Closet cooling and fan static pressure

[kaspor]

Hello

I have a closet which is approx 2.7m high by 1m deep by 1.2m wide. I have l equipment in there that needs to sit at max 30degC. There are two double doors on the closet.

In order to keep it at 30degC I’ve installed 4 extraction fans on a thermostat. The fans draw air in through gaps in the doors and from underneath and exhaust it at the top. The fans are similar to these

S Series 100 Wall/Ceiling Exhaust Fan – White - Universal Fans

The Fanco S Series exhaust fan is a small yet capable fan. It features a simple design and can be mounted on the ceiling or wall. It can also be connected to duct so the air may be exhausted out of the house.

universalfans.com.au

The problem is the noise. They’re generation 30dba each. I’ve insulated them as much as I can but they’re still loud.

So I then started looking at quieter fans such as these

NF-S12A FLX

The NF-S12A marks the third generation of Noctua’s signature S12 series quiet case fans, which have received more than 250 awards and recommendations from the international press. Featuring a further refined blade design with Anti-Stall Knobs that works in tandem with the new AAO (Advanced...

noctua.at

They have a much lower db but looking at the performance curve it looks to have a much lower static pressure rating

NF-A12x25: performance comparison to NF-F12 and NF-S12A

Designed in Austria, Noctua's premium cooling components are renowned for their superb quietness, exceptional performance and thoroughgoing quality.

noctua.at

Does this mean they will inherently take longer to get the temp in the closet down as they are not able to draw the same amount of air in?

I’m worried that if I change the fans over that the newer fans will be useless as they won’t be able to draw air in like my current ones.

Also cutting holes or vents for intake isn’t an option unfortunately.

 

[3DDave]

The 30dB noise is the noise of the fans, not including the sound that vibration of the wall they are attached to. 30dB is around a whisper; even 40dB is similar to a quiet library.

If the fans seem too loud, perhaps they need better isolation from the wall, perhaps with soft duct work.

 

[Snickster]

Affinity laws applied to fans predict that for a given speed reduction the flow will vary same as speed reduction and the static pressure output will vary in accordance with the square of the speed.

So as I indicated above your existing fan outputs 8 Pa static pressure at 80 m3/hr, if you reduce the speed by 50% the new fan output at same point on curve will be 2Pa at 40 m3/hr which is very low. So you can create a new curve for 50% reduction in speed by selecting points on the original 100% speed curve and use the affinity laws to draw the new curve of pressure versus flow. Here is information on the affinity laws:

Affinity laws - Wikipedia

en.wikipedia.org

One thing though is that you really don't know the heat dissipated by the electrical equipment so you really don't know the true required ventilation rate. When designing a ventilation cooling system such as this you would normally add up all the heat from the equipment then calculate the required flowrate of ventilation air. The flow of ventilation air is determined by the following equation:

Q = 1.1(CFM)(delta T)

Q is heat input (BTU/HR)
CFM is the flowrate to be calculated (CUBIC FEET PER MINUTE)
delta T is the temperature rise of the air between inlet and outlet (Deg F)

The heat dissipated by the electrical equipment might be obtained from manufacturer's literature. For purely electronic components such as servers I would think that all electrical power input would eventually be dissipated as heat just like a light bulb? For items such as motor starters or electrical panels only the loss due to resistance of the wiring is dissipated as heat while the rest of the power is transferred to the user via the current flow.

 

[kaspor]

Affinity laws applied to fans predict that for a given speed reduction the flow will vary same as speed reduction and the static pressure output will vary in accordance with the square of the speed.

