Dr. Sasha Savic, Alstom Power Service, Baden, Switzerland, and Dr. Paul Kippax, Malvern Instruments, Malvern, UK

Injection of water into the inlet duct of gas turbines is a well-established tool of power augmentation, but droplet size and distribution is critical. Laser diffraction analysis has resulted in an improved understanding of the dynamics of fog formation, and could allow inlet fogging use to become more widespread in Europe.

Inlet fogging is now a well-established method of power augmentation in gas turbines. Increased power output is achieved by injecting a water-droplet ‘fog’ into the air inlet of the turbine. Evaporation of the water droplets within the inlet stream causes the air to be cooled, increasing the mass flow through the turbine and the turbine power output.


Figure 1. Alstom’s wind tunnel set up (Source: Alstom)
Click here to enlarge image

null

A novel variation of the technique of inlet fogging is high fogging where most of the water injected into the inlet stream passes through the turbine compressor and performs compressor inter-cooling. This can lead to a further increase in the power output compared to standard inlet fogging. However, high fogging presents additional technological challenges such as adjustment and modification of the gas turbine air-cooling, combustion, control and protection systems. Care must also be taken to maintain compressor stability and blade mechanical integrity. Of prime importance are the droplet size, size distribution and concentration of the water droplets contained within the fog as this defines how rapidly the droplets evaporate within the compressor system.

Using a Spraytec laser diffraction-based droplet size analyzer from Malvern Instruments, researchers at Alstom have optimized the droplet size distribution of the fogging process and increased the power of a gas turbine by eight per cent. The real-time, in situ measurements from the Spraytec system resulted in an improved understanding of the dynamics of fog formation and enabled an assessment of different injection nozzle types. Conventionally, laser diffraction droplet size measurements are carried out in a laboratory. However, by using the Spraytec system directly on a gas turbine, Alstom was able to quantify the effect of different droplet size distributions on the power of a gas turbine for the first time.

High fogging

In the well-established technique of inlet fogging, evaporation of the water droplets ahead of the compressor causes air cooling, increasing the mass flow through the turbine. This can lead to an increase in power output by up to ten per cent depending on the environmental conditions.

With the relatively new technique of high fogging, special nozzles are used to form a fog of small droplets (10-60 microns in diameter) just upstream from the compressor bellmouth. These droplets are drawn into the compressor, where they evaporate performing compressor spray inter-cooling. This can lead to a further increase in power output compared with the inlet fogging technique. Compressor spray inter-cooling helps to reduce parasitic compressor work and at the same time increases the air mass flow through the engine, thus increasing the total power and efficiency of a gas turbine.


Figure 2. Malvern’s Spraytec particle size analyser was located upstream from the compressor bellmouth (Source: Alstom)
Click here to enlarge image

null

It is a rule of thumb that one per cent of injected water (relative to the intake airflow) boosts the turbine power by 5-7 per cent[1]. However, the main prerequisite for the correct performance of a high fogging system is that the water droplets are small enough to evaporate on their journey through a compressor. If the water droplets are larger than required evaporation occurs within the compressor outlet, rather than the early stages of the compressor. This reduces the efficiency of the inter-cooling process and therefore the degree of power augmentation obtained. The presence of large particles can also increase the risk of erosion of compressor blades. Selection of the correct fogging nozzle system is therefore critical.

Particle sizing

The particle size distributions produced by different fogging systems were measured using the Malvern Spraytec spray droplet sizer[1]. This system relies on the technique of laser diffraction, where the intensity of light scattered from the droplets passing through a laser beam is measured as a function of scattering angle. The angle at which droplets scatter light is inversely proportional to their size, allowing the droplet particle size distribution to be directly calculated from the spray’s scattering pattern using an appropriate scattering model, in this case Mie Theory. A patented multiple scattering algorithm incorporated into the Spraytec allows the accurate measurement of both dense and diffuse spray plumes.

Measurement using this technique is very fast, and the Spraytec is capable of acquiring size distributions at a rate of up to 2500 Hz. This allows for the measurement of short-duration sprays, revealing fine temporal fluctuations in the atomiser output. Continuous measurements over long periods of time are also possible at lower acquisition rates, enabling users to understand the stability of different nozzle systems and also directly track the effect of changing atomiser conditions on the delivered particle size.

