try this website:

http://www.mcnallyinstitute.com/CDweb/o-html/o003_1.htm

however, i also recommend validating the need for the orifice before installing it and obtaining needed information from the pump mfg.

good luck!
-pmover

I haven't crunched the numbers but I strongly suspect 25 gpm per 100 hp is based on some temperature rise across the pump and a fairly high one.

It seems to be very low for minimum flows currently recommended by vendors.  I've just put in a bunch of API pumps, all relatively low head (100 to 300 feet) and up to 1200 gpm or so and the recommended minimum flows were around the 25% or so, some pushing 40%.

At those flows, an orifice is typically not economical due to the continuous energy use and additional pump capacity needed to account for the flow through the orifice 100% of the time.  If the fluid is clean, Yarway's ARC valves work well. If not, a traditional minimum flow loop using a meter and control valve is likely easily justified on economics.

If you want to post the head, flow, SG, configuration (single suction, double suction, etc) and speed, I can do a quick calc using an approach I have.  It's gives a recommended minimum flow comparible with the numbers I gave in paragraph 2.

   The Pump Handbook (1976 Ed.), on page 13-12, suggests 30 GPM per 100 brakehorsepower minimum flow for hot water pumps with a recommended maximum water temperature rise of 15 degrees F. For cold water pumps, temperature rises up to 50F or even 100F are considered permissible. This kind of criterion does not address pressure pulsation concerns associated with suction and discharge recirculation in the impeller which may drive minimum recommended flow fractions much higher than thermal heating considerations. As noted by TD2K above, minimum flowrates of the order of 25 to 40% of rated (best efficiency) flow are being recommended by pump manufactures who can now estimate critical flowrates for peak pressure fluctuations due to recirculation. Pressure fluctuation testing of a 125 hp, 3600 RPM, double-suction pump around 1996 found discharge and suction recirculation pressure pulsation peaks at 30 and 40% of rated flow, respectively (see Atkins,R.A. etal, "New Monitoring System Warns of Cavitation and Low-Flow Stabilities", Pump and Systems Magazine, April, 1996, pp. 12-15).