| dc.description.abstract | The water intake structure at hydropower plants has the important function
of providing clean water for safe and reliable power plant operation
during all normal situations. This thesis deals with intake technologies
for small hydropower plants and assessment of proposed intake designs
before construction.
All intake sites are unique, with site-specific hydrology, geology, incoming
sediments, types and amount of floating debris, ice conditions, fish species,
local human activity, and prevailing legislation. Moreover, the intake structure
needs to be a well-balanced compromise between functionality and
construction and operational costs. A large portion of the maintenance and
operational costs of small hydropower plants are related to problems at the
intake, and the overall goal of this PhD research is to contribute to more
problem-free intakes.
An intake design process should include the consideration of all sitespecific
challenges, the selection of a suitable intake technology with an
appropriate design, and the evaluation on whether the proposed intake
design meets all the defined performance standards. The papers presented
in this thesis contribute to several stages of the design process. First, the
studies of the Coanda-effect intake screen and backflushing for trash rack
cleaning contribute with applicable information and the development of
promising intake technologies. Moreover, the field study of actual debris
adhesion to trash racks, the studies of threshold values for efficient backflushing, and the use of CFD models for further development of the
standard methods for designing settling basins, all contribute to improving
intake design assessment before construction.
The self-cleaning properties of the relatively new Coanda-effect intake
screen are well-documented, and the screen is installed at many hydropower
plants worldwide. At the same time, very little is known about
the Coanda screen’s cold weather performance. The first Coanda-effect
intake screen was installed in Norway at Dyrkorn hydropower plant in
2010. In addition to performance monitoring during the first winter of
operation, the ice formation on a section of a full-scale screen was studied
in a frost laboratory. Both field studies and laboratory tests showed that
frazil ice particles stuck to the screen surface, but they never entered in between
the wires. Moreover, the screen reopened without any intervention
for all the observed ice-blockage events, as long as there was established
a solid ice cover over the screen, and the frazil ice production stopped or
was prevented from reaching the screen by ice cover on the intake pond.
The idea of backflushing is to induce a brief flow through the rack with
the opposite direction to normal flow. By opening a flushing gate, the
reverse flow will remove clogged debris from the rack and convey it back
to the river downstream from the weir. During this PhD research, the
actual debris adhesion to trash racks is for the first time measured in situ
with a custom-made measurement device. Moreover, the velocities and
pressure differences required for efficiently cleaning with backflushing
was found based on physical model studies of a section of a full-scale
trash rack. The field observations at hydropower plants with backflushing
facilities revealed that most of the debris is removed at the initiation of the
backflushing process, and that evenly distributed velocities over the trash
rack is required for cleaning the entire trash rack. The latter finding and the
threshold values for velocities and pressure differences gave the inspiration
for a new intake concept utilizing a horizontal trash rack. The concept was tested in a physical scale model and proved that the flushing could be
conducted while operating the power plant. Additional modifications of
the intake structure were also found.
Furthermore, a new method for assessing settling basin design was developed.
A settling basin design process is based on trap efficiency calculations
using standard methods. The assumption of uniform flow, which allows for
the use of one-dimensional equations for calculating the turbulence parameters,
is common practice today. The geometry of the settling basin strongly
affects the performance and may induce recirculation zones, secondary
currents, and high turbulence levels; hence, the assumption of uniform
flow is no longer valid. In this study, computational fluid dynamics (CFD)
models were used to simulate the velocity distribution and turbulence
characteristics of the settling basins at the physical hydraulics model of the
Lower Manang Marsyangdi project. The calculated results compared well
with measurements from the physical model. The effective surface area
of the settling basin and turbulence levels were extracted from the CFD
model and used as input parameters for the trap efficiency calculations in
the standard methods.
Because all intake sites are unique, no particular intake design is suitable
for all hydropower plants. The general recommendation for all new
development projects is to carefully perform the design process and assess
the intake design with respect to all performance standards. A promising
way to evaluate intake structures and hence obtain problem-free intakes
is to combine physical models and CFD models of well-defined parts of a
complex intake system. | nb_NO |