So as I indicated above your existing fan outputs 8 Pa static pressure at 80 m3/hr, if you reduce the speed by 50% the new fan output at same point on curve will be 2Pa at 40 m3/hr which is very low. So you can create a new curve for 50% reduction in speed by selecting points on the original 100% speed curve and use the affinity laws to draw the new curve of pressure versus flow. Here is information on the affinity laws:

Affinity laws - Wikipedia

en.wikipedia.org

One thing though is that you really don't know the heat dissipated by the electrical equipment so you really don't know the true required ventilation rate. When designing a ventilation cooling system such as this you would normally add up all the heat from the equipment then calculate the required flowrate of ventilation air. The flow of ventilation air is determined by the following equation:

Q = 1.1(CFM)(delta T)

Q is heat input (BTU/HR)
CFM is the flowrate to be calculated (CUBIC FEET PER MINUTE)
delta T is the temperature rise of the air between inlet and outlet (Deg F)

The heat dissipated by the electrical equipment might be obtained from manufacturer's literature. For purely electronic components such as servers I would think that all electrical power input would eventually be dissipated as heat just like a light bulb? For items such as motor starters or electrical panels only the loss due to resistance of the wiring is dissipated as heat while the rest of the power is transferred to the user via the current flow.

Thanks for the response. I tried the speed controller and that did nothing for noise reduction. I can definitely say it’s the motor hum that drives me mad.

I’ve come to the conclusion that I probably need better quality fans. I had another look tonight and The Noctua ones won’t work as mounting will be an issue.

Does anyone know of any decent quality quiet wall fans that have good static pressure and air flow ratings? Everything I look at is 30dba and above

 

[dvd]

Some good general information on fan acoustics attached.

If you could create an insulated plenum with fans discharging into the plenum with plenum discharging into the room, you could cut noise without significantly increasing pressure drop. The devil is in the details on all of this, so all anyone can do is speculate.

 

[dtimberlake]

"it’s the motor hum......"

120 Hz hum?

I suspect the fans are mounted to the doors, and the doors are acting as loudspeaker diaphragms.
When the are running touch a finger tip to a door.
If vibration can be felt it is capable of generating sound

Very soft isolation mounting with no mechanical short circuiting paths could work but would be tough.
A hinge mounted door is mighty flexible/soft in rotation.

Clamp the doors to the jamb.

Mount the fans to the jam.

 

[HVAC-Novice]

The ambient temperature (room around your closets) is at 28°C? Why not put the hardware near or in an air-conditioned room?
if you need the room to be 30°C, and the air you exchange is 28°C, you need to run a lot of air depending on the cooling load generated.
If you know heat rejected (by measuring average power consumption from the receptacle), you can calculate the airflow required. But again, at only a dT of 2°C, this likely is a lot of flow.

And whatever air you exchange, larger slower fans typically are less noisy for a given airflow. If you look at PC fans, look at 140mm or larger. But start at figuring out what airflow you need, then select a fan setup/ducting etc. You kind of do it backwards, you start selecting fans before you know the airflow requirement.

 

[LittleInch]

The only really quiet fans you're going to find anywhere are similar to those 12V cooling fans you foiund earlier.

One key issue is that your air flow looks bad to me. Basically 90% of your air will simply travel from the gap at the bottom up to the fans and bypass the equipment shoving out heat.

I think what you need is one of the biggest 12V fans at the back of the closet blowing air into each shelf and then the same number extracting the air at the top.

I estimate at the moment you have basically the same square area of inlet vent as you do exhaust fan area. That is bad as it simply is too much resistance to air flow. You need for an extraction fan probably about 5 times the square area of open hole / grill / louver to the fan area to allow air to enter with virtually no differential pressure.

As for options there are many many server cabinet fans sets available like this.

19 Inch 3U Fan | Rack Mounted | Low Noise

19 Inch Rack Mounted Low Noise 3U Fan from Data Comms Direct. Order Online today from the data comms specialists. Next day delivery available.

www.dcdi.co.uk

 

[kaspor]

Some good general information on fan acoustics attached.

If you could create an insulated plenum with fans discharging into the plenum with plenum discharging into the room, you could cut noise without significantly increasing pressure drop. The devil is in the details on all of this, so all anyone can do is speculate.

