AHU / Retun air fan speed control 1

[remp]

If I have 1 AHU for one return air fan in a VAV system is it ok to ramp the supply air fan and return air fan up and down prportionally.? If so if I have 2 AHUs and 1 return air fan is it ok to do the same...e.g If the 2 AHUs (same size) ramp down by 50% can I ramp down the retunr air fan by 25%.??

 

[lilliput1]

For VAV systems the supply fan VFD is modulated to maintain a setpoint duct static pressure about 2/3 the worst run. You can have multiple pressure sensors & have fan control for the worst case. The supply CFM is monitored (air measuring station in supply duct with appropriate straight run, or at each inlet of the fan. The return air CFM is similarly monitored but in addition the return fan VFD is is controlled to maintain the same differential with the supply. This differential is equal to the design minimum outdoor air. If minimum OA is 4000 CFM then the return air CFM is controlled to be = supply air CFM - 4000. The minimum outdoor air CFM is either the amount required for ventilation (typically 20 CFM/person) or the sum of all exhaust + Bldg pressurization to offset infiltration (typical pressurization = .05 CFM/SF), whichever is the maximum.

 

[imok2]

billyq: I bring this to your attention because of the instibility of VAV return fan tracking systems.

"403.3.2 Common ventilation system. Where spaces having different ventilation rate requirements are served by a common ventilation system, the ratio of outdoor air to total supply air for the system shall he determined based on the space having the largest outdoor air requirement.

403.3.3.1 Variable air volume system control. Variable air volume [VAV] air distribution systems, ……. shall be provided with controls to regulate the flow of outdoor air. Such control system shall be designed to maintain the flow of outdoor air at a rate of not less than that required by Section 403 over the entire range of supply air operating rates."

 

[lilliput1]

Fan tracking precisely does 403.3.3.1. DDC controls incorporate adjustments to control speed of response and degree of control action to allow tracking it to be stable, avoiding overshoot and hunting.

 

[jburrows]

Fan tracking can work if both fans are exactly the same and care is taken to ensure that minimum outside air requirements are met. However, in my experience it is often unstable and leads to building static pressure problems, especially if two supply fans are being used with one return fan or vice versa. Trane has an engineering newsletter that discusses a few return fan control strategies:

http://www.trane.com/commercial/library/vol31_2/ADM_APN003_EN_040302.pdf

. (If the link doesn't work, go to Trane's webpage and look up commercial building pressure control in the engineering newsletter archives.) Controlling the return fans to maintain a pressure setpoint and then modulating relief dampers to control building static has provided good results here at the university campus I work at. Then return dampers are modulated with outside air dampers to maintain the mixed air temperature setpoint (with minimum outside air settings on the outside air dampers). The newsletter has a few other suggestions, but this basic idea has proved to be quite effective.

 

[imok2]

Lilliput1, In a paper by Gil Avery reprinted from Heating/Piping/Air Conditioning February,1992 he discusses how control loop interaction can cause instibility in VAV systems. Fan tracking seems to be the major culprit and he goes so far as eliminating this type of control strategy.

. Trane recommends(jburrows) Note: An intake airflow sensor controls ventilation more accurately than supply/
return airflow tracking. Monitoring
intake airflow also costs less and is
usually more reliable than using an
injection fan, pressure sensor, and
variable frequency drive.

 

[lilliput1]

We have always used fan CFM tracking controls for Pharmaceutical, lab & hospital application in lieu of direct space pressurization control to setup relative space pressurization / direction of transfer air. If you control by direct space pressure sensing, your system will be veru unstable because of dor openings, stack effects, multiple fans, hunting & fighting of controls. I have actually witness setup of fan CFM tracking and observed that the DDC programing allows you to vary speed of response as required to provide stable operation without overshoot hunting. You must be sure to use air measuring station sizes within the appropriate velocity approx 2500 to 500 CFM or 5 to 1 turndown, correct straight run upstream & downstream, & use DP transmitters to have minimal signal loss.

 

[remp]

Has anybody read the Trane article mentioned by JBurrows baove.

http://www.trane.com/commercial/library/vol31_2/ADM_APN003_EN_040302.pdf

Can anybody tell me how do I work out the return air plenum set point pressure.?
Is this method ok for multiple AHUS with a single Return air fan?
Is it ok to position the space pressure sensor in the return air shaft...? Shaft serves 12 stories.

Thanks.

 

[jburrows]

billyq,
The return air plenum pressure needs to be measured on the discharge side of the return fan. I have never done this with one return fan and multiple supply fans, I've done it the other way around. It should still work though. The plenum (or duct, depending on your set-up) static pressure after the retun fan should be maintained just slightly positive. It needs just enough pressure to overcome the pressure losses through the duct to the return side of your AHU's and across the return dampers. If this is an existing building, you should check to make sure that the pressure drop through the return system is about equivalent to the loss if the air is being relieved. (I assume your return fan is actually a return/relief fan.) If this is in the design stage, size the ducts so the pressure drop will be about the same for all operating scenarios. In your case with multiple AHU's you should find the pressure drops along the paths to each AHU and set your setpoint for the highest pressure drop that will be encountered. You want the mixed air chamber of the AHU to be as close to neutral pressure as possible. There may be other ways of doing this, but this particular method has worked for me and I think it will work in your case too.

