SailingEngineer91
Mechanical
- Nov 17, 2014
- 1
Hello everybody,
I've been browsing these forums for the last 6 months and now I finally need your help
Warning: semi long post
So we've been having this problem with our chilled water system where the actuators on a bunch of our valves on the secondary condenser water side are burning out on a lightly loaded building. This is a data center.
It started with us getting alarms on our secondary condenser water temps. We pulled some trends and the problem I am seeing is that it is taking too long for the cooling taking place in the tower to reach the Secondary CW System temperature sensor (at least 5 minutes). By the time the cooling reaches the Secondary supply the cooling tower is introducing too much cooling to the system.
We made some changes to try and alleviate the problem:
1. Adjust Cooling Tower PID Proportional change from 1.4 to 2 to better maintain the speed control
2. Adjust the Bypass mode delay on make from 1 to 8 minutes to allow the temps to maintain set point before going into bypass
3. Bypass set point hysteresis change from 2 to 1 F to allow the system bypass mode to respond better
4. 2nd stage reset delay on break change from 1 to 0 minute to allow staging down back to one tower when needed.
5. All CT fan ramp up change from 1%/1sec to 0.5%/2sec and ramp down change from 1%/1sec to 2%/1sec to maintain setpoint better
6. All CT fan start up delay change from 1 to 2 minutes to allow stage one the chance to maintain set point before calling for stage two.
We implemented these changed which have improved the situation slightly in that the system no longer cycles in and out of bypass. The system is still cycling back and forth from stage 1 to stage 2. One bit of fruit that was harvested from making these changes is that the primary condensing loop isn't running stage 1 flow at building specification. The spec states stage one cooling should run 1,200 GPM and are running at closer to 700 GPM. Stage 2 spec says run 2,400 GMP and we are closer to 1,500 GPM. Is this a big problem? Trends have shown a consistency when in stage 1 at 600 give or take 50 gpm and stage 2 at 1500 to 1600 gpm since the facility opened a year ago.
Here is a breakdown of how the system is set up as of right now:
The SCWS setpoint is 55 degrees (this varies according to outside air wet bulb). The SCWS Set-point resets instantaneously and constant with an exception of a ramp up/down limit of 5 per 1 sec, when wet-bulb is in range of 46 to 70 the output for the SCWS set-point will be 55 to 75 respectively
The PCWP speeds were determined by the balancer prior to commissioning and are set at: stage one runs at 53% and stage two runs at 83% This is consistent with the Commissioning report for 2,400 gpm. Our Primary Condenser Water loop flow meter indicates that the flow in the primary loop is around 650 gpm during stage 1 and 1550 gpm during stage 2.
The system shifts to stage 2 once the cooling tower fan speed has been at 100% for greater than 5 minutes. PCWP first goes to stage 2 speed. Once the run command is sent there is an additional 1 minute delay before calling for tower isolation valves to open. Once the valves prove open the cooling tower fans has a 2 minute delay before starting. Once started the fan will ramp up (limited to 1/2% increase per 2 seconds) to match the fan speed signal of the tower already running. All fans are controlled from a common PID with a min of 10 percent.
The system shifts back to stage 1 once the fan speed drops below 40%
----------
My potential fixes:
1) The SOO calls for the outdoor wet bulb temperature to be read and subsequently for the Secondary CW setpoint to be adjusted every 2 hours. If this setpoint is constantly changing and the building is lightly loaded, won't this contribute to the system hunting that we are experiencing. Should the polling interval to be set to 2 hours?
2) Another concern that I have is the reduced flows of stage 1 and 2, versus what the building was designed for and what was confirmed in Commissioning Agent's report of stage 1 being 1,200 gpm and stage 2 being 2,400 gpm. Without the proper water flow rate through the piping system, could ‘slugs’ of cold or warm water be hitting the temperature sensors and are causing further system hunting? Obviously this wouldn't be a concern as the building load increases, but could this be a cause now? Would we need to increase water flow?
Does anybody have any ideas? I'm all out and really don't have too many people to turn to. I tried to do my research so I didn't come in here empty handed, begging for answers but any help would be greatly appreciated!