Thank you I will have a read of this

"it’s the motor hum......"

120 Hz hum?

I suspect the fans are mounted to the doors, and the doors are acting as loudspeaker diaphragms.
When the are running touch a finger tip to a door.
If vibration can be felt it is capable of generating sound

Very soft isolation mounting with no mechanical short circuiting paths could work but would be tough.
A hinge mounted door is mighty flexible/soft in rotation.

Clamp the doors to the jamb.

Mount the fans to the jam.

No - the fans are mounted to the drywall above the doors. It is not vibration. The fans are screwed into the wall and I have put pressure on the fans with my hands. No change in sound, the hum is from the motors (I'm sure of it!)

The ambient temperature (room around your closets) is at 28°C? Why not put the hardware near or in an air-conditioned room?
if you need the room to be 30°C, and the air you exchange is 28°C, you need to run a lot of air depending on the cooling load generated.
If you know heat rejected (by measuring average power consumption from the receptacle), you can calculate the airflow required. But again, at only a dT of 2°C, this likely is a lot of flow.

And whatever air you exchange, larger slower fans typically are less noisy for a given airflow. If you look at PC fans, look at 140mm or larger. But start at figuring out what airflow you need, then select a fan setup/ducting etc. You kind of do it backwards, you start selecting fans before you know the airflow requirement.

Yep poor design, but I cannot re-wire the house unfortunately.

The only really quiet fans you're going to find anywhere are similar to those 12V cooling fans you foiund earlier.

One key issue is that your air flow looks bad to me. Basically 90% of your air will simply travel from the gap at the bottom up to the fans and bypass the equipment shoving out heat.

I think what you need is one of the biggest 12V fans at the back of the closet blowing air into each shelf and then the same number extracting the air at the top.

I estimate at the moment you have basically the same square area of inlet vent as you do exhaust fan area. That is bad as it simply is too much resistance to air flow. You need for an extraction fan probably about 5 times the square area of open hole / grill / louver to the fan area to allow air to enter with virtually no differential pressure.

As for options there are many many server cabinet fans sets available like this.

19 Inch 3U Fan | Rack Mounted | Low Noise

19 Inch Rack Mounted Low Noise 3U Fan from Data Comms Direct. Order Online today from the data comms specialists. Next day delivery available.

www.dcdi.co.uk

Unfortunately the back of the closet is the a mirror that is mounted in another room. No cutting holes there :\. The fans at the moment keep the equipment cool, I have 4 temperature sensors in there monitoring temperatures. Thanks for the recommendation I will take a look.

 

[Snickster]

I have an idea that is a little out of the box but worth a thought.

The fans actually blow inward if you have the square face mounted on the outside of the closet towards the room. The exhaust is through the 100 mm duct connection on the back of the assembly. Therefore in the present installation you are actually sucking in from the room and discharging through the 1" undercut.

If you would reverse the installation and put the square mounting face on the inside of the closet gyp board wall and put the discharge connection facing outward you would then be sucking through the 1" undercut. Then you could install a hood vent on the outside wall at each fan outlet like this one below that is available from Universal Fan in Australia for $25 ea. that distributes your existing fan Fanco. I believe putting the fans where the motors are on the inside of the closet will deaden the fan noise some and installing an external hood will further muffle the sound. May be worth a try for $100 total plus shipping cost.

External Vents / Grilles | Universal Fans Australia

External Vents / Grilles

universalfans.com.au

 

[kaspor]

I have an idea that is a little out of the box but worth a thought.

The fans actually blow inward if you have the square face mounted on the outside of the closet towards the room. The exhaust is through the 100 mm duct connection on the back of the assembly. Therefore in the present installation you are actually sucking in from the room and discharging through the 1" undercut.

