CIVIL WORKS GUIDELINES FOR MICRO-HYDROPOWER IN NEPAL
97
2. The pipe fabricator company proposed to roll the penstock
pipes at 455 mm diameter for the price of 450 mm! This was
because the plate width that the fabricator was able to obtain
in the market was just right for a 455 mm diameter (1430 mm
width). Thus, it was cost effective for the fabricator to com-
pensate extra 5 mm diameter than the additional work re-
quired to cut plate width to make them right for 450 mm diam-
eter. The point to be noted here is that along with optimization
calculations, practical considerations need to be taken into
account as well.
Pipe wall thickness calculations
First calculate the thickness required at the downstream end of
the penstock (i.e. hstatic = h gross = 180 m) using d = 450 mm.
Try t = 6 mm .
a = 1440 1 + [(2150 x d) / (E x t)]
a = 1440 1 + [(2150 x 450) / (2.0 x 6)]
or, a
= 1071 m/s
V = 4Q / πd2 = 4 x 0.450 / (0.450)2π = 2.83
m/s
hsurge
= (av / g) x (1 / n)
hsurge
= [(1071 x 2.83) / 9.8] x [1 / 6] = 52 m
htotal = hgross + hsurge = 180 + 52 = 232 m
teffective
= [6 / (1.1 x 1.2)] - 1.0 = 3.55 m
(1.1 for welding, 1.2 for flat rolled, and 1 mm for corrosion allow-
ance: the corrosion allowance is less than previously
recom-mended for larger schemes because Jhankre was de-
signed to provide construction power to a larger project). Calcu-
late the safety factor:
SF = (200 x teffective x S) / (htotal x d)
= (200 x 3.55 x 400) / (232 x 450)
SF = 2.72 > 2.5, although SF is less than 3.5, it
is acceptable in this case since:
1. There were experienced staff at site. The site staff had installed
penstock pipes in various other micro-hydro projects.
2. The valves at the powerhouse are of slow closing type.
3. The pipes were pressure tested as follows:
Tensile test of steel plates was performed at a laboratory and
an ultimate tensile strength of 400 N/mm2 was ensured as
mentioned earlier.
Rolled pipes were pressure tested at the workshop
at htotal using a hydraulic pump.
Finally, the pipes were also pressure tested on site
after installation by simulating hsurge at the entrance
(forebay) using a hydraulic pump.
4. The alignment was assessed to be fairly stable. In case of
pipe burst it was not expected to instantaneously cause
landslide.
Since the Jhankre penstock is long, to optimise the design, it
was decided to decrease the pipe thickness at lower heads
(i.e.. upstream) using the same safety factor (SF).
Calculations of the static head at which the penstock thickness
can be decreased by 1 mm (i.e. thickness = 5 mm) using the
same safety factor (SF = 2.72) are as follows:
a=
1440
1+[(2150 x 450) / (2.0 x 105 x 5)
= 1027m/s
Note that t = 5 mm in this case
V = 2.83 m/s, (same Q & d)
hsurge = (av / g) x (1 / n)
= [(1027 x 2.83) / 9.8] x [1 /6]
= 49 m
t =effective [5 / (1.1 x 1.2)] - 1.0 = 2.79 mm
SF = [200 x teffective x S] / [htotal x d]
or rewriting this equation in terms of htotal :
htotal = [200 x 2.79 x 400] / [2.72 x 450]
= 182 m
hgross
= htotalt - hsurge
= 182 - 49 =133 m
Therefore the pipe thickness was reduced to 5 mm at a static
head of 133 m in Jhankre, keeping the same factor of safety
(i.e. 2.72) as shown in Figure 6.4. The Jhankre penstock align-
ment for the last section can also be seen in Photograph 6.7.
These calculations were repeated for lower static heads and a
wall thickness down to 3 mm has been used to reduce the cost.