Constant Primary-Variable Secondary Loop CWS Problem

[basseng]

System Description:
2 Trane Centravac 145 Ton chillers in parallel each with a constant flow 290 gpm at 35 ft. head (5 hp) centrifugal chilled water primary pump. 2 variable speed (Danfoss VLT6000 Drives)secondary chilled water pumps (B&G 5x5x9.5 in-line centrigugal)in parallel (1 on standby)with a max. flow of 540 gpm at 45 ft. head (10 hp). Two main loops are 6" diam pipe, and common piping is 6" diam with check valve.

Problem:
When system sees low load and secondary variable flow pumps are operating at low flow, chillers trip out due to low flow conditions (from constant flow primary pumps).

Question:
Does anyone have any idea why the constant flow primary pumps would be operating at lower flow conditions than when the system operates at full load? Can anyone suggest a good reference for design and trouble shooting Constant Primary-Variable Secondary pumping systems?

 

[AlanN]

Why have a check valve in the common piping - this is meant to be minimal resistance? Either the hydraulics is set up right or it isn't - to my mind, using check valves to try and correct bad hydraulics is similar to painting arrows on the side of the pipe (inside?) to tell the water which way to go.

More seriously, you obviously don't have a primary-secondary system, otherwise there would be no effect from one ciruit to the other. Study your schematic.

 

[basseng]

AlanN, the check valve is my first suspect, but I wish to review all culprits before deciding which one will be charged.

Thanks for the reply.

 

[DouginMB]

Primary-secondary piping systems exist to eliminate the type of trouble you suspect (secondary flow affecting primary flow). The pumps should have check valves (probably triple-duty valves) at the pump discharge. If these valves don't fully close (due to, say, trash blockage or bad springs) you can get a backflow through the pump if the other pump comes on. This could occur if one chiller shuts-down and the other one starts-up. You might also need to examine the flow switch serving each chiller to make sure they aren't failing.

 

[AlanN]

DouginMB, surely if the system is hydraulically correct, the primary problems you suggest would exist at all loads (at least up to half load, if you're talking about one chiller)? Maybe they do, basseng would need to tell us that. I agree that it's worth defining as much as possible how the parts of the system are performing, to derive a true picture of the flow.

 

[DouginMB]

AlanN, basseng said the problem occurs during low-load conditons. This would suggest the chillers (and associated pumps) would be energizing and de-energizing. Under these conditons, during a switchover from a lead to lag chiller, one pump could conceivably push water back thru a slightly open check-valve at the other pump. If the flow switch is sensitive, it could at least cause it to vibrate enough to cause a shutdown of the chiller. Is it far-fetched? Sure, but I'm just throwing out ideas.

 

[ChasBean1]

Is the check valve preventing the primary flow from recirculating back to the chillers during low secondary flow? AlanN may have had it right in his initial thought. Sounds like someone "checked" the decoupler line, but again, not sure without seeing schematic.

 

[aknyunt]

Basseng,
What type of control valves are using at your load/process side.Is there any by pass line in primary circuit.If so, how do you control the bypass flow and what is your differential pressure in primary circuit.

 

[EagleCove1]

It is my opinion that DougMB came the closest in nailing this problem. I would bet a dollar to a doughnut that the trip is as likely low temperature and low flow. The chiller control safeties just may not be indicating which was first. Still, the check valve in the primary circuit may cause this anomaly...hence low flow and low temp cutouts.

This sounds like a common control problem when there is no automated control to command lead-lag chiller operation. Depending on the amount that these chillers can effectively un-load, they would likely fail on refrigeration safeties before they fail on flow alarms. However, with the check valve in the primary circuit....I must question the Cv and the actual pressure drop on this. (See my PS on this matter)

It is more likely that one chiller at low load would be capable of handling the low flow required by the VFD secondary system. This sounds like a classic control mis-match that cannot control design rate vs. low demand. Refrigeration machine Mfgs are not know for allowing this level of flexibility given the safeties that are build into their products.

Would recommend that a sequence that measures the chiller delta t vs. the secondary chilled water pump VFD operation. As the flow rate drops, a chiller can be alarmed to be turned off if the pressure differential of the system is being maintained. (We do not recommend the chillers be fully automated for Stop/Start functions) A bell alarm would work well here.

Recommend that you verify the actual percentage that each chiller can unload to. Without knowing the specifics, I would expect not less than 30% unless special provisions have been provided in this selection. Coupled with the chiller flow (Check Valve PD) and time-out for restart in low temperature conditions, these are your likely culprits.

PS: Contact Bell & Gossett and they will confirm that the use of check valves in a primary and secondary systems are NOT recommended. They cause more problems than they solve. Likely one of the contributing factors here.

 

[basseng]

Thanks all for your input. Proper operation of the flow switches and the temperature sensors have been confirmed. The mechanical contractor will remove the flapper in the check valve. If the problem continues to exist we will then perform a diagnostic on the low load cutout of the chiller versus low flow control of the VFD pumps.

 

[DouginMB]

Good luck and let us know how things turn out.

 

[greenaire]

Basseng

Could you let us know how your chiller system is going on?

Thank you.