View Full Version : Glycol Header Sizing w/ Flow limiters on tanks

12-13-2013, 12:29 PM
Hey guys, having a bit of an issue with setting up our glycol system.

We're setting up a brewpub. 10BBL system, 5 jacketed 10BBL fermenters, 10 jacketed 10BBL serving tanks. We're going with Cool-Fit headers, drops, and valves, then the flexible hose supplied by the tank manufacturer (Specific) to connect the tanks. Possibly we will run Cool-Fit all the way to the tanks in the future. Glycol chiller manufactured by Pro.

Pro sized us with their 5HP MA model. We then began to talk about header size. Specific suggested flow requirements of 2GPM for holding temperature, 6GPM for crashing. Pro came back and told me that that's all well and good, but since the tanks have 1" connections, they will draw 10GPM whether they need it or not. 10GPM * 18 (3 future tanks) * 70% = 126GPM. Pro recommended I go with a 2.5" header, upsize the reservoir to 200gal, and upsize the process pump to 2 or 3 HP. Having worked in a brewery with a suspect glycol system, I decided to go with all the recommendations, and figure out how to deal with the price increases later.

Of course, being an idiot, I forgot to actually confirm with Pro that I wanted these upgrades, and I learned yesterday that my chiller is built, without the upgrades. I have not ordered the glycol piping from GF yet.

My options are to either:
a) Find another buyer for the completed system (possibly not an easy task), and order a new one, delaying me a month or so, as Pro has a bit of a backlog
b) Explore sizing the system for only the required 2/6GPM per tank specified by the manufacturer. Also we'd be able to maybe reduce the header size, saving a boatload of cash.

Could we look into adding a 6GPM flow limiter onto each tank? Does a part like this exist that is reliable? If we only had 6GPM max in each tank, that brings the whole system down to 6 * 18 * 70% = 76GPM. If we have some form of adjustable flow limiter that is generally set to 2GPM, and can be switched to 6GPM when crashing the tank, we would reduce flow requirement down to (2*17+6*1)*70% = 28GPM. From my understanding, this would eliminate the need for the upsized process pump and reservoir? Is there a good reason this type of thinking isn't valid? Is there something else I'm over looking?

I'm discussing this with Pro, GF, and Specific right now, but just wanted to see if anyone here had any tips they felt like sharing.


12-13-2013, 01:39 PM
I have "flow limiters" on my BBTs. I used big globe valves. Keep closing them until you see your tank is not getting down to temp in a reasonable amount of time, then back off a little. I am also thinking of adding valves to three new FVs as they came with 1.5" glycol pipes on the tanks. Three jackets all with 1.5" inlets and outlets is way to big for a 60 bbl uni. I think I could crash the tank from 70 to 32 in an hour. Haha The globe valves are more expensive than balls or gates but they are the perfect design for adjusting flow whereas the others will fail over time as they are not designed to restrict flow.

I am not saying you should go with the small chiller however. That you really need to crunch the numbers for. The serving tanks are a huge load as they will be crashed most of the time. With FVs there is more time spent at warmer temps so less load. I think I would take the chiller they have now. If you run into problems when you get all your tanks rolling you could buy a small used chiller and run two or three of the tanks off it. In the mean time you will have beer leaving your building and money coming in. Just my opinion.

12-14-2013, 05:10 PM
"they will draw 10GPM whether they need it or not."
Not really. It would be rare to have every tank on cooling at the same time. And if they don't need it, the solenoid will turn off the flow. I think you would be fine with what you have now. I run 90hl from a 5hp unit on 1 1/2" lines in a very hot tropical climate and have no issues. I wouldn't push it any further, though. Pro Refrigeration are certainly very good at what they do, but maybe this is overkill. A 200 gallon reservoir just gives you a bit of thermal flywheel. It doesn't add any capacity to your mechanical refrigeration; just adds buffer capacity. This is great if you have a 2 stage heat exchanger that takes glycol for the secondary. And adding a 3hp pump seems like a lot. 1 1/2 or 2hp would probably do. I'd try it with 1hp first. The bigger this pump is, the more heat it pushes into your glycol. Smaller is better. And for a header feeding 18 tanks (why no doubles?), you may want at least 2" header feeding the 1" drops. I don't see the value in "flow limiters" as they just reduce flow rate at the expense of duty cycle length. Do you really want the tank temperature to fall slower so that you can have more tanks call for glycol? Or stick with a quick cycle that has higher flow rate? Only reason I can think of for having any kind of limiter is to keep the cycle time longer for the purpose of NOT undershooting your tank temperature when the beer cools to setpoint, the solenoid closes, and there's still lots of icy glycol in the jackets. I've seen tanks drop another 2-3C because the glycol temperature and volume were sufficient to overcool. I'd stick with tank manufacturer's recommendations. Good luck!

12-15-2013, 06:28 AM
We run a " Lot " more than that on 2 inch supply with a 1 HP pump.
I would not have serving tanks on a system that is used for cool down heat exchange for the brewhouse.
Its not a good idea to hard pipe fermentation or other vessels in. Hoses from a header are better for a number of reasons.

12-18-2013, 01:30 PM
Yes very true Steve. It will only be heavy btu when actually crashing.

12-19-2013, 09:01 AM
Another way to limit flow to each fermenter is to have a programmable temperature controller. When you are crashing a fermenter from say 68 F to 40 F, have the controller lower the setpoint by 2 F per hour. The tank will get all the flow it needs to lower the temperature at whatever ramp rate you want.

12-19-2013, 04:31 PM
That's brilliant Linus! Use the controller as the limiter. There's another technique I've used to have the glycol temperature as a permissive to the fermenter temperature controller. You only cool your fermenter when the glycol is cooler than the fermenter setpoint.