6 Cooling Twrs/ VFD's..how to stage

[ccnman]

Question for all ..i have a plant with 5 chillers that reject heat through 6 brand new cooling towers the are driven with ABB VFD'S.the owners in house electrical engineer is asking me to confirm the most efficient way to stage the tower fans on.
example ....he is suggesting with one fan energized and fan speed reaches say 30 htz (or an alternative freq less than 60 htz ), then to start the second..and then when two fans reach this setpoint to start the third fan ..all fans running would operate at same htz...
so my question is would you run the lead to 60 htz then start 2nd fan and run both to 60 htz together then start 3rd...etc etc...or would you start each fan at a lower setpoint and then run both together ...
thanks

[Shophound]

Seems to me the intent behind having VFD's on cooling towers is to set them up so the blower modulates below 60Hz to maintain ECWT. One tower running full blast vs. two running considerably slower...fan laws would drop combined amp draw below what just one fan motor pulls at full speed.

Will you have DDC/BMS staging the towers? One method is that lead tower stages on and attempts to maintain ECWT setpoint. If it can't after given amount of time, lag tower stages on, then next lag tower, if needed, will stage on as needed. Same goes for staging off. If lead and lag towers hold setpoint or better, lag towers stage off and lead takes full load until it's maxed out.

If ECWT setpoint is reset by dewpoint or wet bulb, towers would modulate to maintain setpoint in accordance with outdoor ambient conditions and condenser load. I've had good success with tower design approach over wet bulb, with my towers modulating below 60 Hz on the hottest of days when wet bulb drops a bit.

The Hz would be determined by load. One tower trying to hold setpoint at 60 Hz might see ECWT creep over setpoint if staging interval between lead and lag tower is too long. Lag tower bumps on and runs full blast to help bring setpoint down, then Hz of both towers will likely be considerably below initial full blast Hz of lead tower as setpoint is re-established and held.

[vbhvac]

I do not know if 50% or 30 Hz is actually good number usially the most efficient speed for the fan is between 75 and 85 % - you can find it on the fan curve - and that will be the point to bring next fan.

[ccnman]

Originally posted by vbhvac
I do not know if 50% or 30 Hz is actually good number usially the most efficient speed for the fan is between 75 and 85 % - you can find it on the fan curve - and that will be the point to bring next fan.

ok, so it is true then ...what he asked me for...which was to confirm the most efficient motor speed....
so on a rise in cwst i will increase the motor speed until this point on the curve then start my next fan and ramp them at the same speed ...on a continued rise when i get to that same speed i would add a third and so on...
what about a drop in temp.....would the point where i drop them be at a lower speed...is that something i would plot on the motor curve or would be be arbitrary?

thanks again

[railroad48]

Here's's the old man's idea, for whatever it's worth. Each VFD should have a minimum speed of 10-18 hz. I would stage each fan on as the CWST rises above setpoint. Whenever all 5 fans are running at minimum speed, begin modulating all fans together to maintain setpoint.

My idea is, all the fans at minimum speed will cost less and have more cooling efficency than one motor at 60 hz.

[trane tek]

Just my .02 but if the CT fans are gear driven, the VFD minimum speed should be 22 Hz (37%) to keep the oil splashing in the gear boxes properly.

The subtraction rate for each fan should be higher. In other words, make it easier for fan to come on than turn off.

[chillersandcontrols]

There are so many variables here.

Are the towers all one common sump? Or are the chillers piped to individual towers? I have seen situations where the piping to a common header is a nightmare of diverting and mixing.

What type/brand of chillers? Some chiller designs can tolerate much lower ECWT than others and be more efficient in the lower temp ranges.

What model ABB drives?
If the new ACH550, then most of this control can be done completely within the drives (they have double PIDs internal)!

Combine all the previous suggestions and you will have a great system.

PS: The "engineers" may design it on paper, but WE make it work in the field!

[qtip]

Originally posted by chillersandcontrols

PS: The "engineers" may design it on paper, but WE make it work in the field!

Isn't that the truth

[vbhvac]

I would maintain CWT SP with VFD PID loop same for all fans, with one fan running, if VFD of this fan will be greater then 75-80 will bring second one on, and drop it off when the speed will go below 60-65, same for the rest of the fans.
On seperate question how to run chiller plant with gereater energy savings, i would agree we have a lot more variables, but the bootom line is run it according to the manufactures recommendations. York for example is asking for cooler condensor water, Trane is OK with warmer. So in the first case you would save money on chillers, spend them on cooling towers, in the second it will be opposite. What is better who knows.