York YS varying evaporator flow
We have two York YS screws that are in a primary secondary loop system with a hydraulic decoupler. We have also been trying make this system as efficient as possible. Our biggest problem is that either the primary loop over pumps the secondary loop and return water to the chillers is too cold (low delta t. Chillers won't load) or the secondary over pumps the primary and makes the supply water warmer than the discharge leaving the chillers.
We can remedy this by varying the flow through the chillers. I was thinking of a Vfd on the evap pump. But our service guys say that you can't do that with screws. The engineering manual shows a minimum flow and a max flow through the evap, so i was thinking it would be ok as long as I stayed in between those flow numbers. JCI won't tell me why you can't vary flow they just say you can't. Can someone enlighten me?
you can vary the water flows but other issues can creep up such as loading/unloading issues among others. how do you plan to control your primary water flows? control to what value? your secondary water pumps should be the ones that control the overall water flow. if you vary the primary, you would usually remove the secondary chilled water pumps and the decoupling loop...this would give you a primary chilled water system.
varying your primary water flows won't fix an undersized chiller plant or a poorly controlled building.
"If you pull one more stunt like you just pulled with Tommy, you won't have to get on a plane because I will personally kick your ass from here to Korea!" - Best of the Best
Thinking of trying to make the primary loop match the flow of the variable secondary. Secondary is controlled by diff pressure further out in the loop. It would hard to get rid of the secondary pumps the way our plant is set up and we also have an air tank. Not saying it can't be done. I just think my idea may be simpler.
But my Jci guys say we'll blow up our chiller if we try
When systems are converted from primary/secondary the primary pumps are eliminated and the secondary pumps are set up for a minimum/maximum flow through through the evaporator using the VSD that is on those pumps. Typically called a Hartman Loop. There is a great deal of things going on in that system.
I have never heard of a variable primary system utilizing screw chillers unless they are screws with VSDs. The YS isn't a candidate IMHO. Every VPF system I have seen and worked on has had a centrifugal with a VSD.
The min/max flows that York publishes in their engineering guides are to show flows at 3'/sec and 12'/sec. In a VPF system the minimum flow in the evaporator is actually set at about 1.5'/sec.
I think what I would do with a primary/secondary system with a screw chiller is "float" the leaving chilled water setpoint. On colder days when the primary is over pumping raise the CHW setpoint so the cooling coil valves will open more. When the secondary starts over pumping it's time to bring on another chiller. It would be possible to bring on another primary pump as long as the leaving chilled water valve is throttled to keep flow under the max. of 12'/sec.
I have seen some of these balance problems solved somewhat by introducing either a fixed orifice or an adjustible orifice (butterfly valve) installed in the decoupler. What really needs to be looked at is records of flow rates. Read and record flow rates on both sides at different load spectrums. Then determine if introducing a slight orifice in the decoupler will create just enough pressure drop to cause the chiller(s) loop to pull and push into the load gathering loop. Gathering data can help an engineer to decide how to do it. Adding the orifice by the way is not energy efficient. The reason why any chiller mfg. will not support variable flow in the cooler is because it has to be a real time calculation of cooler conditions occuring. Delta P across the cooler is a good indicator, but unreliable in the long term. For example if the chiller were relying on delta p to control the load algorithm the pressure transducer(s) and it's connections MUST be accurate always at all conditions. Most chiller mfg. do provide a variable flow routine in the newer machines or their software upgrades. Those flow algorithms assume a reduction in the condensor medium temperature also with the cooler load. In an air cooled screw chiller that is not a controllable compressor compression ratio plot(per se). I think your exploration of controlling the decoupler flow is the least expensive move, otherwise you would need to reevaluate the chiller plant. Todays variable flow cooler systems eliminate the primary/secondary loop and vary the entire loop flow rate directly thru the chillers.
When you drop down to 1.5 ft./s in a vpf suspended solids can fall out of the chilled water in a close loop system making evaporator tubebrushing and maintenance more frequent
I know the guy who know's the "Chiller Whisperer"
The recommendation by KNEWYORK bears good advice. If you raise lvg chw water temp a few degrees out of primary. That would raise the secondary lvg and keep chill water valves from staying throttled down too much(overall). In any event it would allow you to capture flow rates data.
You could vary the evap flow with no problems if you set it up right, use a chilled water reset like knewyork said. You just have to be careful when slowing the evap pump down and make sure it is done very slowly so it gives the chiller a chance to unload. Figure out what frequency on your drive is your min flow and set that as your min frequency. I always set my min flow a little high. If your secondary flow drops below your primary flow adjust your chilled water reset. Control your primary pump off of the difference in the primary and secodary water temp. (ex. #1 primary temp 45, secondary temp 48, diff = 3, speed pump up try to maintain 1.5 diff) (ex #2 primary temp 45, secondary temp 45, diff = 0, slow pump until diff is 1.5) The 1.5 diff setpoint has work well for me, you could make it water ever you want but there has to be a difference. This will make your system have slightly more flow on the secondary side, so you may have to adjust the chilled water set point a degree or so if the secondary water is to warm.
Just remember the control of the evap pump has to be slower than the chiller loading and unloading or you will have problem. Hope this helps
This would require adding VSDs to the primary pumps. Typically only the secondaries have VSDs. Also I seriously question the ability of the YS to successfully turn down with varying flow. Not saying it can't be done, but I think one might run into more trouble than it's worth.
Originally Posted by chad_nc
VPF systems tend to count on the cooling coils valves being open further than a normal scenario and then the CHW setpoint is raised. The systems I have worked on typically run their CHW temps around 52-53*F. Lower setpoints are allowed if space temperatures can't be maintained with CHW valve at 100% at those setpoints.
Pumping more water through the secondary loop would use more power.
It may be more efficient to pump more water than run 5 chillers at 50% load. I'm not sure, but using less electricity is our goal.
We had an engineer come in and do a survey on the chilled water system. His remedy to the problem was, install PIV valve at all air handlers, sequence secondary pumps to only pump enough to maintain the min DP, and sequence our chiller by the flow in the secondary loop.
We have installed the PIV's. I am working on installing more sensors in the secondary loop to assure we are only pumping the water that we need. But I am unsure on how to stage the chillers to maintain the flow of the secondary loop without varying the primary loop somehow.
We have 1 centrifugal chiller a YK. And I can vary the evap flow through it. But only varying flow on 1 chiller would pretty much make us have to run it 100% of the time, which just isn't realistic.
I understand what you are saying about the temp reset, but I just don't know if that will save on electricity.
I wish I could throw a stick of dynamite in the chiller plant and just start over. But new homeland security regs have pretty much made that an unrealistic solution as well.
You have to look at where to save energy. Raising the setpoint results in less chiller hours. The biggest motor in the system is the chiller. Any efforts to reduce the use of that motor makes the most sense. I've run into the same argument with regard to turning cooling tower fan motors at higher speeds to reduce the condenser water temperature for chillers with VSDs. The savings of the chiller motor dwarfs the slight increases of energy consumption of the fan.
I believe the same to be true of your situation. Your secondary pumps have VSDs, no? Running a pump wih a VSD at a slightly higher frequency is not the end of the world especially if the chiller is running for less hours. Investigate the Hartman loop I think you'll see what I'm talking about.