VAV System Control Sequence
I would like to get some feedback on the best way to control a way over-sized
Cooling only VAV system.
The System has about 15 various size vav box's. The original sequence calls for 55°f supply air at 1"wc static pressure. When we set this system to auto the average space temperature gets way too cold. The vavs don't control very well if we set the minimum air flow below the .2"wc velocity pressure cut-off.
For a quick fix we started reseting the supply air temperature set-point up based
on the return air temperature. if RAT = 70 then SAT = 63. if RAT = 76 then SAT = 55.
This helps to some degree, but some zones have the damper at 100% open and not enough air flow with fan at full speed.
Mr. Engineer is now worried about humidity and mold issue's.
Anyone have any ideas
Try seting up your box's as pressure dependent and not use the the flow setpoints. Use space temp to control the damper.
This should allow your box's to completely close if this is acceptable.
I would adjust the flow setpoints to whatever is needed before I'd convert to pressure dependent. Sometimes the setpoint might be 0 CFM. Engineers regularly specify VAV systems with minimum flow setpoints that are 2/3 of max setpoints. That doesn't leave a lot of room for temperature control without reheat. I'm sure it's because of ASHRAE standards regarding makeup air, but the tenant doesn't care. I use a weighted average of terminal load to control VAV air handlers. Discharge air control with DX cooling is just dumb - it tends to cycle on the min and max run times much of the time. If it's chilled water, you're all right. If you don't have a terminal load point, you can use average room temperature or some other indicator of demand for heating and cooling.
What do you mean by "not enough air flow"? Damper 100% open and not meeting flow setpiont? Were there enough air flow before the SAT change? I don't see any connection between SAT change and not enough air flow.
Or you are trying to say, with the SAT change, VAV boxes in some rooms can not satisfy the temperature even with damper at 100%. I would assume the actual flow is exceeding the designed max flow, otherwise you need to raise the static pressure setting.
Originally Posted by AutoPilot
AutoPilot, I didn't see you specify what type of cooling you are using. Is it DX, as assumed, or chill water?
What I've done in similar situations is drive the damper position on the VAV based on the space temp. vs. setpoint and kept the damper position from 20-100%.
Do you have any humidity readings in the zones or return air? Do a dew point calc and make sure you keep the coil below dew point and you should alleviate your engineer's mold concerns.
Look at this post regarding a discussion about static setpoint reset:
Does this help?
How do you get 63?
There are 2 critical temperature points:
With T1, all rooms with oversize boxes are in comfortable temperature zone, one of them is just at the .2"wc velocity pressure cut-off, others are higher.
With T2, all rooms with normal boxes are in comfortable temperature zone, one of them is just at the 100% damper pos, others are lower.
It seems T2 < T1 in this case. Where is 63 relative to T1 and T2? And do you know how big the difference is between T1 and T2?
Originally Posted by AutoPilot
The unit is chilled water.
The Supply air set-point is adjusted down as the return air temperature rises.
RAT a 70°f will equal 63°f SAT, RAT at 73°f will equal 59°f SAT, RAT at 76 or above will equal 55°f SAT. just simple reset. I'm not sure why 63°f was selected as the max SAT set-point.
Some of the zones are at full cooling as the dampers are at 100% open.
What I don't understand is how they could be low on air flow. I.E. the CFM set-point is like 2200 CFM but box is getting 1795 CFM. When the TAB was done all zones were getting the correct air flow with 55°f SAT and 1"wc static pressure 2/3 of way down the duct from the fan.
I like the idea of converting some of the zones to pressure dependent, but want to keep some kind of minimum air flow.
Thanks for your help
>What I don't understand is how they could be low on air flow. I.E. the CFM set-point is like 2200 CFM but box is getting 1795 CFM. When the TAB was done all zones were getting the correct air flow with 55°f SAT and 1"wc static pressure 2/3 of way down the duct from the fan.
