SwirlFlash: one year on5 April 2002
In May 2001 the first commercial SwirlFlash overspray system was commissioned. This was on a 30 MWe gas turbine, at the Amer power station in Geertruidenberg, the Netherlands. Since then, performance has been better than predicted and inspections after 200 hours and 1300 hours have revealed no erosion of the compressor blades. J van Liere, C G Meijer, Alpha Power Systems, Arnhem, The Netherlands; G H M Laagland, KEMA, Arnhem, The Netherlands; C A Anemaat, Essent Energie, Geertruidenberg, The Netherlands
A seemingly ever-increasing number of inlet air-cooling and overspray technologies are being introduced into the market place for augmenting the power output of combustion turbines.
Evaporative coolers saturate and cool the inlet air, as the name suggests, by evaporation. In contrast, overspray systems saturate and cool the inlet air while the remainder of the water is evaporated in the compressor. Overspray systems can be cold systems or, as in the case of SwirlFlash, hot systems.
The performance improvement achievable with evaporative cooling systems is limited due to the fact that the functionality of these systems depends on the relative humidity at the ambient temperature. An evaporative cooling system cannot evaporate more water in the inlet air than the amount corresponding to the maximum relative humidity downstream of the system.
SwirlFlash versus the competition
Overspray systems offer a more consistent way to increase turbine output than evaporative systems. Overspray systems inject water in the inlet air duct directly before the compressor. The operation of the system is independent of the relative humidity of the air.
Overspraying with cold water is the preferred option since this is the simplest, cheapest and thermodynamically best way to obtain the largest power augmentation. However, nobody has yet solved the problem of producing cold water droplets small enough to evaporate in the compressor during their short residence time. In addition there is the danger that such droplets will cause erosion of the first rows of compressor blades.
The hot water SwirlFlash technology, however, can produce an amazing spectrum of extremely fine droplets. All droplets are smaller than 3 micron and larger than 2 micron.
Due to aerosol-like behaviour the droplets avoid collision with the compressor blades, thus preventing blade erosion. Also, the evaporation of the hot tiny droplets in the compressor takes place in a very short time (a matter of milliseconds) in the first compressor rows. This results in a lower compressor discharge temperature, a higher power output and better turbine efficiency. In addition NOx emissions are reduced significantly.
Another advantage is that the SwirlFlash system can operate at ambient temperatures as low as 0°C. This is due to the fact that the hot water heats up the air slightly when the temperature is low and/or the humidity is high. As such it behaves at low ambient temperatures as a "de-icing" system.
Evaporative cooling systems, in contrast, are normally more limited in terms of the lowest ambient temperature they can operate at - needing a minimum of around 8-10°C. As a result, evaporative systems can only operate for a very limited period compared with SwirlFlash.
Further technical and financial analysis of the various options for augmenting the output of combustion turbines produces some very interesting conclusions.
First of all, it appears possible to gain about seven times more MWh annually with SwirlFlash compared with evaporative cooling or fogging systems given the same middle European weather conditions. The reasons are simple:
• the average ambient humidity is relatively high, thus the effect of evaporative cooling and fogging systems is limited;
• the number of operating hours available with SwirlFlash is greater, resulting in more MWh annually.
The return on investment (ROI) for the various systems is of course the key issue when it comes to decision-making. Table 1 summarises the options in terms of system differences and associated costs (average figures for a middle European 100 MW combustion turbine).
The Amer experience
The proof of the pudding is in the eating. A 30 MWe Alstom/ABB 9D gas turbine at the Amer power station in Geertruidenberg (unit 8) was equipped with a SwirlFlash overspray system, injecting 2 per cent of hot water (at 220°C). The water is taken from high pressure preheater 7 of the adjacent coal fired steam plant.
Both the performance improvements and the emission reductions were closely monitored and analysed.
In addition to dry and wet tests at 100 per cent load, tests were carried out at 75 per cent and 50 per cent load. The ambient conditions were 17.0°C and 1015 mbar for the dry test and 21.2°C and 1015 mbar for the wet test. All data were standardised to ISO conditions, 15°C and 60 per cent relative humidity.
The results are shown in Table 2. Summarising, it was found that 2 per cent injection of hot water at 220°C in the inlet air duct results in: 9.34 per cent more power; 0.96 per cent better heat rate; and a 44 per cent reduction in NOx emissions.
The effects of SwirlFlash injection on the measured performance data and the NOx emission data at Amer are shown in Figures 5 and 6. The measured data agreed with the modelling done to predict the effects of the SwirlFlash retrofit on the gas turbine.
In terms of gas turbine behaviour, performance at the Amer installation during the past year has been very good. No vibrations, no flame instability and no distortion of the compressor house occurred, which means operation was stable under all conditions.
A visual inspection after about 200 and 1300 hours revealed no erosion of the compressor blades. Wear was detected in the spray nozzles, but the wear pattern is within limits and according to expectations.
More troublesome was the performance of some of the valves in the hot water supply lines. The manual shut-off valves in the tube assemblies failed since they were erroneously ordered in the wrong pressure class. They have all been replaced. Filter seals also caused leakage due to wrong materials selection and the very fast start-up time and consequently short heat-up time of the system. The filters have been replaced and the start-up sequence has been changed.
A design review by an independent engineering company showed no additional flaws in the system and at the time of writing it was operating satisfactorily.
TablesTable 1. Comparison of water injection power augmentation options Table 2. Effect of SwirlFlash retrofitted on an Alstom/ABB 9D gas turbine, corrected to ISO conditions (15 °C and 60 per cent RH). The water was injected at 150 bar and 220 °C. The results are for full load