It should be specified by the turbine OEM. At least that is the case with aircraft gas turbine engines and their nozzles.
Type of fluid is specified (usually a calibrating fluid, per some MIL or ASTM or equivalent spec.), temperature range and supply pressure. Lower and upper flow limits (usually in mass flow units) are set for these conditions and the actual testing is rather simple.
A pump capable of supplying let say 50% larger flow at that pressure or the original pump is required, an adequate flow-meter would be nice, a thermometer and that is it. If there is no flow-meter than a calibrated vessel with a volume scale and a chronometer are required to measure (calculate) the volume flow rate. Sometimes it may be necessary to preheat the fluid or to cool the flow. But I have never heard that air or water is utilized for the testing, not for flying gas turbines.
For some nozzles even desirable spraying pattern is defined. Therefore it is not sufficient to meet quantitative limits only but qualitative also. I doubt that something like it is necessary for large gas turbines’ (not aero-engine derivatives). I believe that their nozzles are more like burners, well I might be wrong about that.
What balu1 has to say is true. I've build two gas turbine liquid fuel nozzle testers which have worked beautifully. Here's what I did. Contructing the tester around a small electric motor driven gear pump (industrial gas turbine, positive displacement) and a tank (40-60 gallons), the test fluid used is straight diesel. The important thing here is that the system is protected from any leaks which could be ignited by any sparks (use fully enclosed motor). The chamber to measure the flow was a cylinder about 6" in diameter and about 2.5 feet long with two precision float switches, one mounted at the bottom and the other at the top of the float chamber. The switches trigger a digital counter, the bottom switch starts the counter and the top switch stops it. You could have it count in 10ths of a second or any other value so long as it's count rate is high enough (needed for precision). The burners need to be calibrated one against the other until you have a set within a few tenths of a second (or other value) to reduce temperature deivation in the engine. The spray pattern check is easy. Simply fit the fuel injector into a box with a heavy plastic window for the flowing fuel to stike it. By observing the flow pattern, the injector is either good, (even, round flow pattern, no streaks, or fuel shooting off to one side) or it's not. It's a bit difficult to describe my design, but it works well and consistent. I've been testing industrial turbine diesel fuel burners with this tester for about 12 years with good success. If you need help in copying my design, let me know. I might be able to assist.
Regards,
TurboTec