The schematic of our cooling system is attached.
I've been browsing these forums for the last 6 months and now I finally need your help
Warning: semi long postSo we've been having this problem with our chilled water system where the actuators on a bunch of our valves on the secondary condenser water side are burning out on a lightly loaded building. This is a data center.
It started with us getting alarms on our secondary condenser water temps. We pulled some trends and the problem I am seeing is that it is taking too long for the cooling taking place in the tower to reach the Secondary CW System temperature sensor (at least 5 minutes). By the time the cooling reaches the Secondary supply the cooling tower is introducing too much cooling to the system.
We made some changes to try and alleviate the problem:
1. Adjust Cooling Tower PID Proportional change from 1.4 to 2 to better maintain the speed control
2. Adjust the Bypass mode delay on make from 1 to 8 minutes to allow the temps to maintain set point before going into bypass
3. Bypass set point hysteresis change from 2 to 1 F to allow the system bypass mode to respond better
4. 2nd stage reset delay on break change from 1 to 0 minute to allow staging down back to one tower when needed.
5. All CT fan ramp up change from 1%/1sec to 0.5%/2sec and ramp down change from 1%/1sec to 2%/1sec to maintain setpoint better
6. All CT fan start up delay change from 1 to 2 minutes to allow stage one the chance to maintain set point before calling for stage two.
We implemented these changed which have improved the situation slightly in that the system no longer cycles in and out of bypass. The system is still cycling back and forth from stage 1 to stage 2. One bit of fruit that was harvested from making these changes is that the primary condensing loop isn't running stage 1 flow at building specification. The spec states stage one cooling should run 1,200 GPM and are running at closer to 700 GPM. Stage 2 spec says run 2,400 GMP and we are closer to 1,500 GPM. Is this a big problem? Trends have shown a consistency when in stage 1 at 600 give or take 50 gpm and stage 2 at 1500 to 1600 gpm since the facility opened a year ago.
Here is a breakdown of how the system is set up as of right now:
The SCWS setpoint is 55 degrees (this varies according to outside air wet bulb). The SCWS Set-point resets instantaneously and constant with an exception of a ramp up/down limit of 5 per 1 sec, when wet-bulb is in range of 46 to 70 the output for the SCWS set-point will be 55 to 75 respectively
The PCWP speeds were determined by the balancer prior to commissioning and are set at: stage one runs at 53% and stage two runs at 83% This is consistent with the Commissioning report for 2,400 gpm. Our Primary Condenser Water loop flow meter indicates that the flow in the primary loop is around 650 gpm during stage 1 and 1550 gpm during stage 2.
The system shifts to stage 2 once the cooling tower fan speed has been at 100% for greater than 5 minutes. PCWP first goes to stage 2 speed. Once the run command is sent there is an additional 1 minute delay before calling for tower isolation valves to open. Once the valves prove open the cooling tower fans has a 2 minute delay before starting. Once started the fan will ramp up (limited to 1/2% increase per 2 seconds) to match the fan speed signal of the tower already running. All fans are controlled from a common PID with a min of 10 percent.
The system shifts back to stage 1 once the fan speed drops below 40%
----------
My potential fixes:
1) The SOO calls for the outdoor wet bulb temperature to be read and subsequently for the Secondary CW setpoint to be adjusted every 2 hours. If this setpoint is constantly changing and the building is lightly loaded, won't this contribute to the system hunting that we are experiencing. Should the polling interval to be set to 2 hours?
2) Another concern that I have is the reduced flows of stage 1 and 2, versus what the building was designed for and what was confirmed in Commissioning Agent's report of stage 1 being 1,200 gpm and stage 2 being 2,400 gpm. Without the proper water flow rate through the piping system, could ‘slugs’ of cold or warm water be hitting the temperature sensors and are causing further system hunting? Obviously this wouldn't be a concern as the building load increases, but could this be a cause now? Would we need to increase water flow?
Does anybody have any ideas? I'm all out and really don't have too many people to turn to. I tried to do my research so I didn't come in here empty handed, begging for answers but any help would be greatly appreciated!
The schematic of our cooling system is attached.
