Gas turbine technology
Novel compressor aims to break small-turbine efficiency barrier1 September 2008
Quasi supersonic compression holds out the prospect of more efficient small turbines, enabling turbine CHP systems and fuel cell/turbine combinations to be applied below the 10 kW size and opening new markets for distributed power.
Why are gas turbines below 30 kW rare in any application field? Despite the attractions of their high power to weight ratios, simplicity and reliability. why are they not employed in light aircraft, UAVs and hybrid vehicles? The answer lies in their low efficiency. The advantages of low weight are soon cancelled by the increased fuel load for acceptable range and piston engines are preferred in these applications.
The inefficiency of small turbines has also prevented their application in micro CHP. Here the problem is not overall efficiency, because using the waste heat makes the overall efficiency high. Rather, it is the low ratio of work output to heat output. There are more efficient and cheaper ways of converting fuel to heat than in a gas turbine. Unless work output/efficiency of electrical power generation is high one might as well use a boiler.
The smaller the gas turbine the lower its efficiency. The reasons why gas turbines suffer from this negative scale effect include: larger blade tip clearance relative to size, which leads to increased leakage losses; larger relative thickness of blades, which leads to increased aerodynamic losses; and smaller blade heights and passages, resulting in boundary layers taking up a larger fraction of the passage cross-section. This increases drag and reduces flow.
These scale problems particularly affect the compressor. Compression is inherently more difficult than expansion (ever wondered why the compressor of an axial gas turbine requires more stages than the power turbine?). Leakage in a turbine expander is merely lost opportunity to convert pressure to work. Leakage in the compressor, however, travels upstream, requiring the leaked gas to be re-compressed. As the leaked gas also carries energy upstream with it, more work is required to re-compress it than to compress fresh, cool gas. Thus, as leakage increases with size reduction, the impact is greater on the compressor than the power turbine. This leads to a vicious circle as more of the output of the turbine is consumed by the compressor rather than being available to generate electricity.
With a reduction in size a dynamic compressor has to be smaller and faster to achieve an adequate pressure ratio at the required flow. The turbine, however, does not have to reduce much in size, but for reasons of simplicity and manufacturing cost it is easier to mount both compressor and power turbine on the same shaft, forcing a reduction in size on the turbine because it has to run at the same speed as the compressor. Gearing the two to operate at different speeds is not regarded as economically practical. Cost and pulsating input to the combustor prevent the use of screw compressors to solve this problem.
There is no prospect of efficient small gas turbines using current technology. However a new method of compression, in the early stages of development, could lead to efficient small turbines. Called quasi supersonic compression, it has been developed by SVT Licensing in collaboration with ICON, the consultancy arm of Imperial College. Research has been partially funded by a grant from the UK DTI (now Technology Strategy Board) under the spring 2006 funding competition.
The quasi supersonic compressor (QSC) fills a gap between dynamic and positive displacement compressors. It has the pulse-less flow of a dynamic compressor, but at a higher pressure ratio and with higher efficiency. It cannot surge. The input speed required is moderate, enabling it to be directly driven either by a low speed turbine, or electrically or mechanically.
How it works
Quasi supersonic compression exploits the phenomenon that a pinch point between a cylinder orbiting within another can rotate supersonically while all physical components and the process gas move much more slowly. The supersonic speed of the pinch enables high and low pressure to co-exist at opposite ends of the same chamber. A short gas path, low acceleration, low turbulence and low mechanical friction ensure high efficiency and low gas noise.
In a quasi supersonic compressor the cylindrical rotor orbits and rolls close to an outer cylindrical chamber wall creating a pinch point. The pinch point advances at supersonic speed. The chamber and rotor walls are stationary, relative to each other, at the pinch point. However the rolling motion makes the rotor surface move in the opposite direction to the direction of travel of the pinch point. Thus each revolution the rotor surface moves a distance equal to the difference in circumferences of chamber and rotor, while the pinch moves the distance of the chamber circumference in the opposite direction, ie much faster than the rotor speed. The larger the rotor size relative to the chamber, the slower the speed of the rotor surface. Also the larger the absolute size of the chamber, the lower the rpm required to get the pinch point moving at supersonic speed.
Gas is aspirated through the wall of the rotor into the expanding duct that follows the supersonically rotating pinch point. Gas in front of the pinch point is in a narrowing duct and is compressed and forced through the wall of the stator cylinder. Because the speed of the pinch point is supersonic, the pinch point overtakes information about pressure increase. This enables a high pressure outlet zone to co-exist in the same chamber as a low pressure inlet zone.
Process gas follows a relatively straight radial path from inlet to outlet with low acceleration. Turbulence is low and gas/surface friction is minimal. Compression is positive displacement but continuous and pulse free. There is no clearance volume. These attributes combine to produce high compression efficiency.
1D and 2D modelling shows that pressure increase occurs below supersonic pinch speed because of pressure wave reflections and chamber geometry. Modelling shows that efficiency is high and that the flow and pressure ratios are comparable or better than those of a centrifugal of similar size. A rarefaction follows the pinch point ensuring good inlet flow. With the benefit of optimisation, adding an inducer to the rotor or shaft, centrifugal effects within the rotor and diffusion effects on leaving the chamber, it is expected that performance will be considerably enhanced. Furthermore, a QSC compressor cannot surge because the compression method is positive.