Secondary Chilled water pump dp sensor position 1

[dulenz]

How do you calculate the best position for differential pressure sensor for controlling secondary chilled water pump and how do you determine actual set point?

I am aware of various rules of thumbs (2/3 of the index run, index run, end of a riser etc.) and recommendations but would like to see a documented procedure with examples if possible


Thanks

 

[DRWeig]

Hi Dulenz,

It depends on your goal. If you want to be able to deliver design pressure to the most remote coil during worst-case conditions, then that's where the dp sensor needs to be (across the combination of the control valve and coil).

I've almost always seen it done this way. If you put it 2/3 down the pipe, you can compensate by setting this level higher than needed so that downstream coil/valve combinations get what they need too.

Basic validation is that the farthest coil needs enough pressure to achieve rated flow with its control valve wide open. Add the pressure drop of the wide-open valve to the pressure drop of the coil at design flow rate, and there's your setpoint.

Good on ya,

Goober Dave

 

[ChasBean1]

10-4

 

[cry22]

ASHRAE 90.1 recommends locating the dp at the fartest coil.
It also recommends that you use multiple sensors if you have multiple distinct zones. You VFD should react to the lowest DP reading of all dp's.

 

[dulenz]

When you say the furthest coil is it physically the furthest or hydraulically (index run)?

 

[dulenz]

What about differential pressure sensor reset strategy (when you reset the set point on a differential pressure sensor until one of the control valves is fully open). Is that sort of control applicable in critical applications such as art galleries and museums? Anybody with actual experience?

 

[DRWeig]

Dulenz,

Done that -- it's a good energy minimization strategy if the pressure reset control loop can be tuned to be stable. Be careful, though -- in the big picture, you're using the system pump VFD to control space temperature in that worst-case space. Can be hard to tune, or impossible. With lots of programming space, you can build yourself in a safety factor perhaps...

I personally wouldn't do it in a critical application unless the customer (end user) were sophisticated enough to buy in with full understanding of the scheme -- including the means to adjust and override it if necessary. A large university, for example, with actual engineers running the plant, might be a good candidate. A museum with only art experts on site (and maybe a mechanic) might be a bad candidate. It's a judgement call.

Good on ya,

Goober Dave