Differential pressure control valves for Chilled Water System 2

[mechth]

We are going to install VSD's on some existing pumps for cooling. These pumps are all constant speed. Most of the loads have two-way control valve. To make sure there is a minimum 30% flow rate for VSD's, differential pressure control valves are needed. My question is whether we should install the valves on the suction and discharge headers or the end of the loops?

 

[CESSNA1]

JOBCANADA: flow control devices are always installed on the downstream (discharge) side of the heat exchanger or whatever. Installing them on the upstream (inlet) side can lead to erratic operation and cavitation.

Regards
Dave

 

[KenRad]

jobcanada,

Is this a primary-secondary pumping system (with separate pumps for circulating through the chillers) or one single loop?

Assuming you have a primary-secondary, I would take one of the air handlers, and install a 3-way valve. Pick a valve that will give you whatever minimum flow you're trying to achieve, and you're done, and you have no control system to worry about.

If your system is large and diverse enough and you have some year-round cooling loads, you probably will always acheive your minimum flow anyway. You might want to wait and see how the system operates before installing anything.

More importantly, you should make sure that you choose your method of VSD control carefully. Pick a couple of remote points in the system, and install differential pressure sensors between supply and return. Modulate the VSD to satisfy the differential setpoints of each of the zones.

---KenRad

 

[mechth]

Thanks, guys. The pumps are constant-speed secondary pumps. We are going to install VSD's on the pumps and want to make sure the minimum flow rate is 30% of total. The control valves are all two-way valves on each coil. The system is huge so most likely we'll have year round cooling load and will have enough flow rate.

If there is no constant cooling load, i.e. there is no core area, my feeling is that we should install a differential pressure relief valve in the end of the loop, not a by-pass on pumps. What are the pros and cons of these two configurations?

 

[krb]

You'd be better off picking one or more of the two way valves and replacing them with three way valves to meet the minimum flow if this is required. As Kenrad mentioned, this way you have no control system to worry about and you are able to meet the minimum flow. I operate two large systems, and we have no three ways at all, but then we are in the southern US.

KRB

 

[quark]

30% minimum flow seems to be redundant to me because the minimum flow requirement gets reduced when you decrease the speed of the pump. Even at full speed the minimum flow requirement is around 20% with most of the manufacturers.

I always prefer the header bypass(for minimum flow) to be at the end of the headers to avoid possible short circuiting(particularly at low loads) when it is installed in the pump discharge.

Regards,

 

[ODG01]

The bypass should be provided at the end of each loop (where control DP transmitter(s) provide feedback for pump speed/logic to select the critical loop/DP transmitter & maintain its associated minimum DP should be included in the modification) and a three-way valve is a simple solution. By providing the minimum loop flow at the end of each secondary loop the water temperature is maintained at chiller leaving setpoint plus loop gain. The alternate bypass location results in potentially increasing the chilled water temperature to ambient conditions on a static leg. This leads to instability in heat exchanger control loops, increased humidity when air handler coils must go thru a coil cool down to reach design LAT, and the risk freezing for chilled water loops that do not contain a glycol mix.
A bypass near the secondary pump must be set to allow your minimum flow when loop DP increases to calculated maximum for a minimum demand state. Under maximum demand the DP due to the increased flow(dynamic head press) and pump speed will invariably be close enough the setpoint to result in unwanted bypass.

 

[lilliput1]

Make sure your bypass valves are in the mechanical room so water velocity noise will not be a problem. Even with "constant volume" pumps, mix of (2) way & (3) way pumps are a good idea to allow for flow diversity future expansion. The constant volume pump can ride up its curve, saving energy when flows are throttled. The (3) way valves when located at the end of the run would avoid dead legs where water temperature would climb up as ODG1 indicated.