Buffer Tank / Pump Placement

RonnieJ

Happy 4th everyone!

I would appreciate your advice please. I'm re-piping my chilled-water AC system (Spacepak) this weekend to improve performance - I'm installing a Heat-Flo buffer tank in the attic with short, fat 2" headers for the secondary supply and return. I chose this tank as it fits vertically in the limited space available. Before I make the final connections, I wanted to run this past The Wall.

My concern is with the primary loop. Most of what I've read recommends the cold supply from the chiller to be connected to the bottom port, and the secondary loop supply header to be placed directly across from it for laminar flow. On this tank, there are four 1 1/4" ports - two on the left, one on the right aligned with the bottom port on the left, and one out the top of the tank. I'm assuming the chiller return should come off the top of the tank, with the SpiroVent and expansion tank at the high point. I plan to bring that line horizontally into the SpiroVent, then exit into a 45 toward the attic floor, leveling off with another 45 and then into the primary loop pump (pump A). The pump is pumping away from the expansion tank, and is about 22" lower than the pipe exiting the tank.

Do I need to worry about pump cavitation with this arrangement? The pump is a Taco 2400-50, sized for the higher head of the chiller (21' @ 12 GPM) and the 1" pex for anything that is not 1" copper. If there is any air in the system, it should be in the pipe at the top of the tank and ultimately will be removed by the vent when I perform the final top-off. I run a 30% Prop. Glycol solution and I monitor the pressure - about 20lbs at rest. I don't have a water-makeup system - I add solution if the pressure drops - it doesn't really ever lose fluid.

The attached diagram shows the relationship of the components. The chiller is out-of-doors at ground level -single story house.

Thanks,

Ron

Ironman

As long as proper static pressure is maintained, what issues would there be?

RonnieJ

Thanks Ironman. In the current arrangement, that pump is noisy. I'm sorting out why - but I'm worried about cavitation. I'll keep the system at design pressure with the new arrangement.

Ron

hot_rod

What is your static fill pressure? Try bumping it up 5 psi, that can sometimes help. The exp tank would need to be adjusted also.

Some installers like theexp tank set a few psi lower so some fluid is in the tank as the water contracts as it cools. Do you see a pressure drop as the system temperature lowers? A larger exp tank or a second one may be a good idea.

You have 1" pex in the chiller loop? 12 gpm is on the high side for 1". Here are some charts to calculate that pressure drop.

I like a 3 pipe buffer for chillers, but that should work okay if everything is sized correctly.

hot_rod

PPI has a handy calculator selection, here is an example of 1" pex
40' at 12 gpm. velocity is getting a bit high, around 5 fps you can start to hear flow noise, mores in the metallic piping

Pipe over 2" can generally run higher velocity, typically sized by pressure drop not the 2-4 fps rule that we use in hydronics.

http://www.plasticpipecalculator.com/PressureDropHeadLoss.aspx

Zman

The 2400-50 is designed for considerably more than 21' @12 GPM. https://s3.amazonaws.com/s3.supplyhouse.com/product_files/Taco-2400-50-3P-Product-Overview.pdf
I would be concerned about cavitation with that circ and 1" pipe on the suction side. Especially if the pipe is pex and/or the distance from the expansion tank is significant.

hot_rod

Zman said:

The 2400-50 is designed for considerably more than 21' @12 GPM. https://s3.amazonaws.com/s3.supplyhouse.com/product_files/Taco-2400-50-3P-Product-Overview.pdf
I would be concerned about cavitation with that circ and 1" pipe on the suction side. Especially if the pipe is pex and/or the distance from the expansion tank is significant.

Nice that it has pressure taps so you could check actual performance.

RonnieJ

Thanks Hot_rod and Zman. The static fill pressure is 20 psi. The chiller calls fora recommended flow rate of 10GPM. My calculated TDH is roughly 37'. I know I am pushing it with the1" pex - there is about 40' of it - the rest is 1" copper. This is a vast improvement over the current layout. I will be installing pressure gauges on the 2400 - and will be taking measurements before and after the changes. I'm avoiding a complete re-pipe due to the investment in copper - a small fortune. I'm adding a third Aircell to address the de-rating of the units that occurred after I purchased them. This was my winter project - getting a late start!

hot_rod

Ideally you would plot be system curve over the pump curve, I-16 shows you how to do that. The intersection develops the OP operating point. It looks like you have one size too large on the circ? Possibly some pretty high flow velocities.


https://www.caleffi.com/sites/default/files/coll_attach_file/idronics_16_na_0.pdf

Zman

At 37' of head you are in the ballpark. How much and what type of pipe is between the exp tank and circ inlet? I do agree that the gauges will be the tell-all. You want all of that head energy to be added to the system static pressure. You do that by minimizing the resistance between the circ and expansion tank.

RonnieJ

Hi Zman, I'm about to construct that now - 1" copper. Short answer, it can be between 6' and 10' 6". If that's not long enough, I can place a coupler and run it another 3'.

From the tank out, 1" copper, 6" to a 45 elbow, 3' 3" down to another elbow to level out at floor height. From there, I have enough copper to make the horizontal run up to 6' 8" before the Y-strainer, and into the pump. The exit from the pump transitions to 1" PEX to run about 10' before exiting the attic, 10' down to the ground where it transitions back to 1" copper to the chiller.

I was going to make that horizontal run just a couple of feet. What is the preferred distance between the expansion tank and the pump?

Thanks,
Ron

hot_rod

The tank connection should be as close to the circ as possible. You could calculate the pressure drop between the tank and the circ to determine what psi is at the circ.

1" type M copper flowing 10 gpm of 60° water, has about 2psi drop per 100' as an example.

You could be a few feet away without any problems. You want positive pressure on the inlet side of the circ.

RonnieJ

Thanks hot_rod. So if I understand you correctly, a shorter distance between the expansion tank and pump inlet is better?

I had assumed the pump "altitude" would be better at floor level, with the height of the buffer tank that rose above it providing some positive pressure (static head). The tank is only about 26" tall. Does it matter where the pump is placed in relation to the buffer / expansion tanks, vertically, in a closed loop system?

If I place the pump at floor level, (Option A) the expansion tank to the inlet of the pump is about 4.5 feet. If I suspend the pump level with the top of the buffer tank/expansion tank outlet, (option the pump can be 16 inches away or so. I have a Y-strainer just 4" in front of the pump suction - should it go elsewhere? Option B places the pump at the highest point in the system - this seems like a bad idea - or does it matter?

In your diagram, the length between the two points would be 4.5' plus the equivalent length of two 45 degree elbows for option A (pump on floor) or 16" for option B, pump level with top of buffer tank. Pipe size is 1" copper, type L. PSI at rest is 20lbs.

Thanks,
Ron

hot_rod

Are the aircellls above the tank and chiller.? If so they are the high points. Not a huge difference in your drawing, I like the circulator low, so it “feels” that static pressure🚰

You know he tank can be located separate of the Spirovent, it does not need to be below it, if that helps.

Can the circ be mounted at or on the chiller, a tee and iso valve just upstream for the tank? Keep all the serviceable components in one location.

If you haven’t purchased it yet, get a combination air/ dirt separator , like a DiscalDirtMag. They do a better job and have less pressure drop compared to a strainer.

RonnieJ

Hi hot_rod, The AirCells are lower - they are floor level - water inlets probably a foot off the floor - so.a third or so of the height of the tank. I have everything already - just revising it. I’m soldering as I wrote this - will be back in operation before bed - whatever day that is 😃. Thanks for the advice - and ZMan too - I really appreciate it.

Ron