text
stringlengths
0
1.75k
V.H. ADAMU et al.
2
Methodology
The technical and financial analysis is mainly performed using the RETScreen
software supported by extensive bibliographic review of relevant technology and
financial information. Required technical input data are obtained through direct
field measurements and from existing environmental and technical information
databases online, as well as previously developed papers on the subject. Google
Earth was also used for the topography study of the target site and to determine the
gross head of the river with respect to the position of the hydropower turbine. It was
fundamental to maintain close communication between the authors and interna-
tional partners in Nigeria to obtain reliable onsite data.
3
Results and Discussions
3.1
Field Assessment
The dimensioning of the hydropower plant is based on the net head and the water
flow duration curve (Nasir 2014). These are determined specifically from the
physiographical and hydrological analysis, respectively, of data measured for at
least 1 year. This is usually available from databases of the government or other
private agencies.
Climate and geological conditions are also needed to have a comprehensive
perspective of the location. In the absence of data, in situ measurements can be
made as in this case; for which flow measurements were made in collaboration with
workers at the project site. The float method was used to measure the average
volumetric flow rate of the river (Froend 2012) as shown in Fig. 2. This measure-
ment was made in November 2014, during the dry season in Nigeria.
Hence, the flow rate measured can safely be taken as the firm flow available at
95–100% of the time since flow is expected to increase during the rainy season. A
20 m length of relatively uniform stretch of the river was marked with three points:
start, middle and end points. A float was made to travel between the two extreme
points, and the travel time was measured in order to determine the surface velocity
(Vsurface). To estimate the average cross-sectional area (Aave) of the river, the depth
was measured at sections 0.5 m apart along the width of the river at the points
marked as shown in Fig. 3.
The results of the measurement and the data used for the calculations are shown
in Table 1.
From this method, Aave and Vsurface were determined to be 2.81 m2 and 0.62 m/s,
respectively, and used in Eq. (1):
Q ¼ Aave  Vsurface  Correction factor
ð1Þ
Technical-Economic Prefeasibility Assessment of an Off-Grid Mini-hydropower. . .
1195
Fig. 2 Flow measurement by float method (Joy et al. 2005)
Fig. 3 Cross-sectional area measurement (Reckhow et al. 2010)
Table 1 River flow measurement results
Point
Start
Mid
End
Width (m)
4.0
3.9
4.6
Depth across stream width (m)
y1
0.58
0.63
0.25
y2
0.67
0.68
0.25
y3
0.74
0.74
0.50
y4
0.80
0.72
0.63
y5
0.83
0.62
0.59
y6
0.83
0.48
0.45
y7
0.55
0.76
0.50
y8
0.76
0.34
0.23
Time of travel (s)
Track 1
33.10
Track 2
30.98
Track 3
32.15
1196
V.H. ADAMU et al.
The cross-sectional area was calculated as shown in Fig. 4 above using measured
data from the field.
Subsequently, the average flow (Q) of the river has been calculated as 1.2 m3 /s.
The correction factor of 0.85 was used based on the characteristics of the river