Experimental set-up

Three set-ups were used to test various types of nozzles. Experiments were carried out in still air on a test rig; in a wind tunnel that had a maximum wind velocity of 20 m/s; and at the intake of a full-size 55 MW gas turbine, equipped with a high-fogging system.

Measurements within the gas turbine system were performed using a high-fogging nozzle rack, placed just upstream from the compressor bellmouth. Pressure swirl-type nozzles were used and the air velocity during the tests was set to 20 m/s. The Spraytec transmitter and receiver optics were mounted 2.5 m away from each other, with the laser beam penetrating the high concentration fog observed just beyond the fogging rack. A vibration-free operation was ensured by mounting the Spraytec size analyser independently from the inlet duct, with PC and power supply on the roof of the generator enclosure. Large clearances between transmitter and receiver modules allow the accommodation of sprays with large cone angles and minimize contamination of the optics.

Results

The effect of both air flow rate and nozzle orientation on the particle size distribution produced by a typical high fogging system was measured using the Malvern Spraytec.

The most important parameters which must be assessed when characterizing fogging sprays are:

  • D[3,2] (Sauter Mean or surface area mean diameter)
  • D[v,10] (the particle size below which ten per cent of the volume of droplets exists)
  • D[v,90] (the particle size below which 90 per cent of the volume of droplets exists).

The D[3,2] and D[v,10] are directly related to the evaporation rate as this is defined by the overall droplet surface area. This in turn, allows the calculation of the concentration and size of the droplets that will enter the compressor. The D[v,90] provides information on the largest droplets present in the spray and therefore relates to the likelihood that erosion will occur within the compressor during fogging.


Figure 3. Droplet size distribution using water injection perpendicular to the airflow (Source: Alstom)
Click here to enlarge image

null

Figure 3 shows the results of testing different air velocities for perpendicular spray injection using an impact pin nozzle in a wind tunnel and indicates that the droplet size decreases with increasing air velocity. This is due to increased droplet break-up at higher air velocities. The dilution effect observed with increasing air flow rate also decreases the probability of droplet coalescence near to the nozzle.

Figure 4 shows the same experiment with parallel injection of the spray. It can be seen that, while droplet size does decrease with increasing air velocity, the effect is not as obvious as for perpendicular injection. This is believed to be due to the increased droplet-shear observed at the nozzle when injecting perpendicular to the air stream.


Figure 4. Droplet size distribution using water injection parallel to the airflow (Source: Alstom)
Click here to enlarge image

null

Finally, using pressure swirl-type nozzles, a comparison of droplet size distribution was carried out during gas turbine operation and using a laboratory test rig but without airflow. The size distribution measured under laboratory conditions is much wider than that seen in turbine measurements. Turbulence within the turbine air intake system causes large droplet break-up and enables the rapid evaporation of any fine droplets. This is not true in the laboratory measurements where the air flow rate is zero, increasing the fine and coarse particle fraction and yielding broader size distribution.

Long term outlook

Malvern Instrument’s Spraytec system has proved to be a valuable tool in monitoring the droplet size distribution produced by inlet-fogging or high-fogging systems used for power augmentation of gas turbines.

This research has shown that the droplet size of the resulting spray can be reduced by injecting the water perpendicular to the airflow. It has also been shown that increasing the air velocity reduces the droplet size. Therefore, it is necessary to carefully consider both of these effects in addition to the intake duct geometry for optimal positioning and arranging of the high fogging nozzles in the gas turbine.

Alstom now uses laser diffraction droplet size measurement to characterize specific nozzles for each individual installation of a high-fogging system. This is carried out in a wind tunnel as part of a quality control process. Laser diffraction is also used to assess long-term operation of nozzles.

Using Malvern Instruments’ Spraytec system has enabled the company to improve the efficiency and reduce the cost of its high-fogging systems.

References

[1] S Savic, G Mitsis, C Härtel, S Khaidarov and P Pfeiffer, “Spray interaction and droplet coalescence in turbulent air-flow, an experimental study with application to gas turbine high fogging”, Proceedings of ILASS Europe, Zaragoza, Spain, September 2002.