Diversity is my guess, ie the AHU total CFM does not equal to the total of the VAVs max CFM. Nothing strange there, at least in our area it’s pretty common. Have you checked these numbers? Did the TAB report have anything that smelled funny?
>if we set the minimum air flow below the .2"wc velocity pressure cut-off.
Where are you coming up with .2” as the min cutoff of the VAVs? Typically .2” would be a pretty damm good flow on a VAV box, do you mean .02”??
Do all the zones have similar cooling needs, or are interior and exterior zones mixed together on this unit?
I agree with orion. Do you know what the diversity factor is? Also has anyone verified that the supply duct static pressure sensor is reading correctly. You stated it is in a good location (2/3 down stream), but could it be out of calibration reading high? Could you tee a magnehelic gauge just to confirm static pressure reading is accurate.
To orion’s point about the diversity, I also see this frequently. Typically if the DF is a problem it might be worse on full cooling load days.
What is the system static pressure setpoint?
Who setup the VAV Box parameters and did they work with the TAB. I do not know what control system this is, many today are using Application Specific Controllers for VAV's, are you sure the inlet duct size, K factor, etc.. are correct for each VAV Box?
What type of VAV Boxes are these? (fan powered, parallel, series, reheat, etc..)
Are all the boxes the same, or are interior and perimeter different?
Orion242 you're correct the minimum cut off is .02"wc... typo on my part.
Static pressure is 1"wc new and calibrated in large main duct in good location with magnehelic Dwyer 604d I think (don't like em terminals are to small and I can't see in the dark). TAB reported all rooms between 68° and 72° with fan at 48hz. 1"wc at SAT 55.2°f while unoccupied
during TAB with fan scheduled on at 7am and off at 5pm. They set all box Kfactors, min, max airflows etc. Only reported one room with noise problems.
TP.. All box's are cooling only single damper with zone sensor. floor layout is elevator lobby, two large waiting room type areas at each end of elev lobby.
No interior/ exterior.. All have windows with view.
Some good sized meeting rooms and the rest are office areas. Elevator lobby and large waiting areas have good control. meeting rooms and offices get damn cold if SAT is at set-point 55°f.
I noticed on a Saturday with a few people at the site, things look good.
RAT @ 71 (warm for this system)
SAT @ 62 Fan 60hz static @ .68"
warm zone @ 74.6 (only zone with damper at 100%)
cool zone @ 66.2
4 or 5 zones satisfied at around 50% cooling. all the rest are at minimum flow
with temperatures well below set-point. (2°f or more below). Tenants control
room set-point and most have them set at max 78°.
On a VAV unit:
I reset my Static Pressure Setpoint based on VAV Damper Position, and reset the DAT setpoint based on terminal load. I also have time limits on how fast the setpoints can change. Example: Static setpoint can not change more than x-"w.c. per x-time, etc...
Without going nuts like adding another variable in there for "riding the fan curve", that about as good as it gets IMHO.[/QUOTE]
codewriter.. Could you please post an example algorithm
>Some good sized meeting rooms and the rest are office areas. Elevator lobby and large waiting areas have good control. meeting rooms and offices get damn cold if SAT is at set-point 55°f.
Plenty of dissimilar cooling needs…just what a cooling only VAV system needs…
>I noticed on a Saturday with a few people at the site, things look good.
RAT @ 71 (warm for this system)
SAT @ 62 Fan 60hz static @ .68"
warm zone @ 74.6 (only zone with damper at 100%)
cool zone @ 66.2
Something seems odd with these readings at least compared to the TAB numbers. The fan is maxed out only holding .68”, which to me says the system is under a high cooling demand, yet your DAT set point is only 62. All this on a weekend with light occupancy? If you have a good deal of boxes with 78F setpoints most of the boxes should be at min CFM and I would expect the fan running slower.
>TAB reported all rooms between 68° and 72° with fan at 48hz. 1"wc at SAT 55.2°f while unoccupied.
Funny how they report the fan running so much slower with 1” of static and reasonable space temps @ 55F DAT. Where boxes closed during the unoccupied time or is this a drive by balance job?