If you would reverse the installation and put the square mounting face on the inside of the closet gyp board wall and put the discharge connection facing outward you would then be sucking through the 1" undercut. Then you could install a hood vent on the outside wall at each fan outlet like this one below that is available from Universal Fan in Australia for $25 ea. that distributes your existing fan Fanco. I believe putting the fans where the motors are on the inside of the closet will deaden the fan noise some and installing an external hood will further muffle the sound. May be worth a try for $100 total plus shipping cost.

External Vents / Grilles | Universal Fans Australia

External Vents / Grilles

universalfans.com.au

I like your thinking. However, this is how the install already is! The fans are mounted inside the closet. They extract air from inside (via the 1" gap) and blow outside towards the open room. If I put fan covers on the outside of the closet that are visible from the room, I'll be sleeping on the street . At least I'll still hear the fans out there right? lol

I have found better quality fans here and plan to try them:

Vent-Axia PureAir Sense | Vent-Axia

www.vent-axia.com

 

[kaspor]

Ok so I tried those more expensive fans, what total rubbish. Louder than the ones I have. Real shame, Vent-axia are supposed to be the world leaders in fan manufacturing (from what I've heard).

I was really hoping that would fix the issue. Never the less - back to the drawing board.

I have taken some videos and photos of the install. Links:

Pictures taken from inside the closet - you can see the 4 fans up the top:
https://drive.google.com/file/d/1M1uEPBvW2q6JhQI7yVebeULCq0TbsX4i/view?usp=sharing

p3.jpg

drive.google.com

Picture taken from outside the cabinet - the fans are venting through the grills up top:
https://drive.google.com/file/d/1twK47vOSBpmLbF-7k1hmrvv0apPipjPx/view?usp=sharing

This is a video from outside, I turn the fans on, you can hear the noise:
https://drive.google.com/file/d/1Sfgjz3zCYj2TtnaxVbVRPhxAMfEP4Hmb/view?usp=sharing

This is a video from inside, I turn the fans on, you can hear the noise:
https://drive.google.com/file/d/1ZIWkifW1IfIV4s5xBbtiCScOb0KlDS0O/view?usp=sharing

This is the plans showing where the closet is in relation to other rooms:

plans.png

drive.google.com

Note: there's no access to the roof space and venting into the roof cavity is illegal here.

It may not seem like much in the videos, but the sound travels throughout the house as it's low frequency

Crappy place for this setup I know, but what's done is done.

I used to be a builder/cabinet maker, so happy to entertain building or rigging up whatever

A thought that did cross my mind was building an insulated box at the top of the wardrobe. I could then mount the fans to the underside of the box so that they would blow air through the box and out through the front grills. But not sure if that would reduce the sound all that much. It's alot of work for something that may not work :\

Any thoughts?

 

[LittleInch]

OK, it's taken a while to understand your set up and constraints, but how about this.

I see a few issues here:

1) I think the majority of your air flow is under the door, up the inside of the doors and out the fans at the top. Hence your air flow inside the closet is poor, hence you need a load of air to actually keep the units cool. I reckon you're probably extracting about 80 to 100 m3/hr with 4 fans. Thats an ach of about 30 or once every 2 minutes. That's a lot of air and might be more. It tells me you either have a load of equipment there or your airflow distribution is poor.

Alternative - Cut some holes or a section out of the back of the shelves( say 1 to 2" deep and maybe 75% of the width of the shelf), which is usually where the equipment fans are. Make a shelf which virtually touches the doors about 2" off the floor and extends about 75% to the back of the closet. This will guide air to the back of the closet, up the back of the closet picking up all the heat and then out the top. You might well find that only running one or two fans will work.

2) Your inlet air flow is too small and is limiting the air flow.

I can now see why the boss doesn't want more holes cut into the doors, so an alternative is to use your low power really quiet fans, but have two in extract mode and two in input mode with some 4" or 6" ducting going from the top all the way to the base of the closet. Plus cut the holes in the back of the shelves. This way the DP of the fans is effectively doubled as the air inside the closet is now under a very small positive pressure ( but might leak out the 1" gap), but at worst is atmospheric pressure, not negative pressure having to "suck" in the air through your very narrow gap.

