Jet propelled developments14 December 2000
Increasing the specific speed of impulse turbines without losing their inherent advantages of simplicity, efficiency and tolerance is commercially attractive in the modern hydro power world. The full admission axial impulse turbine demonstrates such characteristics and, on the basis of current research, may become a useful alternative to traditional turbine designs for small scale sites. Simon Lucas and John Burton give more details
AN increasing demand for smaller water power units in both industrialised and developing countries has been prompted by environmental, amongst other, concerns. Consequently the industry has been forced to adopt different approaches to developing hydro power schemes, including revisiting smaller sites.
Smaller sites usually mean less resources for project research and design, hence the flexibility to accommodate estimates and inaccuracies becomes increasingly important.
The Impulse turbines, most commonly represented by the Pelton turbine, are ideal in most of these respects but have a relatively low specific speed. This limits their use to relatively high head, low flow sites. Increasing the specific speed would allow these rugged, cheap and easily controllable turbines to be used at a far greater range of sites - improving overall power output and reducing operating costs.
The whirling jet concept
The full admission axial impulse turbine (FAAIT) addresses the issue of increasing the flow rate of impulse turbines by drawing on work carried out in the 1950s by Swiss company Sfindex (Societe Financere d'Expansion Commerciale). Sfindex developed an annual jet axial impulse turbine which achieved relatively high full flow efficiency but, due to a complex flow control mechanism, suffered from high losses at part load.
The design is based on the use of a whirling ring jet to drive an axial impulse turbine similar to a Turgo turbine. The whirling jet is developed from an annual volute casing, similar to that used in a Francis turbine or centrifugal pump, as shown in Figure 1.
The flow leaving the volute is a free jet with components of velocity in the radial, tangential and axial directions, as shown in Figure 2. This 360? jet produces a distinctive waist and cone pattern with the cone angle, jet width and whirl or rotational component determined by the volute design and inlet conditions.
The jet formed gives the ideal inlet conditions for an axial impulse turbine, ie a full 360? jet with a low axial inlet angle. The jet can also be designed to have a slight radially inward component at the inlet to the turbine. This counteracts the centrifugal forces acting as the flow passes over the rotating turbine blade - increasing overall efficiency.
Figure 4 shows the operating position of the turbine in the test volute. It is important to note that this system does not require seals and does not create pressure changes in the volute with changing turbine speed or flow rate as the jet is free, ie at atmospheric pressure as it enters the turbine.
The use of the 360? jet greatly increases the inlet area of the turbine and, consequently, the flow rate that can be utilised for a given head. In addition, relative to a Crossflow turbine of similar specific speed, it has a smaller diameter and hence a higher rotational speed at a given site - an advantage at lower head sites.
As shown in Figure 5, the whirling jet is not necessarily one directional. Allowing the jet to form in two directions doubles the achievable flow rate. In addition, it balances the axial loads on the shaft and reduces frictional losses in the volute relative to a one-sided design. In this form the turbine has the potential to accept a higher relative flow than other conventional impulse turbine designs.
This concept has been undergoing investigation at the University of Reading in the UK for the past four years. Initial volute and turbine tests indicated that there was considerable promise in pursuing this approach. Figure 6 shows the results of initial tests using a Turgo turbine optimised for a single jet, compared to results for the same turbine using a whirling jet produced from a test volute. Although not conclusive, these results do suggest a significant increase in efficiency of the full admission jet over the traditional Turgo turbine jet arrangement.
The previous work by Sfindex generated similar efficiencies for small scale test turbines, with a far more complex volute design and a specifically designed turbine. Subsequent work by the company provided the basis for an empirical scaling formula for similar designs. Employing this for initial test data gives the theoretical efficiency curve shown in Figure 7. Clearly the FAAIT is not competitive in its present form. However, given the unoptimised nature of the initial test turbine these results did provide a good justification for continued research.
Since these initial tests were carried out small scale research has continued to improve the understanding of the fundamental principles underlying the formation of a whirling jet from a volute. This has included physical tests and CFD modelling of simplified volutes to understand a number of design parameters. These include whirling jet formation, optimum volute configuration and flow control options.
Whirling Jet formation
The relationship between various parameters such as volute scroll case angle, volute inlet area and whirling jet formation have been analysed based on well established volute design approaches and CFD analysis. Initial findings indicate that variation of these parameters can be used to match specific sites to optimal turbine/volute combinations - gaining high efficiency.
Optimum Volute configuration
A number of configurations have been modelled to analyse the effects of volute design on the jet. Figure 9 shows the preliminary CFD results of a conical volute design. This work has indicated the trade off between the complexity of volute design, frictional losses in the volute and jet formation for optimal turbine operation.
Flow control options
An important advantage of any impulse turbine is the ability to vary flow rates without altering the running speed or greatly affecting the efficiency of the turbine. Tests of the volute at part flow utilising a simple valve mechanism have been undertaken. These, as shown in Figure 10 produce a partial cone with similar characteristics to the full 360? jet and hence the prospect of a good part flow efficiency.
The above results indicate that there is considerable potential in further investigations of the FAAIT concept based on:
• Ease of construction based on well understood designs, e.g. Francis turbine casings and Turgo turbine runners.
• Good full and part load efficiency.
• Higher specific speed than alternative impulse machines - extending the potential range of operation.
• Higher rotational speed and inherent strength relative to a Crossflow or Turgo design at the same site.
These advantages are however only relevant to some sites and it is unlikely that this design will be a serious competitor at the large and medium scales.
However, at the small and micro end of the market the concept does offer significant advantages over existing designs and is therefore worth investigating further. The main thrust of future work is focussed in two areas:
• Consolidating the understanding of the theoretical design parameters involved in optimising the turbine configuration, particularly the volute design and flow control.
• Proving the concept using an optimised volute at a size large enough to allow confident extrapolation to commercial scale.
As is so often the case with successful laboratory trials, making the leap to industrial development and commercialisation is challenging. The researchers are therefore interested in working with turbine manufacturers and other researchers active in the field to develop the concept further.