Improving a Closed Loop Cooling System – Flow Regulation between 1L/min and 10L/min

[thomx5]

Hello everyone,

I’ve been tasked with improving an existing closed-loop cooling system, which currently includes a pump, expansion tank, 3-way valve, and a heat exchanger. The system uses potable water, which circulates through the heat exchanger to cool the closed-loop water before being drained into the sewer.

The 3-way valve in the system regulates the temperature by diverting a portion of the water away from the heat exchanger. However, we now need to improve the flow regulation of the system to vary between 1L/min and 10L/min.

• Total Pressure Drop: Around 1.5 bar
• Pipe Size: Considering 3/4" pipes to match the size of most PHEs (plate heat exchangers) and reduce pressure drop by lowering the velocities.

I’m looking for advice on the best type of pump to achieve precise flow control between 1L/min and 10L/min. Some thoughts I’ve had:

• Centrifugal Pumps with Frequency Drives: Would a centrifugal pump with a variable frequency drive (VFD) be the best way to regulate this flow range? I understand they are effective for flow control over varying conditions.
• Solenoid Valve Control: Alternatively, could a solenoid valve be used in conjunction with the pump to regulate the flow? How well does this approach perform for fine flow control, and what are the trade-offs?
• Gear pumps: But these are mainly for hydraulics not for water.

Any suggestions on specific pumps, technologies, or control methods would be greatly appreciated!

Looking forward to your insights!

 

[LittleInch]

A 10: 1 turndown is a lot for anyhting to work effectively.

I don't understand why you need or want to change your current system? What's wrong with it?

You can probably get a VFd or even just throttle in a fixed spee dpump to give your 4 or maybe 5:1 and then use the 3 way valve after that?

 

[Snickster]

Your system is only 2.6 gpm max. What reason would you want to make it more efficient. I would not mess with a 2.6 gpm process system with pumps/heat exchanger if it is working.

 

[thomx5]

Your system is only 2.6 gpm max. What reason would you want to make it more efficient. I would not mess with a 2.6 gpm process system with pumps/heat exchanger if it is working.

Sorry I worded this badly, current system is 3.5L/min and needs to change to max 10L/min. Currently using a centrifugal pump without control on the flow, only 3 modes on but since the pump is underpowered it doesn't reach the full 1-10L/min range I'm looking for.

 

[thomx5]

A 10: 1 turndown is a lot for anyhting to work effectively.

I don't understand why you need or want to change your current system? What's wrong with it?

You can probably get a VFd or even just throttle in a fixed spee dpump to give your 4 or maybe 5:1 and then use the 3 way valve after that?

Can probably get away with a 5:1 turndown, we test watercooled projects here with this system and flow control is essential here. Right now the 3 way valve is being used to precisely control temperature given to what project is connected to the system/load.

But not sure if a centrifugal pump works well with the low flow? Even when controlled by a frequency drive?
Fixed speed with a solenoid valve might be better suited?
An external gear pump might be what we need? But im not too sure, pumps are not my expertise.

 

[Snickster]

I believe a rotary positive replacement pump with a VFD is the way to go. A positive displacement pump will output a set flowrate at a set RPM as controlled by the VFD, and at the same time it will develop the exact discharge pressure required to push the set flowrate through the system. This is because a PD pump pushes out a set volume for each revolution - the higher the revolutions per minute the higher the volume per unit time it will flow. The pump will develop whatever opposing pressure of the friction loss of the piping to produce the flow as long as the motor horsepower is not exceeded.

With a centrifugal pump the developed pressure is dependent on RPM so with VFD you might not get head/output pressure required at the reduced flow and flow is not as controllable directly with VFD speed unless you have flow control too with flow meter/control valve. With PD pump you can have flow meter/sensors and controller controlling the speed of the VFD which will directly control flow by varying pump speed.

 

[thomx5]

I believe a rotary positive replacement pump with a VFD is the way to go. A positive displacement pump will output a set flowrate at a set RPM as controlled by the VFD, and at the same time it will develop the exact discharge pressure required to push the set flowrate through the system. This is because a PD pump pushes out a set volume for each revolution - the higher the revolutions per minute the higher the volume per unit time it will flow. The pump will develop whatever opposing pressure of the friction loss of the piping to produce the flow as long as the motor horsepower is not exceeded.

With a centrifugal pump the developed pressure is dependent on RPM so with VFD you might not get head/output pressure required at the reduced flow and flow is not as controllable directly with VFD speed unless you have flow control too with flow meter/control valve. With PD pump you can have flow meter/sensors and controller controlling the speed of the VFD which will directly control flow by varying pump speed.

I see, right now I'm looking at rotary vane pumps (Fluid-o-tech). And this looks like exactly what I need. Can combine this with a flow sensor and simple PID and think this will work exactly like I would want it to. Most centrifugal pumps are indeed very low head pressure at low rpm's and probably not worth the hassle.

 

[LittleInch]

Just be aware that if you go the PD route you also need to have a pressure relief / bypass installed to prevent high pressure if the outlet is blocked or has a high flow resistance unlike centrifugal which just stop at a certain pressure at dead head.

If you're going this way throw your 3 way valve away and do your flow control using the VFD on the pump

 

[thomx5]

Just be aware that if you go the PD route you also need to have a pressure relief / bypass installed to prevent high pressure if the outlet is blocked or has a high flow resistance unlike centrifugal which just stop at a certain pressure at dead head.

If you're going this way throw your 3 way valve away and do your flow control using the VFD on the pump

Currently 3 way valve is just for temperature control, by bypassing part of the water from going to the PHE and mixing it down the line with the cooled water from the PHE. Currently there is an expansion vessel, but this only protects against thermal expansion (I think). I'd rather combine this with a automatic pump shutdown and pressure sensor. But thanks for bringing the pressure up, haven't put any thoughts towards that..

 

[LittleInch]

Some PD pumps come with built in bypass relief valves which recirculate flow back to the inlet to prevent excessive pressure build up. But if not you need to have something when using a PD pump.

 

[thomx5]

Some PD pumps come with built in bypass relief valves which recirculate flow back to the inlet to prevent excessive pressure build up. But if not you need to have something when using a PD pump.

I see, some of the smaller ones don't. But I'll make sure there is some sort of protection.
For the rest I think I have enough information now, thank you all.