>Try seting up your box's as pressure dependent and not use the the flow setpoints. Use space temp to control the damper. This should allow your box's to completely close if this is acceptable.
I think this a good bet, you cannot deal with the dissimilar cooling needs any other way. I would still use the flow setpoints for the max CFM, and make min CFM 0. Otherwise you could have diffusers sounding like a jet engine. I’m sure the freezing office occupants won’t complain about min OA if you can fix the temp issues.
-- Is this a new system? Or one that's been in operation for a number of years?
It makes a difference. With a system that's been in operation for a while I'd be wondering if some of the more critical components were working adequately. ie Testing duct press sensor, checking supply fan for belt slippage and full speed amp draw. Etc
I'd also be wondering if building owner/operators have changed some of the original SPs . Do you have the data for the original min/max air flows programmed into the VAV controllers? Have you compared those to what's in the controller's now?
You say in a later post that the TAB results you have show the system to be providing adequate cooling, with fan power to spare. What were the ambient air conditions on that day? Was it warmer or cooler than the conditions you're experiencing as you troubleshoot this system?
-- You mention a VP cut-off of .02"WC. Now I'm not quite sure what you mean by "velocity pressure cut-off". I'm inclined to think you mean that low point in VP as sensed by the VAV flow meters and their associated sensing tubes (pitot tube or similar device) at which they might be expected to give adequate information about the current air flow.
If my guess is right about what you're referring to, where did you get that .02"WC number?
In most reasonably new systems, that's fairly high. The .02"WC min VP sounds more like a number often just cut and pasted upon the chart on a VAV manufacturer's box, usually reflective of what old time pneumatic controls could handle. Many DDC system VAV controllers have air flow inputs that can handle as little as .004"WC VP.
You mention that the controls don't control very well if you reduce min air flow to a point where the VP drops below .02"WC.
Did you actually measure that, or is it a presumption?. If the pitot tube is an older design, and if the VAV controller's flow sensor is an older design, could be true. But it's not what I'd expect in any system less than ... say ... 5 or 6 years old.
Do ALL of the VAVs control air flow poorly if min air flow is reduced from current settings? Or only some of them?
If only some, I'd suspect problems too turbulent of a flow at the box inlet where the pitot tubes are installed and check for that.
Highly turbulent flow JUST BEFORE the pitot tubes (or whatever) can be a problem that results in poor control. Some would debate that yah need 3 to 10 duct diameters of straight duct before a flow measuring point to get accurate data. True enough as far as that goes. But we all also know you're not gonna get that most times for a VAV box. In my experience, if yah can get 6" of straight line before the pitot tube/ring you should be able to get usable, if not perfect readings, as long as there are not really sharp bends just before that.
Another possibility as to why you might have troubles at lower VP. Many DDC type VAV controllers have a setting for hysteresis within the air flow control loop. Usually some place where you can enter some CFM amount that'll tell the controller, "Don't do anything if the air flow change is less than this amount". If this value is too low, damper actuator hunts excessively, and steady state operation is difficult to achieve.
And some controllers allow you to select either Current Value of CFM (real time readings) or Average CFM as inputs to the control loop (and as the number to be displayed on the head end screen). If such is the case, always go with Averaging, unless you have a specific requirement otherwise.
I'd quite expect to see more fluctuations at low flow conditions when looking at real time CV. But wouldn't be too concerned about them as long as the Average looked pretty good, space temp was being maintained properly, and the damper actuator was not hunting excessively after control adjustments.
Now, if this can't be done reasonably, then there probably is some issue that needs to be addressed and corrected if possible. Excessively low VP at the sensing point (pitot tube or similar), in the case of an oversized VAV box, can often be fixed by installing a throttling collar (small section of ducting of a reduced size) just before the VAV box's pitot tube. Or by installing a better pitot tube/sensing ring. There are several models available from different manufacturers, offering better signal amplification and better averaging.