Also do your louvers have a mesh inside? Get rid of that, especially for the lower power fans.

This post isn't exactly HVAC engineering but is quite fun so please keep us in the loop over progress and the results.

 

[kaspor]

OK, it's taken a while to understand your set up and constraints, but how about this.

I see a few issues here:

1) I think the majority of your air flow is under the door, up the inside of the doors and out the fans at the top. Hence your air flow inside the closet is poor, hence you need a load of air to actually keep the units cool. I reckon you're probably extracting about 80 to 100 m3/hr with 4 fans. Thats an ach of about 30 or once every 2 minutes. That's a lot of air and might be more. It tells me you either have a load of equipment there or your airflow distribution is poor.

Alternative - Cut some holes or a section out of the back of the shelves( say 1 to 2" deep and maybe 75% of the width of the shelf), which is usually where the equipment fans are. Make a shelf which virtually touches the doors about 2" off the floor and extends about 75% to the back of the closet. This will guide air to the back of the closet, up the back of the closet picking up all the heat and then out the top. You might well find that only running one or two fans will work.

2) Your inlet air flow is too small and is limiting the air flow.

I can now see why the boss doesn't want more holes cut into the doors, so an alternative is to use your low power really quiet fans, but have two in extract mode and two in input mode with some 4" or 6" ducting going from the top all the way to the base of the closet. Plus cut the holes in the back of the shelves. This way the DP of the fans is effectively doubled as the air inside the closet is now under a very small positive pressure ( but might leak out the 1" gap), but at worst is atmospheric pressure, not negative pressure having to "suck" in the air through your very narrow gap.

Also do your louvers have a mesh inside? Get rid of that, especially for the lower power fans.

This post isn't exactly HVAC engineering but is quite fun so please keep us in the loop over progress and the results.

Yeah apologies I should've posted the photos and video earlier, just been very busy this past week.

You are correct, the inlet air flow availability is poor. I don't have a load of equipment in there, I have the following:

1. PC that runs barely above idle
2. Dahua NVR (cameras)
3. Modem, router and a 8 port simple network switches (not fan forced)
4. A small amplifier for playing sound
5. Coax modulator

The main heat sources in the list are definitely the PC and NVR, maybe the network switches, but the rest is negligible really.

I should put this into perspective, we're bang in the middle of summer here, so ambient rarely drops below 25degC inside. In Autumn/Spring/Winter, the fans rarely run. Passive cooling is enough.

Thanks for the shelf option, but again that is a nuclear one lol. I will keep it in my options list.

I don't think these are super quiet fans, but that's perception I guess. I cannot seem to find anything quieter unless it's a PC style fan (like the Noctua's), but then I think static pressure ratings would be too low to draw air in with my setup. Reversing two of the fans won't work I think. The fans are blowing into the wall cavity then out through the grills. If I run two in reverse, I'll likely suck the hot air in that I am expelling with the other two?

No louvres or mesh inside those fans.

Yeah probably beyond HVAC related, but I had no idea where else to post.

 

[LittleInch]

You would need to sketch a cross section of your wall for me to understand fully, but should be easy to add a short length of ducting to bridge the cavity in either direction, either blowing in or blowing out of the closet.

Having them side by side isn't great, but the velocities are quite small so I don't think there will be a lot of short circuiting of air flow.

Not sure why cutting a gap at the back of a shelf with computer gear on it is a nuclear option - At some point if you want things to change you're going to need to change things....

 

[3DDave]

I would flip two of the fans around and put a short duct to direct the air down so it's not immediately taken back out; perhaps use plastic duct to carry the air all the way to the bottom. This eliminates the restriction of the door and places the fans in series for pressure losses.

In other words - two fans pull air in and two fans exhaust air out.

For extra you can branch off the inlet to the inlets to the equipment so they are all getting outside air directly rather than taking in hot air from lower locations.