 

[remp]

JBurrows,

I was thinking that during the commissioning the return/relief fan discharge static pressure (set point) would be maesured as follows;

Put the building into full outsdie air (since this is when the return air fan will have to work hardest))

Open the relief damper fully

Ramp up the relief/return air fan till the pressure drop between the space and outsdie is at the desired pressure differential

Now record the discharde pressure. This is the set point pressure and will be entered into the BMS.

What Do you think?

 

[imok2]

jburrows
What do you think about multiple sensors on the return air and input to an averaging control to maintain a better RA presure control?

 

[jburrows]

billyq,
That's what I would do as far as getting your pressure setpoint. After you check the system in full outside air mode, you would probably want to put it in full return air mode and check the performance of each air handler just to make sure that adequate return air is available to each. This may be what imok is referring to with his multiple sensors question. Is that right, imok? Having a sensor in the return duct going to each air handler would probably work, but I don't know if it would improve the performance enough to justify the extra complexity of having multiple sensors. Does anybody have any thoughts on this?

 

[remp]

I think putting too many sensors in the building makes it too complex.
I have put a lot of tought into this. (???) In the building Im doing at the moment I have a 24 storey building. A typical AHU serves 12 floors of say Western perimeter. Im going to put the indoor sensor into level 6 (half ways between level 1 and 12)probably in the ceiling space fairly close to the return air shaft in a fairly open part of the office. The return air shaft is sized with low pressure drop so the pressure on this floor should be fairly representative of all floors.
The thing to remember is to ensure during after hours operation of the A/C no matter what floor is activated level 6 must be activated also because it contains the indoor sensor. (I have shut off dampers on each floor.

Havent looked into outdoor sensors but would probably have 2, windward(prevaling wind) and leeward side of the building.
Putting the Return air fan sensor just upstream of the return air fan.

I have seen fan tracking in a building done lately by my company and its not working. The building is fit to explode. The lift door dont even close properly due to build up of pressure.

 

[imok2]

Gentleman you know that the stacking effect is different in a high rise building between summer conditions and winter conditions that's why I thought multiple sensors would give you a better average senseing pressure say like4,8,12 floor
I don't believe this is complicated just a little more expense. Billy have you discussed the complexity with a controls eng or is that your assumption, just curious. But listen has some one written a paper on this problem recently? I sure would like to read it, if so. Any way HAPPY TURKEY TO ALL!

 

[ChasBean1]

I haven't seen any indication (unless I overlooked) of whether or not there is return air flow control at each floor (not just volume dampering, but volume flow sensing and control dampering).

To answer your first post Billyg, change the last word in that post from "25%" to "50%" and the answer is yes, it is possible and common for open return systems (no return flow volume dampering). It's not the best control, as others have observed.

I would recommend one of three control scenarios for VFD systems, in order of preference, and obviously in order of expense:

1. Fully ducted with flow controlling terminal boxes. Boxes (supply and return) are controlled at the point of use. This system controls outside air to the building by summing the terminal box offset volumes. The supply fan should control to maintain supply duct static pressure and the return fan should maintain return duct static pressure only.

2. VAV systems without return air terminal boxes (e.g., open-shaft systems and/or mechanical volume dampers): The supply VFD should control to maintain supply duct static pressure. Ideally, the supply and return air flows are monitored at the AHU via air flow stations; the return VFD modulates the return fan to maintain a fixed return flow offset from the supply air flow. OA flow should only be monitored at the intake and used for control if there is a place to accurately monitor it, without turbulence and obstruction. This is seldom the case.

3. Supply VFD controls the static pressure in the supply duct; supply and return VFDs track in unison.

I wouldn't play with controlling plenum pressures. What if someone leaves an access door open? If the relief plenum is positive (say, 0.05 in. w.c., arbitrary) how do we know the that the OA intake pressure is right - it has to be negative to take in air, right?

Sorry, didn't mean to get too wordy. Thanks for letting me chime in... -CB

 

[remp]

Imok2,

Click on the piece below to see an artile by trane. I have since spoken to the controls people and a few sensors on more than 1 floor seems a better idea and average them out.

You are correct about the stack effect and the need to vary the pressure between winter and summer. My building is in Sydney so outsdie temp vary from about 15-30 degC during the occupied period. Im going to hold a slightly positive pressure, but location of outsdie sensors is going to be a problem. Will need at least 2 out side.

Also is attached a responce from Trane (same person who wrote the artice) on how to achieve the return air fan set point pressure.

