Posted with kind permission of
International Water Power & Dam Construction
Quadrant House, The Quadrant, Sutton, Surrey SM2 5AS, UK Fax: +44 181 652 8904

ISSN 0306-400X, 48, 3, March 1996

A US patent has been granted to Hydro West in Washington, for a new 'potential flow' turbine. Martin Hindley talks to the company about the 'Gokhman' invention.

The hydraulic turbine, some may argue, is a fairly mature product, offering little scope for design extension. Research and development efforts have been largely directed at improving the performance and efficiency of existing turbine designs rather than completely re-inventing the 'wheel.'

Now, researchers in the US have bridged the gap between the tried and tested and the truly innovative by launching an entirely new hydro turbine concept.

The 'Gokhman' potential flow turbine, named after its inventor, Dr. Alexander Gokhman, has been patented by the Washington, US-based Hydro West Group, in a move that promises to revolutionise the business of generating hydro-power. The turbine could improve generation efficiency of the best conventional designs by between 1 and 3%, Hydro West claims.

The Gokhman turbine works by providing 'potential' water flow to the runner, eliminating losses that 'engineers had previously accepted.'

In conventional turbine designs, the energy lost to 'whirl' - the product of the circumferential component of velocity and the radial distance from the turbine axis - discharged to the draft tube can be as much as 25% of total turbine losses, Hydro West says. Efficiency gains should be highest in the low-to-medium-head Francis turbines as well as Kaplan and fixed-blade propeller units, the company claims.

According to Hydro West the new turbine should allow hydro-power operators to upgrade their plants with higher efficiency units with greater specific speeds and a better cavitation index than more conventional designs.

To demonstrate the 'improved performance' of the new turbine, Hydro West is carrying out model tests in conjunction with the Ganz Machinery and Energy Company in the Slovak Republic.

According to Hydro West Chief Engineer, William Holveck, the trials have confirmed that the vortex in the draft tube can be virtually eliminated with the new turbine.

'Even with some flow separation (the runner design is not yet completely optimized) the efficiency is still good - above 93% on a 300mm model runner,' Holveck says.

In an interview with Hydro West, Water Power explores the features and benefits of the Gokhman turbine.

Water Power: Just what is a potential flow turbine?

William Holveck: In general terms, it is any turbine where the flow approaching the runner is potential - that is, having a constant 'Whirl' value throughout the area of the runner inlet. Whirl is the product of the circumferential component of velocity (V_u) multiplied by the radial distance from the turbine axis (R).

WP: What is the Gokhman Potential Flow turbine?

WH: The Gokhman Potential Flow turbine is named after its inventor Dr. Alexander Gokhman of the Fluid and Power Research Institute in San Francisco. It is a potential flow turbine with a radial intake and eitherž mixed flow or axial flow runner, like a Francis or Kaplan turbine.

The Gokhman turbine has wicket gates that provide potential flow to the runner, which utilises the potential flow delivered to it.

WP: How does it differ physically from a conventional turbine?

WH: Very little. The Potential Flow turbine fits in a conventional turbine case. The wicket gate discharge angle changes along the gate height span from top to bottom to provide potential flow to the runner inlet, and the runner is designed so its discharge has no whirl entering the draft tube at design mode.

WP: What's wrong with the conventional turbine?

WH: The conventional turbine is equipped with wicket gates that are cylindrical in the strict mathematical sense of being uniform in cross-section over their entire length. They have a constant discharge angle from top to bottom.

However, in a radial inlet turbine, the flow to the runner veers towards the bottom of the gates as it begins the turn from radial to axial and thus the velocity near the bottom of the gates is higher than it is near the top.

      Figure 1 shows the streamlines of the axisymmetrical flow for a typical Francis turbine. Notice how they are closer together near the bottom of the gate. The closer lines represent higher velocity.

Since the gates have a constant discharge angle, and the velocity changes from top to bottom, the whirl component of the flow velocity varies from top to bottom.

This is an important point, because the turbine produces power by changing whirl as the fluid flows through the runner. There are three problems with this conventional design:

1. The difference in velocity across the leaving from the trailing edges of gates. This causes the relative flow into the runner to be unsteady. As a runner blade passes from one gate to the next, the conditions at the inlet edge of the blade are changing. Thus the design conditions are ambiguous - does one condition for the condition mid-way between gates align with a gate or some other location?

