vacuum producing equipment (liq- uid ring vacuum pump) and it can be adjusted by an air intake control valve. Equation 1 can be used for motive steam flow calculations, as well as suction load flow calculations using steam through an HEI nozzle under critical flow. The unit conversion factor U (see Table 1 ) combines the nozzle flow coefficient and dis - charge function at supercritical flow:

Values of unit conversion factor U

ρ

Unit conversion factor U

m˙

D n in

P

867.9

lb/hr kg/hr

psia

lb/ft 3

0.5804

mm

bar

Kg/m 3

Note: P and ρ are for the motive steam upstream of the nozzle.

Table 1

the lower pressure matches the superheated motive flow rate at design pressure. This calculation is straightforward with steam tables and Equation 1 . Adjusting steam pressure to account for superheat impacts the expansion ratio Er, such that the test points Cr, Er, and ω are no longer equal to those for the ejector curve points. Correcting the test point for the difference in Er requires adjust - ing the entrainment ratio ω by an empirical factor which, similarly to the efficiency factor, is also based on manufacturer’s experience. D. Model test Ejectors are scalable devices, so per- formance of a geometrically scaled ejector can be accurately ratioed up to predict performance of the full- size ejector. For large ejectors, it is common practice for the manufac- turer to use a scaled model of the actual jet to optimise steam nozzle position and complete the factory performance tests. Ejector efficiency is significantly impacted by wall friction inside the jet. As ejector size increases, ejector throat area increases more quickly than wall surface area, so that larger ejectors are more efficient because less energy is lost to wall friction. Therefore, performance of a full-size ejector always exceeds that of a smaller scaled version of the same ejector. In terms of ejector testing and certification, this effi - ciency difference can be accounted for by an empirical efficiency factor to produce accurate ejector capacity curves, or the curves can be used as tested and the increased efficiency thought of as ‘bonus’ or ‘safety mar - gin’. The technique explained in testing method B (scaling of the pro- cess conditions) can also be applied to a model ejector if required by the test stand capabilities.

actual ejector can be tested by scal - ing the process conditions. Scaled test results are then translated to the final ejector curves by using Cr, Er, and ω non-dimensional parameters. C. Superheated or saturated steam test Ejector performance is based on mass flow. For typical refinery lev - els of superheat (100-300°F), the extra energy due to superheat gen - erally has a negligible impact on ejector performance. However, superheat is important because of its effect on motive steam den - sity, which changes the mass flow of steam through the critical flow motive nozzle. Steam with more superheat has lower density, so, for the same pressure, mass flow of superheated steam through a fixed orifice is less than it would be for saturated steam. In terms of testing an ejector designed for superheated steam, options are: If the test stand can supply superheated steam For direct and model testing, superheat the steam to achieve the exact motive steam design pressure and temperature. For scale testing, superheat the steam such that both expansion ratio Er and steam mass flow rate scale correctly. The ratio of superheated steam density to saturated steam density should be the same in both cases. If the test stand cannot supply super- heated steam For model testing, the model’s motive steam nozzle throat area can be reduced to account for the higher density of the satu - rated steam used on the test stand. For direct or scale testing, motive steam pressure can be adjusted to get the correct motive mass flow. Superheated steam has a lower density than saturated steam of the same pressure, so density can be matched by lowering test steam pressure until motive flow rate at

2 * √P * ρ

m. steam,critical =U * D n

[1]

Testing methods Manufacturer’s test stands are lim- ited in physical size, steam boiler capacity, steam pressure, and cool - ing water heat removal capability. Given size limitations, most large first-stage ejectors and many larger second-stage ejectors do not phys - ically fit in the existing test facili - ties. Additionally, large ejectors that consume thousands of pounds per hour of motive steam or ejectors designed for high pressure motive steam may exceed test stand capa - bilities. The ability to add superheat to the motive steam can also pose a challenge. Therefore, test meth- ods based on model ejectors, scaled process conditions or a combination thereof may be required. General guidance is to use the simplest test possible, while resorting to rigorous scaling factors and accurate empiri- cal correction factors when needed. A. Direct test Testing is most straightforward for small ejectors that fit on the test stand and require motive and load steam that are within steam gener- ation capacity both in terms of flow rate and pressure. In these cases, motive steam is set at design, and the measured suction pressures, discharge pressures, and equivalent loads can be plotted directly on the final ejector curves without any cor - rection factors. Due to its simplicity, direct test is the preferred method whenever possible. B. Scale test For ejectors that physically fit on the test stand but that require more motive steam flow or higher motive steam pressure than available, the

PTQQ 2 2022 45

www.digitalrefining.com

Powered by FlippingBook