Petley, Sean and Aggidis, George (2019) Numerical and experimental investigation of flow in horizontal axis Pelton turbines. PhD thesis, Lancaster University.
2018petleyphd.pdf - Published Version
Available under License Creative Commons Attribution-NoDerivs.
Download (17MB)
Abstract
The Pelton turbine is widely considered a mature technology, existing for more than a century. However, manufacturers concerns over operating efficiency are a key driver in the continued design development. In recent decades, the use of Computational Fluid Dynamics (CFD) has become well established in the field of hydro turbine design, analysis and optimisation. However, the complex flow phenomena within impulse turbines have meant that only recently the simulation of the operation of Pelton turbine injectors and runners has been possible. Modern advances in computer hardware enables larger models to be studied in shorter timescales, despite this the application of CFD to the full Pelton turbine including the casing is still very demanding and as far as the author is aware there are no available studies in the public domain, documenting how Eulerian mesh based CFD solvers can be used for this task. The main objective of the present thesis is to understand the impact of the casing geometry on the performance of a twin jet Pelton turbine. This objective has been met by employing advanced CFD models, which has been complemented by experimental testing. In parallel these allowed the description and detailed understanding of the flow features within a horizontal axis Pelton turbine casing and the effects the geometry has on this flow and subsequently turbine performance. The CFD analysis was carried out using the commercial code ANSYS® FLUENT® and included a number of numerical and physical assumptions to simplify the problems, including the usage of symmetry plane and modelling of only six (out of eighteen) consecutive buckets to reduce the size of the computational domain. The investigations can be split into two phases. In the first phase the CFD model was applied to a ‘naked – case’ and examined the impact of the addition of casing inserts such as a jet shroud and bolt on baffle and side shroud which direct the flow away from the injectors and runner, respectively. The second phase examined the variations of the casing width. The CFD models at this stage are only a visual indicative tool and cannot quantitatively predict efficiency as a result of variations in casing geometry, therefore experimental testing has been carried out to complement this work. Experiments were carried out at the Laboratory of Hydraulic Machines within the National Technical University of Athens (NTUA). The first phase experimental testing has found that the addition of the inserts have an overall effect on efficiency of no more than 0.5%, for single and twin jet operation. Furthermore, some inserts and baffles have a positive impact and some are detrimental. The experimental testing of the second phase revealed that a more considerable reduction of as much as 3% in efficiency is experienced when reducing the casing width by 50% and efficiency reduces linearly with width. The CFD results of the casing provide a visual assessment of the flow in the casing and a key outcome of the PhD has been learning how to interpret these patterns and then to make decisions on what experiments to carry out. As a second case study, further CFD analysis was applied to an in-house test rig at Lancaster University designed by the author, and based on this work further experimental testing was then carried out. A second objective of the present thesis was to examine the influence of the injector spear and nozzle angle on the flow behaviour and hydraulic efficiency. The 3D CFD study has shown that injectors with noticeably steeper angles of 110° & 70° and 150° & 90°, respectively, attain a higher efficiency than the industry standard of 80° & 55°. Moreover, experimental testing of the upper jet at NTUA showed gains of ≈1% in efficiency at the best efficiency point of the turbine can be achieved, however there appears to be an upper limit beyond which steeper designs are no longer optimal. Further insight was provided by carrying out additional simulations whereby the 3D velocity profiles obtained from the injector simulations are applied as an inlet boundary condition to a runner simulation and then examining the impact the jet shape has on the runner torque profile during the bucket cycle and the influence this had on turbine efficiency. It can be concluded that the secondary velocities, which contribute to the development of more significant free-surface degradations as the spear and nozzle angles are increased, result in a non-optimal jet runner interaction. Further investigations were carried out in order to compare the absolute difference between the numerical runner efficiency and the experimental efficiency. In doing so, the various losses that occur during operation of the turbine can be appraised and a prediction of casing losses can be made. Firstly, the mechanical losses of the test rig are estimated to determine the experimental hydraulic efficiency. Following this the numerical efficiency of the runner can then be ascertained by considering the upstream pipework losses and the aforementioned runner simulations, which are combined with the 3D velocity profiles from the injectors. The results indicate that out of all of the experimental cases tested, in the best case scenario the casing losses can be approximated to be negligible and in the worst case scenario ≈3%.