Question about ventilation system

[xj25]

Hi,

We are doing some refurbishing of a ventilation system. They use now a line of 8 small (200-300W) fans that suck from a rooftop conduit-chamber with filters.

Two of these fans have a smaller section from where they suck, so they work with a higher presure loss and unbalance the system quite a lot. We found that, as it is now, is really difficult ro regulate because any intervention in one fan affects a lot to the nearbys.

We have been proposed to do a common in-line conduit with two radial fans pumping from each side to the conduit. The section of the conduit will be reduced from start to finish. The conduit has some kind of "holes" (* in the draw below) along to allow the flowing of air.

Schematic in plant (from top) view:
| |
fan1 fan2
| |
----------------------------------------------
------ * * * * *
-------------
* * * * * ------------
----------------------------------------------

Another option is to use a regular conduit and pressurize it from both sides (without section reduction):
| |
fan1 fan2
| |
----------------------------------------------
* * * * * * *
----------------------------------------------
With this option I do not see how we can adjust to equal flow output along the conduit (this is important). It seems for me (intuitively) less optimal than the former solution.

We are not experts in this field, what do you think that is better?,

I am also looking for a practical book in this topic (ventilation conduit sizing, adjusting, pressure loads etc.), could you reccomend me something? All that I found is very much directed to thermal HVAC calculations, more than ventilation....


Thanks in any case for reading..

 

[KiwiMace]

The first idea looks like throw away to me. The second one looks like it can be manufactured if I am interpreting your drawings correctly.

You require even flow through the holes, so the conduit (duct) needs to be sized large enough to negate any pressure drop from air flowing along it. This is the basis of plenum design. The pressure drop of the combined discharge needs to be significant with respect to the fan so that the fans are capable of maintaining the pressure in the plenum. Or put another way, the fans need to be able to outgun the discharge requirements or they won't maintain pressure in the plenum and you'll get preferential air paths.

Secondly if the flow through the holes needs to be very precise, you should consider constant speed on one fan and variable speed pressure control on the other.

 

[xj25]

Hi, I get the idea of the pressurized "plenum" and the regulation proposed. We need even flow, but not so accurate (just improve their present system).

But, why do you think is not sensible the first proposal?,
This is the usual duct distribution seen in a lot of commercial buildings (from big to smaller duct diameter). I thought that it was with the objective of keeping flow at each exit constant.

Btw could anybody recommend any book on the subject (duct distribution and sizing)?

Thanks!

 

[willard3]

You should hire an Engineer who understands ventilation systems.
Don't try to bootstrap this.

 

[xj25]

Hi thanks for the feedback, we are hiring a consultancy to help us design the definitive system.

We have been doing some measurements of the existing system to be able to stimate the present working point and know exactly from which point do we start on.

We have done 4 measurementes in one of the representative fans:
1 - Q(flow) measurement with anemometer and "measurement box" 1m long. Details: measurement at air impulsion, we see quite uneven flow and some backflow in the mid zone. (2851m3/h mean)
2 - Current measurement to get power drawn from fan.
Details: we use datasheet power factor, based in Q stimation. (231W)
3 - RPMs measurement with RPM "reflection type" meter:
Details: we couldnt use the reflection band that should be used to get good results, but the device still got a measurement that seems to make sense. (1480rpms)

From that we got the suposed working poits from fan curve (see pdf please). We get 3 quite different points, depending on source variable used.

In addition we got the static differential pressure room to outside. It was 55Pa in normal situation (all doors and windows well closed). If a door/window is open that goes to almost zero.

We observed also a peculiar change in flow pattern that occurs when the load changes (from any door/window open to "all closed").
If there is an easy return exit (i.e. a door open) the fan flow is conical, however in the normal point (all closed), the flow reduces and changes to a horizontal, stuck to ceiling pattern.

Our opinion/trust from measured data is like this
1 - Flow: we don´t know if it makes sense the measurement as done (please comment)
2 - Power: this is quite strightforward, and measured with a usual multimeter so we trust this quite a lot.
3 - RPMs: we don´t trust completley our meter here due to we didn´t follow strictly the recommendations, but the result is not very different from the power measurement.

The questions are:
- Flow measurement seems to indicate a higher Q than others and we don´t trust it very much, is it usual? could it be due to measurement set up? (i.e. turbulence at impulsion)

- The big change in flow pattern, can be related to the working point of axial fans?(i.e. "saddle point" of the curve)

- Is it possible to use the room static pressure (55Pa) to help to locate the actual working point of the fan?

Thanks and regards

 

[Drazen]

This is something for HVAC engineer, and the more effort you do in complex issues, the more misguided you might become.

Measuring flow is very complex stuff and requires lot of knowledge to reach reliable results.

In general, your system will never work fine if pressure at all branches is not balanced, and the only feasible way for your situation is to have practically depressurized plenum (you call it conduit).

The best would be to have plenum open on both sides and cross section large enough so that just minor pressure drop occurs.

Central filter should be removed and filter installed at all fan suctions, as filter pressure drop is constantly changing, so no any balance could ever be reached even hypothetically - in theory all fans should have the same curves for that - and you will not have depressurized plenum.

ASHRAE handbooks will give you guidance on air hydraulics, the reason you will not found it elsewhere is that google generally does not offer more complex things at a glance.

there is also manual like Shan K. Wang MCGraw Hill handbook, but I would not recommend you such reading if you are not bent on long-term dedication to HVAC field.

 

[xj25]

Hi Drazen, thanks for the comment. We indeed found that the present system is quite unstable as flow changes in each fan if i.e. you closes a fan (with some cardboard sheet) and then removes it. It is like there are different stable flow configurations.

I have been checking the ASHRAE 111 about measurements and have the following comments:
- our flow measuremets are done with a hood at fan outlet (no good place) with about 2-2.5 times the fan diameter. This is far from ideal, but we get a profile similar than shown in Appendix D, figure D-4,F with 78% of points greater than 10% maximun dynamic pressure (so not so terrible).

The questions are:
1) are our Q measurements rubbish as they are done?
2) someone could explain what happens when an axial fan is forced over the saddle point of its curve? is it possible to detect it by some indirect measurement or inspection? (flow pattern, noise, etc)

Regards

 

[willard3]

>>2) someone could explain what happens when an axial fan is forced over the saddle point of its curve? is it possible to detect it by some indirect measurement or inspection? (flow pattern, noise, etc)


Axial flow fans can operate at two different volumes at the same pressure causing fan surge. I say again, hire an Engineer who is familiar with ventilation systems and don't continue bootstrapping this.