 

[kaspor]

Thanks for the suggestion, I will either try this or cutting two holes in the floor and venting cool air from the floor below (probably a neater solution tbh). I've also ordered the Noctua fans suggested by Snickster - so let's see if these are quieter!

 

[3DDave]

Do not cut holes in the floor. The floor functions as a fire break.

 

[dvd]

There are probably a few holes in the floor for air registers, if the supply duct brings air from below. Supplying cooled air to the closet seems like a reasonable approach. Hopefully the cool air doesn't bypass directly out the gap at the bottom of the closet door.

 

[Snickster]

1. PC that runs barely above idle
2. Dahua NVR (cameras)
3. Modem, router and a 8 port simple network switches (not fan forced)
4. A small amplifier for playing sound
5. Coax modulator

If you can provide the power rating for each of these items I can calculate the required air flow. I believe your air flow is way too high.

A server for an office building produces 150 Watt heat dissipation or about 500 BTU/hr heat dissipation. I know because on a project we had a bank of 3 we had to provide cooling for and I have the records of the heat loads.

The heat removal capabilities of your 4-fans at 320 m3/hr (188 CFM) total as I indicated I calculated in the previous post can handle a total heat load at 6 C (10.8 F) temperature rise as follows:

Q = 1.1 (CFM) (delta T) = 1.1 (188) (10.8) = 2234 BTU/hr

This is equivalent to heat dissipated by over 4 servers which is way more than what you have in your closet. I am thinking at most you have equivalent of 1 to 2 servers at the most, so you should only need 1 to 2 fans of same size you have now or you could put in very low flow fan/fans that have very low noise levels. As LI said your airflow is not very efficient as the flow is not over any of the equipment but through a space that runs in front of the equipment so you might need to arrange the air flow more efficiently so that it sweeps the space around the equipment,

 

[kaspor]

Do not cut holes in the floor. The floor functions as a fire break.

Not where I live, I built the house, it's not a fire break. Fire breaks are only between different dwellings.

There are probably a few holes in the floor for air registers, if the supply duct brings air from below. Supplying cooled air to the closet seems like a reasonable approach. Hopefully the cool air doesn't bypass directly out the gap at the bottom of the closet door.

That is a good point, however supplying cooled air from my HVAC unit would cost an absolute fortune as there would be no zoning setup. The house is quite large, I have two separate HVAC systems and running the downstairs system costs ~80c per hour.

1. PC that runs barely above idle
2. Dahua NVR (cameras)
3. Modem, router and a 8 port simple network switches (not fan forced)
4. A small amplifier for playing sound
5. Coax modulator

If you can provide the power rating for each of these items I can calculate the required air flow. I believe your air flow is way too high.

A server for an office building produces 150 Watt heat dissipation or about 500 BTU/hr heat dissipation. I know because on a project we had a bank of 3 we had to provide cooling for and I have the records of the heat loads.

The heat removal capabilities of your 4-fans at 320 m3/hr (188 CFM) total as I indicated I calculated in the previous post can handle a total heat load at 6 C (10.8 F) temperature rise as follows:

Q = 1.1 (CFM) (delta T) = 1.1 (188) (10.8) = 2234 BTU/hr

This is equivalent to heat dissipated by over 4 servers which is way more than what you have in your closet. I am thinking at most you have equivalent of 1 to 2 servers at the most, so you should only need 1 to 2 fans of same size you have now or you could put in very low flow fan/fans that have very low noise levels. As LI said your airflow is not very efficient as the flow is not over any of the equipment but through a space that runs in front of the equipment so you might need to arrange the air flow more efficiently so that it sweeps the space around the equipment,

Thanks for the reply. Power ratings would probably an inaccurate way to do a heat load calc as much of this equipment is sitting well below its full capacity. I can measure the power draw, but the whole closet is probably drawing no more than 1.2kW (electrical)