Some "Better" velocity pressure sensing tubes I've used and had success with are the Kreuger LineaCross flow probe, Enviro-Tec FlowStar airflow sensor, and the Nailor Industries Ampliflow sensor. The Ampliflow is nice, and useful. Not as accurate or sensitive as the others. BUT ... being a simple "straight stick" type pickup tube it is fast and easy to install. Where it produces adequate results, it's a good, fast, cheap (in labor) answer.
-- You mention using a reset on the AHU DAT control loop. Reasonable. But you mention that when reset is in effect that some VAV boxes go wide open, but can't supply max air flow even with the supply fan at 100% and the associated spaces get too warm. While many other spaces get too cold at the same time even with min air flow.
First thing that comes to mind is that the min air flow SPs for those "too cold" spaces is too high.
I'd be asking myself if those min SPs NEED to be so high. Fact is, some design engineers cut and paste design info from one project to another, without actually working out ALL the numbers for each and every VAV box. In relation the the actual space served. In previous projects where I've found problems, have actually had engineers admit that.
In other cases, have seen engineers who just used "canned" rules of thumb for sizing stuff or establishing SPs. Which might or might not work in some particular application. If asked, will insist that they've "always" used those numbers and it always worked for em. End of story, now go away and stop bothering me.
And then yah have the situation where the engineer's numbers are good ... if the actual conditions and loads are what he thought they'd be. But they aren't. Actual owner changed the usage of the space from what design engineer thought it'd be. Or maybe design engineer was told 10 people were gonna occupy an office. And they do. But at any one time only 5 of em will be in there, except for infrequent occasions. I see this a lot.
So maybe yah do need to take the approach of reducing those min allowable air flows.
Worth a try, at least. If this is an existing building and equipment, project complete and sold.
If a new project, I'd drag the original design engineer back on site and tell him, "Okay, YOU tell me what's wrong here. I've set it up the way you specified. It ain't working. So review your numbers and tell me what yah want me to do."
But for existing equipment, building that's been occupied for a while, I'd at least try the reduction of min air flow approach. I'd record all existing SPs first so I could restore system to the way yah found it if this approach doesn't work.
Another situation is where there are the right number of folks, but other loads were reduced from from original designer figured them to be. ie More efficient, less heat producing lighting. Customer added solar blocking curtains/blinds. Those two big copiers on original plans were moved elsewhere. Etc. This is a touchier issue. Yah still want adequate fresh air. But you haven't got a reheat coil. Big problem. You might figure out the numbers for yourself and see if you can come up with a new, smaller min and still get adequate fresh air for normal occupancy. Or yah might have to resort to demand based ventilation using CO2 monitors. Or you may need to suggest duct mounted reheats for that space. Or ... you and customer might just have to bite the bullet and tell occupants to live with too cool conditions or be willing to live with less fresh air.
Did have a project a couple years ago where I ended up setting min air flows way lower than the original design features. Originally the spaces served, couple floors of a large skyscraper, housed engineers, tech support people, sales people, and office staff for a data networking services company. Lotta floor space, and ... originally ... quite a lot of people. But technology moves on. Now, networking design and services requires a lot fewer people, and the equipment needed is both smaller and more energy efficient. With new conditions, I had to set min air flows WAYYYY down. Left the max's as is, just in case.
Sometimes we come up with projects where we have to resort to a two-tiered approach. Using an occupancy requirement. Either based upon actual occupancy sensors, or some time schedule scheme. VAV controllers have their "normal" min and max air flow SPs. But if space is in a scheduled "unoccupied" time period, or if actual occupancy sensors don't detect anyone in the served spaces, we have the automation system change the allowable minimum air flow to zero. More and more necessary these days with a more mobile workforce, flex time working hours, etc.
Something to think about.
Anyway, I have no real answers for you. But ... maybe ... have given you some things to think about. But probably not. I'm neither the brightest nor the shiniest apple on the tree. I've been following this thread in hopes you'll get better info and ideas from others, so that I might learn something.