Packaged CW Central Plants
We have 3 AAON pre-manufactured packaged central plants piped to a common header. Each machine is roughly 150 tons with variable secondary CW flow/CV primary flow with 8 - 25ton compressors (copeland scrolls) and variable condenser fan speed control.
The sequence from day 1 has called for all three machines to run simultaneously.
When we are in low load conditions (180gpm demand @ 3-5deg detla T) all three machines lockout on barrel freeze protection. When we disable 1 of the machines the problem persists and the other 2 lockout on freeze protect.
When we get into these low load conditions (with just 1 compressor running) the condenser fans start to short cycle (ramp to 100% and back to 0% within 1 minute).
1. I believe that the condenser fan VFD PID loop is not tuned properly but AAON says it is supposed to work that way. Why?
2. The whole point of a primary secondary system is to enable the machine to run from 100% load to 0% load without freezing out. Why can't these machines handle it?
Not sure if I read that correctly but are you saying that all 3 machines are running when there is a low load? Or are all 3 just enabled to run?
Does each machine has its own primary pump?
What is your chilled water setpoint? Freeze?
I do not have experience with these machines but soundl like you have a min of 75tons running under your "low load", is there hot gas bypass?
At what temperature below setpoint will the machines cycle off?
Sorry for all the questions!
You're asking some very technical questions for someone that none of us know anything about you. Could you put something in your profile so we know who we're talking to from an experience standpoint?
1) If AAON says it's supposed to work that way, they'll have to tell you why. That don't mean it's right, but they designed it. May be that you have to do something to tweak the operation. (Sounds like your accel/decel time is set too low or the throttling range is too small.)
2) Primary/secondary just smooths out the operation of the chilled water plant when you have large loops and/or load variations. As milar asked, what type of low load controls do you have? Who is in control of the staging? Are there isolation valves on the packages? How is your staging set up? What's your low temp cutout setpoint? Design flows vs. actual thru hx's? Is your decoupler designed correctly? Give us some detail.....
Questions are great - bring em' on!
1. "Are the machines running or just enabled?"
All three machines are enabled via our bacnet integration HOWEVER the machines do not talk to one another so they all try to maintain LWT setpoint.
2. "Does each machine have its own primary pump?"
Yes, and secondary pump as well - keep in mind that these are pre-manufactured "complete" central plants.
3. "LWT setpoint/freeze setpoint?"
LWT = 44 deg. freeze lockout = barrel at 35deg.
4. "what is the cycle off setpoint?"
This is not evident via the bacnet integration. it looks like they are trying to operate within a +/- 2 degree deadband.
5. "Is there hot gas bypass?"
no - and these aren't digital scrolls :-(
6. "What are the low load controls?"
these packaged plants come with MCS-8 controllers and no one has been able to explain the low load SOO to me.
7. "who is in control of the staging?"
since these are packaged central plants they are supposed to control themselves 100% - but again - they aren't talking to one another.
8. "are there isolation valves?"
no - triple duty valves prevent backflow and AAON said they wouldn't be necessary and they are correct as long as we can run all three machines at the same time at all times (on the secondary side).
9. "How is the staging set up?"
Internal to the packaged units (black box).
10. "design flow V actual through HX?"
329 gpm Constant Volume through the chiller barrel via the primary pumps. the CW coils in the facility are under low load right now so flow is lower than design through them (not sure if that is what you were asking - let me know if not).
11. "Is the decoupler correctly designed?"
Well - it is a part of the packaged unit so I assume AAON did their homework.
Thanks for the input - I don't expect to be able to trouble shoot these things in here - I just don't think its right that you would let a single compressor run until the head pressure was so high that you have to ramp the condenser fan VFD to 100% and then kill the condenser fan because head pressure drops so fast....seems like a good way to end up with a slugged compressor.
Likewise it doesn't seem right that a central plant with PRI/SEC loops wouldn't be able to accommodate low load conditions - especially considering that the rate of supply water temp increase could be predicted and controlled to. I need a reality check from someone besides the AAON sales rep.
Thanks again for the input.
Easy. Don't turn them all on when you don't have much load. Figuring out how to do it with the system you have is the trick. Can you simply lock out on OSA with each unit having a different lockout setpoint? Maybe read an average AHU valve opening and use that? Stage the number of systems by the return temp? Anything that gives you some indication of what the load is like. Secondary pump speed? Bypass opening? Look around and see what kind of variables exist in the DDC and go from there.
Sounds like something designed by somebody in an office drinking a latte'.
God Bless our Veterans
God Bless the USA
Although it does sounds like there may be an issue with the tuning of your condenser fan PID loop I'd say it's the least of your problems.
Why does the sequence from day 1 call for all 3 machines to run simultaneously? Reading you information it sounds like there is a flow meter in the secondary loop. If that's the case any halfway decent BAS guy should be able to set up some staging for you. You should run 1 chiller until the secondary loop flow exceeds your known 329 GPM, of primary flow, then start the second. Then start a third when the secondary GPM exceeds the flow of 2 primary pumps. This works assuming your secondary pumps are trying to maintain a differential pressure at some point in the loop. Even with no flow meter this staging can still be set up with a temperature sensor in the secondary loop supply.
Is there currently any external/remote enable or start/stop for the machines???
Great thoughts guys - and at any other facility I would agree that externally staging the machines up would solve the problem Buuuuuuut..... this is a mission critical facility and the sequence calls for all three machines to run in the name of redundancy. The engineer doesnt want to take the chance that the BAS will have a hiccup and fail to enable a machine - the remote enable disable was intended to be used for duty cycling only when the 4th machine is added next year. My hands are tied by the sequence. I can change inputs to the chiller but the engineer and the owner both expect the sequence to work this way because Trane multistack said they could meet it and they were VE'd out. I know that it can be done - my secondary pumps at their minimum VFD setting are still flowing what I need with all the machines up. I dont think that low secondary flow should be causing the freeze lockouts. It seems like the machines just dont have a good internal sequence for low load conditions. AAON has had the sequence all along and should have accounted for low load situations, fundamentally though - air cooled chillers with a PRI/SEC setup shouldnt be locking out due to low load. If the condenser fans would start at 20 or 15% and slowly come up - we would be rejecting heat more consistently instead of dumping a lot quickly by jumping from 0 to 100%. I dont know - I just have to hold AAON accountable for meeting the sequence somehow.
Randy and NW pretty much said everything that can be said - I would tell you the same things. If your on-site chiefs won't let you cycle down as needed, then you're going to have problems. As far as the AAON control issues, I'd say it's time to make them come earn their money, 'cause I agree that the cond fans should enable long before a pressure that would create a 100% call, and the refrigeration systems should cycle off completely before even reaching a low temp point where flywheel effect on the cooling from the hx's would get you anywhere close to 35*. All that said, I really believe that your engineers are creating unwarranted problems by not allowing the BAS to do it's job. Why have it if you won't allow it to work for you?