Pumping cereal mash

[Woolsocks]

I'd like to perform a cereal mash in my 2-vessel brewpub system. What I'm thinking is I use the kettle to cook the cereal. If I use enough water, I'm hoping I can run the whirlpool as I boil the cereal mash so as to keep the grain from scorching. The next trick is getting the cereal mash into the main mash. I don't have a pump specifically meant for moving mash, but if the mash were thin enough, and made of mostly adjuncts (no husks to break), would there be any issue with being able to pump it from the kettle to the mash tun?

[banjolawyer]

Related to woolsocks' issue but more generally, I'm interested in learning more about pumping mash from a mash mixer to a lauter tun where both vessels are on the same level.

What type of pump and tubing would be used? What is the relationship between how much mash you're moving to tubing diameter and pump speed?

Thanks.

[Guest]

In one brewery I worked at, for transferring mash from our mash-mixer/copper to lauter-tun our plant we had a 'positive displacement pump'.

(capable of moving slurries, etc)

cheers
MikeMcG

[gitchegumee]

I've set up quite a few breweries where either mash or adjuncts were pumped. Currently in Palau we do exactly what Woolsocks is contemplating. We use second heads rice as adjunct and don't mill any further. We use a 10:1 ratio of water to rice. We cook our rice in the kettle for 40 minutes before adding to our malt mash. We keep the whirlpool going the whole time to help liquefy the adjunct, and to keep the rice from clumping on the bottom. Our kettle is direct fired, and we cannot actively boil while cooking without scorching the rice (another in a long list of reasons to choose steam-fired equipment). Instead we boil the water, add the rice slowly at fire-out, and stir with a paddle from the middle while the whirlpool mixes things. Our kettle is lower than our combi-tank mash/HLT, and our kettle pump (AKA whirlpool pump & vorlauf pump) pumps the thin slurry to the mash tun as we mash in. We have a VFD on the kettle/whirlpool/vorlauf pump (HIGHLY recommended), and we use another 20 liters or so of attemperated mash water to rinse the kettle of any remaining rice. We consistently hit our mash temperature, and the kettle is quite easy to clean before the runoff--sooner is better. Another great advantage of this "double mash" regime is that you could mash in your malt while the rice is still gelatinizing. You can mash in a bit cooler and stiffer and hit either a protein or beta rest before raising the temperature via the hot adjunct. Cooler and stiffer favoring proteases and hotter and thinner favoring amylases. Likewise, you can add salts to change the pH at different times/temperatures. For example when you want a higher pH early in a cooler & stiffer mash. Lots of variables here to play with even if you have only a 2 vessel brewhouse. To answer Banjolawyer's question regarding a dedicated three or four vessel brewhouse, the best systems I've seen use a (relatively) large flexible impeller pump installed in a 2"-4" line between the mash mixer and the lauter tun. Or between the cooker/kettle and the mash mixer, in the case of decoction recipes. A flexible impeller pump has several advantages over others; first, it is very gentle on solids so it won't obliterate your husk fraction. It is relatively cheap, simple, reversible, positive displacement, easy to vary speed with a VFD, acts as a check valve either way, easy to clean, and cheap & simple to repair. And the relationship between the pump & tubing size vs. batch size is flow rate. Or more specifically, mash velocity. My very quick calculation shows that a 20 bbl batch will pump completely through a 2" tube in 30 minutes with a velocity of about 2 feet/second. I wouldn't pump mash much faster. Anyway, hope this helps and I'd love to hear from others who do something similar. Cheers!

[dick murton]

You should not use a conventional high speed centrifugal pump, used normally for pumping water, wort or beer for pumping across malt mash. It will shear the grist particles, destroying both the larger endosperm particles and particularly the husk particles. I saw this happen at one brewery, where a centrifugal beer pump was used to transfer the mash immediately after mashing, across to the mash vessel, which as it happened, finished the pulverisation process off. And then they wondered why they could not lauter properly - very long runoffs, poor extract efficiency, high solids in the wort and blocked false bed slots and runoff pipes

Specific low speed, low sheer pumps, typically with a fairly open impeller are commonly used for malt mash transfer. Somewhere I have a diagram of one such design, but I can't find it at present, but will post it as and when I find it.

Because the rice or maize mash does not normally have husks in, a convention high speed pump can be used with less risk here.

Briggs quote a maximum mash transfer speed of 1.5 metres / sec, but in my opinion, this would only be ok if being introduced into the lauter through the floor. When introduced through the side wall at low level, the turbulence is still far too high and settling time is required before vorlaufing. You certainly cant start vorlaufing whilst still transferring as there is not a stable filtration bed.

[banjolawyer]

I have been confused about measuring the speed of liquid or mash at a distance per second. I first tried to figure this out when reading an Ashton Lewis description of how fast wort should be pumped to a whirlpool.

I can understand gallons per minute, but not distance per second. How is such a thing measured when talking about liquid and mashes with various liquor/grist ratios?

[dick murton]

Work out the volume of the mash, divide by the cross sectional area of the transfer pipe at the inlet to the lauter tun to give you the equivalent length of pipe of that diameter it occupies. Divide this length by the number of seconds it take you to transfer it to give you the sped in distance / second.

