Lyubomir Yanev

22.04.2010

Hydropower
Introduction and Background

In this exercise we are aiming to find the most appropriate Kaplan turbine for the River Lune.
The data for the fluid flow is taken from the Caton station and are only valid for 2004. The most
appropriate turbine is going to be the one that produces the most
electricity over the course of a year, which is also going to be our
comparative parameter. The type of turbine that is going to be
installed is illustrated in figure 1. In this case we have a given head H of
3 meters, which defines the choice of turbine, since the alternative
Francis and Pelton machines require a bigger head.

In order to define the best turbine for the particular river, we

In this equation is the density of water, g is the gravitational acceleration, H is the predefined head
and Q is the volume flow rate of the water through the turbine. Therefore, if we take an average day for
the river Lune at Caton, with a flow rate of 39,56 m3/s and leave the minimum of 3,956 m3/s in the river,
the power available in the water with the given head is 1 046 757,6 watts. The amount of power that
can be generated depends on the performance/efficiency of the turbine. Thus, the efficiency coefficient
η is introduced to equation 1 to make equation 2:

[2]
[1]
The coefficient η has values between 0 and 1 and its curve, presented in Figure 2 has been predefined in
the assignment and made use of in the following Matlab scripts. Equation 2 is then used in the
development of the methodology for choosing the appropriate Kaplan turbine.

Figure 2: Performance Curve for a Kaplan Turbine

Methodology

In this exercise we are going to use Matlab as the calculation tool because of its ability to
calculate cumbersome loops, which will be required in the calculation. The first step into the process is
handling the data for the river flow, acquired from the National River Flow Archive of the UK. The data
could very easily be pasted in the Matlab editor and with several lines being brought up, a line matrix of
all the data for the flow rate of the river for every day of the year can be created and manipulated. This
approach however, does not order the days chronologically, but list the first, second etc days of the
consecutive months. Either way, the result for the annual output and the capacity of the turbine will be
the same, however for the purpose of generating graphs and data analysis, the chronological order is
preferable. Thus, the data from the National River Flow Archive is put through a data editing software
called UltraEdit, where the columns of values can be selected and then rotated to create a single line of
values for the days of the year in the right order.

The manipulation of data in the initial code Beta, found in Appendix 1, begins with the creation
of 2 matrices that will contain the values of time and daily output of the turbine [1]. One can notice that
they have 366 elements as 2004 is a leap year and there is a value for the water flow rate on the 29th of
February as well. The script then goes on to create a loop [2] that takes value for each day of the year
one at a time, subtracts the minimum required by SEPA [3] and compares the new value to the rating of
the turbine, which in this script is an input [4]. The test parameter is then put through a series of
comparisons, which represent the ranges of operation the turbine can be working in and assigns the
corresponding value to the coefficient of performance η [5]. The script then uses equation 2 to generate
the amount of power that can be utilized by the given turbine on the given day [6]. That value is then
multiplied by 24 hours to give us the amount of energy in watt-hours [Wh] that has been generated over
the course of that particular day [7]. Position [8] in the script makes sure that there are no negative
values for the energy contribution as that is impossible. Those can arise from the fact that the 3,956
m3/s minimum in the water can be more than the flow rate in river at a particular day i.e. no water is ran
through the turbine and no power is generated. The energy generated in the day is added to the annual
total and the daily contribution from [7] is stored in one of the matrices from [1] for comparison and
analysis purposes. The power rating of the investigated in this script turbine is then reverse engineered
as seen in position [11] in script Beta.

After the validity of the calculations made by script Beta were compared against a reference
calculation, it was upgraded to script Hydro, in Appendix 1. Hydro uses the loop and river flow matrix Qs
from Beta, but this time the value for the flow rate rating of the turbine is loaded automatically by the
program. This is done through another loop [12], where the 100 values ranging from 1 to 501 m 3/s. The
generated annual outputs are stored in a matrix [13]. The range of turbine flow rates is plotted against
the respective annual output and the resultant graph is used to identify the maximum annual
production and the corresponding turbine rating. We can see this graph in figure 3 in the Results
section.

Chosen values can then be manually ran through Beta to compare the operation over the year
and calculate the power [W] rating of the turbine with the flow rate rating in question. We can see a
couple of examples in figures 4, 5 and 6 in the Results section.

Results

Table 1 – Results

Flow Rate Rating Qrat [m3/s] Power Rating [MW] Annual Production [MWh]
166 4.5434 5738
131 3.5855 5701
86 2.3538 5302
 x 10
9
X: 166
Figure 3 - Annual Output at Turbine Ratings
6 Y: 5.738e+009

4
Annual Output [Wh]

0
0 100 200 300 400 500 600

Turbine Flowrate Rating [m3/s]

x 10
7 Figure 4 - Daily Output Through the Year 2004 with Qrat=166
12

10
Daily Output [Wh]

0
0 50 100 150 200 250 300 350 400
Time [days]
Annual Flowrate of River Lune in 2004
500
River Flowrate [m3/s]

400

300

200

100

0
0 50 100 150 200 250 300 350 400
Time [days]
 x 10
7 Figure 5 - Daily Output Through the Year 2004 with Qrat=131
10

Daily Output [Wh]

0
0 50 100 150 200 250 300 350 400
Time [days]
Annual Flowrate of River Lune in 2004
500
River Flowrate [m3/s]

400

300

200

100

0
0 50 100 150 200 250 300 350 400
Time [days]

x 10
7 Figure 6 -Daily Output Through the Year 2004 with Qrat=86
6

5
Daily Output [Wh]

0
0 50 100 150 200 250 300 350 400
Time [days]
Annual Flowrate of River Lune in 2004
500
River Flowrate [m3/s]

400

300

200

100

0
0 50 100 150 200 250 300 350 400
Time [days]
Discussion

In the table of results above we can see that the absolute maximum of annual output come at a
turbine rating of 166 m3/s and 4.5434 MW. However, one can notice from the table and visually on
Figure 3 that the relative gain in the annual output starts to diminish very fast after a certain flow rating,
which is about 130. That is why in table 1, there are 2 more values chosen randomly at points of interest
on the graph in figure 3. We can see that with a turbine of 3.5MW, 22% less than the 4.5MW of the
absolute maximum at Qrat, the annual output only drops by 37 MWh, a 0.0175% loss. A further
decrease in the turbine capacity to 2.35 MW, a 48% reduction on the 4.5MW,will result in a annual
output of 7.6% only. Thus, if the goal is to produce as much energy as possible from the river Lune at
Caton regardless of the price, the best turbine is the first one at 166 m3/s and 4.5434 MW. However, if
we are after Return on Investment (RoI), we will want to compromise with a turbine like the third one at
86 m3/s and 2.3538 MW.

The main focus of the exercise was to adjust the rating of the chosen turbine so that it produces
the maximum energy over a year. This could be achieved by trying to use the Kaplan turbine at its
maximum efficiency and balance that out with the maximum capacity. However, this approach only
considers one loss factor. Inefficiencies that can change the performance graph of the turbine can come
from Casing losses of the fluid going around the impeller in a casing, Leakage losses along the piping and
around the impellers. There are factors that can change the annual output coming from outside the
turbine as well. Those include: Servicing time; Wear of the turbine over time; Blockage of the piping
before and after the turbine reducing pressure difference and flow rate; Losses in the electrical
generation, transformation and transportation.