sizing valves to the flow

Mark J Strawcutter

Sizing valves to the flow and not to the pipe size strikes me as a very reasonable thing to do. And Cv ratings give you a way to do a relative comparison of the loss between two valves (of different sized, by different manufacturers, whatever).

But how do you (if you do) use Cv for the absolute sizing of a valve for a particular application? A valve can operate above and below it's Cv rating with corresponding increase or decrease in delta-P. Do you try to find a valve whose Cv rating matches the design flow? Or one whose Cv is slightly above or below design flow?

Does it vary depending on the type of valve (balancing vs differential pressure bypass vs flow-control)?

Mark

David Van Wickler_3

Here's what I do. Most of the time

What I do is choose the valve for small residential systems (under 10,000ftsq) for a little pressure drop – 1psi. So if the system is 100,000BtuH and I have sized my pumps to the system curve to provide 10 or 15gpm (20F drop or 15F drop) I choose the valve that will give me 1psi drop with that flow (10 or 15). Even at 15gpm I would specify a 1” three way valve. Why? Because I design radiant and in fact most system to operate with 3-4fps velocity and head loss that is manageable by small pumps WITH sufficient safety factor, i.e. shorter loops, etc. The pump may be capable of 25% more flow, but I get back a lower delta Tee across the system or zone and the mixing valve, with higher flow rates becomes the “authority” or the controlling “Delta P” device.

The difficulty lies in multiple pump applications where the flow is changing constantly as zone are off and on. It could be said that rarely will all zones be on at the same time, hence the valve sized for total flow will not produce 1psi at low load and will in fact be oversized under these conditions.

It is IMO better design to AT LEAST consider a properly sized mixing valve with a little pressure drop through it. That is, find a valve that will match the design flow rate at 1psi. Realize that a properly designed system will produce higher flow rates that calculated and you will be “in the zone”.

Look at Griswold Controls – They have a line of motorized ball valves with a parabolic insert that throttles and mixes in a linear pattern. The insert also restricts flow and depending on the profile one line size valve offers 5 different CZV’s.

David Van Wickler_3

Here's what I do. Most of the time

What I do is choose the valve for small residential systems (under 10,000ftsq) for a little pressure drop – 1psi. So if the system is 100,000BtuH and I have sized my pumps to the system curve to provide 10 or 15gpm (20F drop or 15F drop) I choose the valve that will give me 1psi drop with that flow (10 or 15). Even at 15gpm I would specify a 1” three way valve. Why? Because I design radiant and in fact most system to operate with 3-4fps velocity and head loss that is manageable by small pumps WITH sufficient safety factor, i.e. shorter loops, etc. The pump may be capable of 25% more flow, but I get back a lower delta Tee across the system or zone and the mixing valve, with higher flow rates becomes the “authority” or the controlling “Delta P” device.

The difficulty lies in multiple pump applications where the flow is changing constantly as zone are off and on. It could be said that rarely will all zones be on at the same time, hence the valve sized for total flow will not produce 1psi at low load and will in fact be oversized under these conditions.

It is IMO better design to AT LEAST consider a properly sized mixing valve with a little pressure drop through it. That is, find a valve that will match the design flow rate at 1psi. Realize that a properly designed system will produce higher flow rates that calculated and you will be “in the zone”.

Look at Griswold Controls – They have a line of motorized ball valves with a parabolic insert that throttles and mixes in a linear pattern. The insert also restricts flow and depending on the profile one line size valve offers 5 different CZV’s.

David Van Wickler_3

Pic of the \"Optimizer\"