Or to work out the transfer time required to achieve let's say 1.5 metres / second (5 ft / second) divide the equivalent pipe line length by 1.5, to give you the number of seconds it will take. If working this out, then in my experience this is the absolute minimum time required, and ideally it should take longer, say equivalvent to 1 metre / second. Unfortunately I have not been able to evaluate slow speeds as the transfer pumps are fixed speed

See also other comments made in the last few days in the message board about mash transfer pump design so you don't smash up the grist into unfilterable mush

Cheers

Happy Christmas

[banjolawyer]

My apologies, but my background is completely law and homebrewing. Thanks for being patient with me. I'm confused about how to measure "the cross sectional area of the transfer pipe at the inlet to the lauter tun." I can't picture it.

And when you say "number of seconds it take you to transfer it," by *it* do you mean the entire mash?

Thanks!

[gitchegumee]

Banjo, to get any amount of liquid through a pipe, it obviously must move. How fast (speed or velocity) it moves determines the flow rate (volume/time). Big pipes move more liquid than little pipes for a given speed of liquid flow. The size of a pipe may be expressed in a diameter, or a cross sectional area like square inches or square cm. Volumes may be expressed in liters, gallons, or cubic feet/inches/meters. So if you know how many cubic liters you wish to pump in a given time, you can calculate how fast it will go in a given pipe size. 10 cubic meters of liquid pumped in 1 minute through a 1 square meter pipe will travel at 10 meters/minute. So to do things right, you need to limit the speed of the mash. These calculations are also necessary in CIP--you need at least 4 ft/sec to introduce enough friction in the pipe to clean it. Hope this helps.

[banjolawyer]

I'm sure I'll get this eventually. I guess I need to take this in baby steps until I have a break-through.

gitchegumee, you did express it very simply that "10 cubic meters of liquid pumped in 1 minute through a 1 square meter pipe will travel at 10 meters/minute"

I just have to bend over backwards when working with these calculations because I'm not used to them, and I always get thrown for a loop when working with imperial units.

From git and dick's posts, I understand that the transfer rate needs to be between 2 and 3.3 feet per second (and I'm assuming that the slower is better).

The mash tun I'm working on will have a 2" TC port, so the ID of the ferrule and piping is 1.87"

I searched for "cross sectional area of pipe" formulas, and found these:

http://www.engineeringtoolbox.com/pipes-equations-d_873.html

http://www.inter-mountain.com/Pipe_Related_Formulas.htm

http://wiki.answers.com/Q/How_do_you_calculate_the_cross_sectional_area_of_a_pipe

The one from engineeringtoolbox.com and the explanation in the answers.com link both calculate the cross sectional area as

A = 0.785 * di squared
where
A = cross-sectional area of pipe (Square Inches)
di = inside diameter (inches)

The one from inter-mountain is different (and even seems to have an incorrect result, but maybe I'm mistaken)

CROSS SECTIONAL AREA (A): The cross sectional area expressed in square inches is used in various tubular goods equations. The formulas described below are based on full sections, exclusive of corner radii.

Round Tube: A = p/4 (D5 - d5)

Where:

D = Outside Diameter, inches

d = Inside Diameter, inches

Example: Calculate the cross sectional area of a 7" O.D. x .500" wall tube.

D = 7.000 d = 7.000 - 2(.500) = 6.000 inches

A = p/4 (D5 - d5)

A = 3.1415/4 (7.0005 - 6.0005)

A = 10.210 inches

Which is the proper way to calculate cross sectional area, and doesn't the inter-mountain calculation arrive at an incorrect result by their own formula?

[banjolawyer]

oh, also.. why only measure the cross sectional area of the pipe at the lauter tun input? If you have the same size ID at the mash outlet and the lauter inlet, but have a smaller or bigger ID through the pump, what is the impact of that?

Merry Christmas brewers!

[gitchegumee]

Banjo, the speed of liquid flow will indeed change as the diameter of the pipe changes. A neck down to smaller diameters will increase the mash velocity. Generally, these things are designed with pretty much constant pipe size--Not the same case as with less viscous liquids like water which may have larger suction piping than discharge. So if you wish to limit mash velocity, then measure the pipe at it's thinnest section--where the velocity will be highest. The correct formula for area of a circle is pi*r**2. For a 1.87" Inside Diameter (ID) pipe, you'll have 3.14 * (1.87"/2)**2 = 2.75 square inches. Now let's say we limit the mash velocity to 2 feet/second = 24 inches/second. That allows maximum transfer rate of 24 inches/second * 2.75 square inches = 66 cubic inches/second = 3,960 cubic inches/minute. There are 231 cubic inches in a gallon, so you get 17.1 gallons/minute. Not so bad, eh? So now the formula you used from Inter-mountain calculates the ANNULAR area. You calculated the area of WALL of the pipe--not the inside area. Useful for knowing how heavy it is. The flow rate is completely independent of the outside diameter of a pipe. It only matters what the inside diameter is. Hope this helps.

[banjolawyer]

Thanks for all your help Phillip. I get the calculation finally!

The system I'm working on is small, so I'd like to figure out some solution of moving only about 45 gallons of boiling mash into the main mash. I assume it isn't a problem to move the mash slower than 2f/second... you just don't want to move it faster.

I'm going to keep researching pumps, but with such a small volume, it also appeals to me to find a big wooden scoop and transfer it manually :-)