Chilled water flow issue

[BronYrAur]

I am troubleshooting a job where chillers were just replaced. As the building load satisfies and control valves close, the chillers are going off on low-flow. I suspect that I know why, but I wanted to run something past everyone.

A rough drawing of the system is attached. It is a very poor design that has been that way since the mid 80's. They have 2 chillers, 2 primary pumps, and 5 AHU coils. They run both primary pumps and run water through both chillers all the time! Even when a chiller is not running, they pump water through it. So right there, they are blending and losing temperature. Then, they are blending again at the coils.

Each coil has a tertiary pump that constantly recirculates the coil. Depending on the position of the 2-way control valves, the coils are either 100% recirculating or blending water to some degree. So the water that actually hits the coil is much warmer than what is being produced by the chiller most of the time. Lots of improvements could be made for sure.

All that aside, I have a specific question about flow. All pumps are constant volume, but the overall system is really a variable-primary system. As 2-way control valves close, the flow in the primary system begins to drop. There are AUTOFLOW valves just upstream of the 2-way valves. I have attached a cut sheet, but I have had no luck researching these valves. I'm not sure what they are, but I think they are pressure-independent valves that attempt to pass a fixed GPM at all times. So these valves are always bypassing water to guarantee a minimum flow back to the chillers.

I think this concept "worked" with the old chillers, but the new ones probably have a more sensitive low-flow limit, which is knocking them offline as the loads get satisfied. So finally my question:

Since the coils pump continues to run, even when the 2-way valve is closed, couldn't that pumps cause the pressure at P2 to be higher than P1? I understand the bypass line with the check valve directs flow back to the suction of the coil pump, but couldn't the P2 pressure still be above P1? If so, then there would not be primary flow returning to the chillers through the AUTOFLOW valves.

Thanks for you help

 

[LittleInch]

NO .

As you've drawn it you can't have the coil pump flowing water back through the chillers with the control valve closed because you would need flow in both directions at the same time in the section of pipe just above where you wrote TYP (between the bypass line with check valve and the AF valve). That's not physically possible.

It's an odd system alright.

The valve data sheet is on page 2 - I nearly missed it! I don't know what's "auto" about this valve. Looks like an orifice plate to me, but what in god's name is "Petes plug", especially as it's "optional"??? The GPM is based on a pressure ( I guess differential) of 3 to 40 psi. That's what the D in PSID means (I think). You don't say what size valve you have? But yes, it appears as if the coils will never get 200 gpm as some will always be going through these valves

Kind of depends on the range ability of your chillers and whether the AHUs are oversized for the area they are cooling.

You might need a storage volume to allow the chillers to cycle less frequently, but I rally don't think you're getting reverse flow.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[GT-EGR]

Auto flow valves like this usually have a cartridge that sets it to a specific flow. So on your cutsheet you attached, if you know the size - the possible flows you can output is under the flow range GPM. For example you have a 3” valve and get the cartridge for 100 GPM, the valve will flow 100 GPM once you get 3 PSI on the inlet, and it won’t let anymore than 100 GPM through until you exceed 40 PSI on the inlet.

I think a relatively easy fix is you need to find your new chiller minimum, and then (assuming it’s possible like I know other manufacturers can do) change those cartridges so all 5 of them combine to meet your minimum flow.

On the other note, if that 2-way valve is closed, the coil pump is completely independent and has minimal/no impact on the rest of the system.

 

[J.J.Newcomb]

I agree with LittleInch that the coil pump will not increase the pressure at P2 above P1 as long as the main pumps are running and there is flow through them.
The “Pete’s Plug” is a pressure/temperature (P/T) opening that self seals and is used to insert a pressure or temperature gauge into the pipe. This is done so you can use the same instrument several places and get readings that you can compare without worrying too much about the calibration of the instruments.
GT-EGR is correct that the autoflow valve will maintain its flow setpoint for a pressure differential of 3 to 40 psi.
I would guess that the main pumps need to generate at least 7 psi pressure differential between P1 and P2 to allow the 2 Way control valves to work (4 psi over what the autoflow valves need). If you can get that pressure differential, you can probably get the full 200 gpm straight supply water to the coils with no mixing when needed. As you note, there will always be some bypass from the supply to the return at the autoflow valve, which will cool off the water coming back. See “low delta T syndrome” in a Google search.
As far as why the chillers are going off on low flow - autoflow valves are very unforgiving about water quality. Most of them require a 20 mesh strainer. This is because they have tiny orifices that need to move back and forth to work. When they clog, they do not allow any flow. If they are clogged, as the load is satisfied, the 2 way valves close and the flow path is shut off. The chillers then rightfully trip on low flow.
The cheapest way to fix the problem is to replace all the cartridges in the autoflows to give you the minimum flow you need for both chillers. To keep the system running longer, make sure you have a 20 mesh filter in the system. Sometimes air separators will have filters in them, you might be able to work with that.
If you want to save some energy on this system, replace the cartridges in the autoflows to give the minimum flow for one chiller. Then, just run one chiller until the load is greater than one chiller can handle and start the second chiller. I’ll bet this system used to have someone manually operating valves and pumps previously to make this happen.
Let us know how this turns out. It is an interesting system.