Transe Responce:
####"When using a return fan, the highest pressure in the return air plenum (downstream of the return fan) must be sufficient to overcome the pressure drop in the relief air path (usually some duct, a damper and a louver) at maximum relief airflow. Most designers assume that maximum relief airflow equals design supply airflow less both local exhaust and local exfiltration airflow. This approach assumes that the supply fan delivers design airflow when the system enters 100% economizer mode and results in an slightly high relief airflow. Remember, space sensible load at the 100% economizer condition is lower than design space load, so supply airflow is actually somewhat less than design airflow.
Note: although it’s possible, the recirculated return path pressure drop (through the wide-open return damper, at design recirculated return airflow) is seldom greater than the relief path pressure drop described above.
The optimized damper approach is simple, logically, but must be tuned for each system. The logic is this: control relief damper position to maintain building pressure set point, and control outdoor air damper position to maintain outdoor airflow set point, control return damper position to maintain outdoor damper at or nearly 100% open, reset return plenum pressure set point to keep either the relief damper or the return damper at or nearly 100% open, and finally, control return fan speed to maintain the current return air plenum pressure set point. The logic is simple and results in the lowest pressure rise across each fan. The control system must be tuned, however, with damper motion speeds and fan speed reset rates that are appropriate for the specific building system, to avoid accuracy and stability problems.###"

http://www.trane.com/commercial/library/vol31_2/ADM_APN003_EN_040302.pdf

Regards

Billyq

 

[ChasBean1]

First, you’re right about “Transe Response,” because it put me in a Trance. The first half of the gibberish is Fluid Mechanics 101 with legalese and wordsmith. We still haven’t established what your system is. Return airflow control at the point of use (a ducted system with terminal boxes), or plenum return?

Billyg, before you post another Trane rep’s pontification, please answer the above, regarding what constitutes the system.

Please note these points:

• The control system indicated above is ridiculously complex and will not work in this world we live in now. Obviously, I shouldn’t go up against such a well-worded sequence without backing it up:

- “Control outdoor air damper position to maintain outdoor airflow set point” – where does the flow station go? If we want to read accurately there is usually not a good place for it.

- “Control return damper position to maintain outdoor damper at or nearly 100% open” – Now we have the outside air flow controlling the outside air damper and the return damper controlling the outside air damper to 100% open… which one wins?

- “Reset return plenum pressure set point to keep either the relief damper or the return damper at or nearly 100% open” – Which damper, return or relief? Which has higher priority? Resetting the return plenum pressure setpoint (not the actual pressure) means we also have a moving return plenum pressure target (not to mention the guy who left the door open). But wouldn’t this pressure vary as the building pressure varies due to modulation of the relief dampers? Does this override building pressure control, or vice-versa?

- “Control return fan speed to maintain the current return air plenum pressure set point” – wow. Well, when the relief dampers open because of the building pressure sensors sensing too high a value, the relief plenum pressure drops and the return fan cranks up. When the return fan is at full speed, the building pressure drops due to higher relief airflow. This will cause the building pressure sensor to command the relief air damper more open… woops, we’re already at 100% relief damper postion, but the pressure is too low in the relief plenum, so let’s crank up the return fan. Woops, we’re at 100% already…

• No controls contractor I’ve ever worked with would be able to execute this sequence successfully, which might mean that you need to buy the Trane package. Even then, your Trane support will consist of various local mechanical firms that don’t have a clue about this level of control.

• The sequence is at the mercy of building volume, damper and drive speeds, timed toilet exhaust setbacks, damper blade leakage, plenum volumes, location of sensors, and a host of other unknowns. Give yourself a break! Please re-read my last post.

Godspeed-CB

 

[lilliput1]

Right on Chasbean1! I'd like to hear from people that say flow tracking do not work. Have they ever found out exactly why not. Because if you thing about it, it should work and I personally know it has. Otherwize our consulting engineering office will be swamped with complaints because we always put in fan tracking controls for all VAV applications.

 

[imok2]

lilliput1, Question! Is there any reliable technical information avialible to aid the design engineer in sizing economizer dampers for VAV systems that utilize return fans, because any engineer faced with a legal liability because of a sick building would find himself in an embarrassing position if he was questioned by the prosecuting atterney about sizing the ventilating dampers. An admission that there was no technology avialible for damper selection or that he left the damper selection up to the control contractor, certainly would not help his case.

 

[lilliput1]

The supply air fan must be able to suck the air from the OA intake. Hence the OA duct must be at negative pressure & pressure loss here must be part of the supply fan duty. Now the return fan being at positive pressure at discharge (the relief air ductwork pressure loss must be handled by the reurn fan) will tend to push the return air to the supply fan suction, not allowing it to get outdoor air. Use of Fan tracking control, minimum OA damper & maximum OA damper control and balancing would be critical. OA damper must be sized to make sure pressure drop is not excessive & velocity is not excessive. If the damper is right at the louver, it must be sized at 250 FPM face velocity because louvers are about 50% free area & 500 FPM free area velocity is the design to avoid water carry over. If the damper is in the OA ductwork, it can be same size as the duct sized for 0.08 to 0.10 inch wg./100 ft of duct pressure drop but not exceed 2000 FPM (for noise concern). and control &