pattern with generated by sand erosion [12].
B.S. Thapa et al. / Energy 47 (2012) 62e69 65
plants and its digital images had been acquired. These shapes have
Fig. 10. Edge boundary of single sand.
further been analyzed by articial neural network. This network
has been rst trained for the known input and known output. After
that it is trained for unknown input and known output. Finally
these networks could recognize any shape given to it and gives the optimization of Francis runners for sediment handling. Developing
shape which is nearest to the seven predened shape [13]. results have shown the possibilities to reduce sediment erosion in
Figs. 8e10 shows the image processing steps used to characterize Francis runner signicantly by optimizing the hydraulic design
the sediment particles and Fig. 11 shows percentage of sediment alone.
particles with different predened shapes in one the hydropower A new program Khoj has been developed to create and opti-
plants in Nepal. mize the design of Francis runner. Khoj is also featured to compare
erosion in runner blades for different design cases. The nal design
can be exported to CFD and CAD tools for further analysis. Para-
5. Progress in design optimization of Francis runner for
metric survey was carried out with Khoj to evaluate the relative
sediment handling
effects of each design parameter on sediment erosion. The results
from Khoj were compared to that from CFD analysis to estimate
Hydropower projects in Nepal and the region have been facing
effects of the design variables on hydraulic performance. Several
the severe problem of turbine erosion due to sediment particles.
optimized designs were developed and analyzed to fulll the
Geographic conditions in Nepal makes high head Francis turbine
desired condition of erosion and efciency.
a better choice. However, Francis turbines are found to be more
sensitive to the effects of sediment erosion [14]. KU in close coop-
eration with NTNU has started a unique project for Design 5.1. Reference design
Fig. 11. Shape and size distribution of sediment particles in one of hydropower plants
Fig. 9. Cropped image of single sand. in Nepal [13].
66 B.S. Thapa et al. / Energy 47 (2012) 62e69
Table 1 Table 3
Basic design data for JHC. Parameters for CFX-Turbo Grid.
5.2. Design optimization range and Methodology Simulations are done to evaluate the hydraulic performance and
erosion on blade surface. Exactly same process has been repeated to
Optimized designs are developed by varying hydraulic design all the Design Analysis to maintain the consistency. Comparisons of
parameters in reference design to get the same power output. The results for optimized designs are done with that for reference
hydraulic design parameters are varied within a dened range and designs. Tables 3e6 presents the parameters selected for the CFD
its effects on erosion factor is evaluated. Table 2 lists the range of analysis.
variation of the design parameters considered for this study.
Following two terms are dened as the indicators and the
5.4. Results and discussion
means of comparison of relative erosion in the Francis turbine
runner.
The results of the parametric study from design program indi-
cate some modications in conventional design to reduce the
5.2.1. Erosion tendency (Et)
sediment erosion. However, each modication has some limita-
It is quantication of tendency of a specic design of runner to
tions that have to be considered during the design optimization as
be eroded in similar sediment conditions. Erosion tendency is
presented in Table 7. Largest reduction in erosion was obtained
dened as follows:
when increasing the number of pole pairs, which implies that the
Pn 3 h i rotational speed of the turbine is decreased. This does however
i1 Wi $Ai
Et Pn m3 =s3 (1) increase the size of both the turbine and the generator, which cause
i1 A increased investment costs as well.
Where n is the number of segment area (Ai) in the runner blade CFD analysis of reference runner has been done to evaluate the
surface. Wi is the relative velocity of ow in each segment area. The effects of change in hydraulic design parameters on sediment
segment area is the area between the intersection of stream lines erosion in runner blade surface. Fig. 12 shows the streamlines of
and stream points in the runner blade surface. ow on the pressure side of the blade of reference runner. Smooth
ow from inlet to outlet section can be observed. Fig. 13 shows the
5.2.2. Erosion factor (Ef) relative velocity at the outlet section of the runner. It shows the
It is ratio of erosion tendency of each new design with respect to average out let velocity at the out let of runner to be in between
the reference design. Erosion factor is dened as follows: 30 m/s to 35 m/s. Fig. 14 shows sediment erosion rate density on the
" # pressure side of reference runner blade. The erosion pattern is
Et New Design predicted to be spread at the entire outlet section of the runner
Ef (2) blade. It can be observed that erosion pattern predicted by the CFD
Et Refernce Design
for the reference design is very close to that observed in the power
The erosion factor estimates a quantitative difference in sedi- plant (Fig. 1a). Same CFD parameters are used to predict the sedi-
ment erosion of runner with the change in hydraulic design alone. ment erosion in optimized design. So consistency in results from
In this study the erosion factor is used as a means to compare the CFD analysis is expected for the optimized design as well.
relative erosion in the optimized designs of runner with respect to It was discovered that by changing the blade angle distribution,
the reference design. and consequently the blade loading, a substantial reduction of
Table 6
Parameters for CFX-Post erosion analysis.
Table 7
Design modications for reducing sediment erosion.
Fig. 12. Streamlines on pressure side. Fig. 15. Streamlines on pressure side.
68 B.S. Thapa et al. / Energy 47 (2012) 62e69
6. Conclusions