Whirlpooling: Pump Curves Make My Head Hurt

[Bainbridge]

So we've been talking about applying some actual science to our whirlpool speed, rather than "eh, looks about right." Considering our current practice is 1) get it spinning with the steel paddle, 2) turn on the VFD until the pump's just starting to make annoying noise, then 3) throttle the whirlpool valve until it stops making annoying noise...yeah it's probably time to take a closer look. Thus I've spent the morning trying to figure out from the pump curve on our Ampco AC+ 1 horse 1.5" pump what the flow rate is. And now it's two hours later and my head hurts.

The MBAA Practical Handbook #1 (totally pick these up BTW!) says whirlpool transfers should go in about 7-8 feet per second flow velocity. It gives the formula:

V = (0.408 x GPM)/D²

Where V is wort velocity, GPM is gallons per minute and D is inner pipe diameter.

This is for transfer to a separate whirlpool, and ours is combined, but I imagine the idea is to spin the kettle about as long as you would take to transfer to a separate whirlpool, 10-15 minutes. So:

If we want to transfer (i.e. spin) 11bbls in 10 minutes we need to go at:

GPM = (11bbl x 31gal/bbl)/10 min = 34.1

Our pipe diameter is 1.5" so D squared is 2.25

So V = (0.408 x 34.1 gpm)/2.25 or 6.18 ft/sec. So a bit slow actually. But it's 7.25 ft/sec at 40 gal/min. So let's set that as the target flow rate. Which makes sense. A little over a barrel a minute to transfer 11bbls in about 10 minutes.

Now to the pump curve. Ugh. So here's the curve for my pump. It's got a 1 horse powering it. Of course, I have no idea without unshipping it what the impeller diameter is, but I'm going to guess probably 3.5" since it's got a one horse on it.

Looking at the curve, if I want 40 gpm it should be somewhere between 3/4 and 1 horse. Which would be like 85% on the VFD, which is WAY faster than we've been running it. Is this right? How much should we throttle down the pump with the VFD? Or am I obviously not a fluid process engineer and reading the chart wrong?

[LongLiveLagers]

Russ, I think the Kunze/Denk came up with 3.5 cubic meters/second is optimum. I think the often overlooked problem is the shear forces in a smaller pump head that can break up all that good floc. I've found that I only need to get the whirlpool moving to get a good pile. My whirlpool time is under 3 min with 3 speeds on my vfd. I only really crank it for 30 seconds. While I was working at a large craft I did some experiments on different speeds and found them to be inconclusive. Just a little movement will result in a good pile, as long as your vessel is shaped correctly

[jwalts]

Hey Russell, I don't think you can interpolate VFD settings from these curves because they're all for a pump operating at 3500 rpm. Even if that wasn't the case, I don't think setting a VFD to 85% necessarily results in a pump operating at 85% of its horsepower. If you can find a way to ballpark the rpm of your whirlpool near the side of the tank (although not necessarily right at the edge because the wall will slow down the wort), you can use the circumference of your tank to estimate the wort's velocity. It may be as simple as watching hop particles swirl around. Otherwise, you could add something inert that barely floats and take it out before knockout.

You may not need to stir the wort with a paddle either. As far as I'm aware, "time at speed" doesn't mean much. What's important is achieving sufficient rotational velocity by the time the rest starts, which a pump should be able to accomplish over the course of 10-15 minutes. The slowdown during the rest is what actually forces particles to the bottom center of the tank.

It's not very science-y, but I hope it helps!

Joe

[Bainbridge]

Yeah that's part of the problem. The VFD at 50% doesn't mean the horsepower is 50%, it means the amperage and presumably thus the RPMs are 50%. So that'd be the 1750 RPM curve http://www.ampcopumps.com/media/1903...15x15_1750.pdf one would think. But my head still hurts.

@LongLiveLagers - Just tryin to get sciency all up in this biatch! We've tried a variety of different speeds, times, rests, stirring/not stirring, pumping/not-pumping and we've got a regimen that works pretty well most of the time now. But just because 'that's how we've been doing it' doesn't mean it's the best way to do it. And if I could get a tighter cone on our IIPA I could jam a few more pounds of whirlpool hops in there...

[Baldrick]

You are correct. A VFD operating at 50% means 50% of rated speed, so if your pump is rated at 1750 RPM, then the VFD at 50% (or 30 Hz) means that the RPM of the pump is half 1750 or 925 RPM.

I think that you're looking too much into it. The idea is to establish the whirlpool, let it develop without ventilating the pump impeller and then shut it down. The whole process should take no more than 10-15 minutes, plus a rest time of 20-30 min for a 15 bbl kettle. Your description of waiting for the pump to make annoying noises and then throttling the whirlpool valve is not the best procedure.
A better way (still not perfect, but this isn't very exact) would be to run the pump at full speed until the whirlpool forms (enter annoying noise) then immediately slow the pump down to whatever speed maintains the whirlpool until it is fully developed, then stop the pump and shut the valves.
Keeping the pump running full speed and with the valve throttled puts un-necessary shear into the wort.