PIC-V Problems

[tmlipcaman]

Ok any and all input is greatly appreciated. I am at a complete loss. A little background to this system. 5 story building with vertical FCU's in each of around 50 typical rooms per floor. Modulating PIC-V's in all units. Chilled water is coming in from a central plant. Main building loop is on the first floor with vertical risers going up to 5 FCU's each. 2 large risers go to the attic to serve 4 larger FCU's and 4 AHU. What we are experiencing is widely varying psid. Everywhere. And nearly none of them within the manufacturer's requirements for proper flow regulation. When I say everywhere I mean to the point of each unit on the same run having extremely differing psid. 1st floor at 1.4, second floor at 47, 3rd floor (within range) at 25.2, 4th floor at .06, 5th floor at 2.7. Manufacturer's optimal operation range being 3-35 psid. We are required to verify correct psid at each of 235 units for commmisioning. Another big fact about this system is it was initially designed having 55% diversity. With total load being 770, total pump capacity at 340. Big miss. So all of the valves are oversized. To compensate and redesign we followed the manufacturer's guidelines and throttled the valves down using their procedures to change how far they could open maximum. Many of the units going from a 3 g.p.m. valve, restricting it to 1 g.p.m. This took our total load all the way down to 400 g.p.m. Ok, so now we are running the pump at full bore, in the service factor and getting 440 g.p.m. out of it. And seeing these psid's. I am baffled. Any insights or thoughts or questions are all greatly appreciated. Everyone Go!

 

[EnergyProfessional]

I think you need to give more information more. what is the pressure the pump generates? It seems high to have 47 psi over one valve (plus all the other pressure losses). Are the pumps regulated by VFD and pressure sensor?

I'd expect the dP near the pump higher than for the "far away" valves.

how did you come up with 770 gpm vs. 340? Maybe diversity assuming not all have full load at once is responsible for the discrepancy?

 

[tmlipcaman]

The pump is controlled by A VFD, but not yet. Right now it's just running in hand at full 60 hz because that's what the commissioning requirements are, verifying proper dP at each valve with the pump at full bore. I agree that we should be seeing higher dP at the units closest to the pump but we aren't. We're seeing sporadic dP throughout. It's baffling. Yet when we verify the pressures at the risers they are very close to the pressures at the pump, minus a small amount as you go up as would be expected. The pressures at the the pump are about 121 psi supply side and 84 return. The same dp as at the valve that's out of range high.

 

[EnergyProfessional]

Keep in mind overall pressuredrop through all paths will be equal (friction loss through pies, coils etc. + dP over valve). If dP over valve is equal to overall dP, that means the valve is closed. So depending on how much pressure is "lost" in the piping the dP over the valve will vary. At fulll flow much of that will be in the pipe, and less over the valve. At partially closed valve, most dP will happen over the valve.

What valves do you have? Most i know (Belimo, honeywell) have an operating dP of 3/5-50 psi and close-off pressure of 200 psi (allthough with a ballvalve I have a hard time seeing how thsi will leak). So unless you exceed 200 psi with a closed valve, I don't see a probblem.

for example look at these spec sheets.

If your designer thinks the Valves operating range of only 35 psi (as opposed to typical 50 psi) is a problem and the pumps are sized to create more.... well, the designer should have chose other valves. but probably not much will happen as long as close off-pressure is high enough. What the operating range means is that up to 50 (or 35 psi in your case) the flow will be higher than the nominal valve rating. Probably not by much. If your valve is controlled by DAT, this probably doesn't matter since the valve will throttle down once it meets the DAT setpoint (in that case no balancing is needed anyway). Assuming you don't have DAT control, then you may have too much flow. but once you implement VFD control, your system pressure should be low enough to not be concerned.

 

[tmlipcaman]

They're griswold space saver pic-v valves. Originally spec'ed for Belimo's and went to griswold on valves under 1". Which happens to be 97% of them. The belimo's are having no problems. The griswold are all over the place. BE at its finest. I'm thinking you're right. I'm leaning toward the pump at this point. It's meeting the flow at full capacity but there's too much differential. When reducing the pump to 75% of capacity I'm seeing the system dP down within operating range of the valves. Too much dP system side is making the valves act as just modulating ball valves basically. No flow regulation aside from the parabolic insert.

 

[EnergyProfessional]

Interesting, I never used Griswold valves, but they looked very similar to the Honeywell and it looks like the smaller Honeywell Valves also have 35 psi range. The Belimo valves are up to 50 PSI.

But in real life it probably doesn't matter because you use a VFD and ideally the pressure setpoint is so low to not have much more than 5-10 psi at the valve when fully open. It seems your pumps seem quite large whne you have that excess pressure. Assuming you had quite some flow int he system still having 47 psi over the valve seems high.

what you are testing is basically driving a car with accelerator pushed all the way down while regulating speed by applying the brake at the same time

 

[Drazen]

the one who advised you to run pump at full Hz, did he/she advice you to open all PIC-V's as well? it makes little sense to me to fix pump speed and leave all control valves to their local freedom for commissioning purposes.

pressure differences on floor could likely be signs of flow differences, water flow escapes elsewhere instead of pushing through almost-closed valves on first floor. did you try to shut down second floor completely? that should give drastic change on first floor dp if this assumption is true.

as a general rule, if you cannot deal with it the other way, you need differential pressure controllers at each section (which is probably a floor in your case).

 

[tmlipcaman]

Yes all valves are 100% open. That's how we have to verify flow. With all valves completely open and pump at full capacity. The system is under control so we could shut down zones, even though the system runs on vertical risers from one main loop. But that's not how we are spec'ed to verify for commissioning. Not that that is necessarily correct. Probably a copy and paste spec.

 

[Drazen]

i believe that way of commissioning is quite obsolete in your conditions, though i keep seeing it in similar situation. i think proper procedures are simply not developed yet on say global level.

as all your valves are self-balancing, you have no means to statically balance pipe network, neither you need it. piping on 1st floor is for some reason oversized, and that's probably all about it.

to verify proper functioning it would be much better to leave all control valves to do their work, than to change settings, i.e. max. temp, than min. temp and monitor how valves will work, and how dp will be maintained by pump.

for pump, critical scenario is when minimum number or no valve is open, and when all are fully open.