   This also dooms attempts at high accuracy flow analysis - small changes in entrance conditions can have large effects on the downstream flow. The variation of inlet conditions changes the circulation around the runner blades, and unsteady vortices are shed from the runner blades into the draft tube.

2. Energy in these vortices cannot be recovered. Each gate vortex passes through the runner and appears in the draft tube, with the blade vortices, where they promote chaotic instability.

3. Most importantly, the whirl variation from top to bottom of the gates and thus from the crown to the band of the Francis turbine or hub to blade tip of the Kaplan runner, cannot two conditions at the design point:
   1. At the centre of the runner, the discharge must have zero whirl. It is well known from experiments that whirl at the centre of the runner discharge results in a pressure field that disturbs the flow through the runner from the designed flow lines and the efficiency falls off rapidly.

   2. The efficiency and the change in whirl from each partial runner (or flow line) must be identical. There are no physical boundaries to maintain the flow within the volume defined for each partial runner. If the efficiencies or work for each partial runner are not identical, then the flow will not follow the streamlines for which the blades are shaped, and the efficiency will suffer.

   Since, in the conventional turbine, the inlet flow has varying whirl, and the change in whirl for all the flow is the same, the discharge flow must then also contain varying whirl.

   And because the whirl at the centre is zero, the whirl at the periphery is equal to the difference in whirl from the top to the bottom of the wicket gates. So, at best efficiency, there is positive whirl from the runner everywhere except at the centre.

   At reduced load there is positive whirl throughout the discharge, and at overload there is a mixture of negative whirl near the centre and positive whirl near the periphery.

WP: Is this important?

WH: Absolutely. This means that for the conventional Francis or Kaplan turbine there is never a point where the flow in the draft tube is meridional, and that there is always some energy wasted in circumferential velocity leaving the runner which cannot be recovered. It is known from analysis and measurement that this energy can be as much as a quarter of all the turbine losses. These losses have been accepted as unavoidable until now.

WP: You say now. What has changed?

WH: The potential flow turbine has changed all this. Since flow from the wicket gates is designed to have constant whirl and the runner is designed for this condition the turbine can discharge flow with no vorticity. The only energy leaving the runner for the draft rube is associated meridional velocity, which the draft tube will recover.

There are no vortices streaming from the wicket gates and the relative flow into the runner is uniform. The draft tube has much less energy disposed towards exciting roughness.

The benefits of the turbine extend to off-design performance as well, because the whirl value as the load is increased or decreased from the design point is less than for a conventional turbine.

The kinetic energy associated with the discharge whirl is, of course, proportional to the square of the circumferential velocity, so, as load is reduced in the conventional design, these losses increase very rapidly compared to the potential flow turbine.

WP: What about cavitation?

WH: The cavitation characteristic of the potential is greatly enhanced for mixed flow turbines. In these turbines near the band where cavitation usually occurs the absolute velocity is reduced with the potential flow design, and the sigma value is reduced significantly.

This allows higher specific speeds for otherwise identical conditions which will result in lower costs for both turbine and generator as well as higher turbine efficiency for new machines.

WP: Can the flow still be shut off by the wicket gates?

WH: Yes, the wicket gates have a cylindrical inlet section, and the tail of the gate comes to a vertical line at the discharge, which closes completely against the adjacent gate.

WP: What about runaway speed? Can I retrofit my turbine without replacing the generator and still have a safe margin of critical speed and stress?

WH: This is an important question, and unfortunately, one that can only be answered definitively by model testing, whether you ate using a potential flow or a conventional runner replacement. There is no way to model mathematically the runaway conditions with adequate accuracy.

WP: What other benefits are expected?

WH: The improvement in cavitation results from increasing the minimum suction side pressure on the turbine blade. Along with the elimination of vortices, this may improve the safe passage of fish through the turbine.

Current research by US Government agencies is intended to reveal the relative importance of low pressure, blade strikes, and other effects on fish passage.

The elimination of vortices in the draft tube may improve operational flexibility at facilities where draft tube surges limit the operating range of some machines.

It is worth noting that for some projects with long water conduits, a 4% improvement in the turbine efficiency may yield an additional 1% or more in the penstock efficiency due to the reduced flow and associated head losses.

The improvement in off-design performance may mean that in some multi-unit plants, some Kaplan runners can be replaced with fixed blade runners reducing retrofit and maintenance costs.