Hypothetically speakingg . . .

[WAYNE3298]

Read post #3. Makes a lot of sense.

Google upgraded my system this morning and it is nothing but trouble. Sent them a nasty gram hoping to go back the way it ways.

[hvacker]

Just to add, a closed damper will still leak. Sometimes a lot depending on the type. Leakage of 10% isn't uncommon. This can be tested using the temperature method if necessary.

[Juan Madera]

So, if you are looking up at the underside of the roof, you can see the unit's curb? If not and there is ducting above the hard ceiling, the RA duct might have been compromised perhaps. If you can see the curb and all the ducting and know that it is intact, the economizer might need some judicious application of some mastic to the damper blades after you de energise the actuator motor... That is provided that you don't need OA for the server room. Easy relief damper is a bit of rubber belting zip screwed at the top over an appropriately sized hole in the SA on the unit... Still amazed that a unit this large and having an economizer does not have a relief damper in it already. Most of my big RTU's have interlocked linkage that work the OA and relief at the same time.

[billygoat22]

And I was thinking server rooms don't have outside air due to pollen, dust , etc damaging the equipment.

[BBeerme]

As you can see . . . Questionable 'engineering' and install.

We can pull in 100% outside air, but no where for it to go.

Anyone need some extra strut, it's laying all over the place.

Are glue lams rated for outdoor use? Never seen it before.

[BBeerme]

Anytime I've seen a package unit on a server room, there has been outside air. Not so much when it is a smaller room served by a mini split.

And I was thinking server rooms don't have outside air due to pollen, dust , etc damaging the equipment.

[BBeerme]

Economizers have been disabled and linkage adjusted to close dampers as tightly as possible.

So, if you are looking up at the underside of the roof, you can see the unit's curb? If not and there is ducting above the hard ceiling, the RA duct might have been compromised perhaps. If you can see the curb and all the ducting and know that it is intact, the economizer might need some judicious application of some mastic to the damper blades after you de energise the actuator motor... That is provided that you don't need OA for the server room. Easy relief damper is a bit of rubber belting zip screwed at the top over an appropriately sized hole in the SA on the unit... Still amazed that a unit this large and having an economizer does not have a relief damper in it already. Most of my big RTU's have interlocked linkage that work the OA and relief at the same time.

[BBeerme]

With roughly 48,000 CFM on a roughly 2,000 square foot room, I'm pretty sure there will be plenty of fresh air due to leakage.

Just to add, a closed damper will still leak. Sometimes a lot depending on the type. Leakage of 10% isn't uncommon. This can be tested using the temperature method if necessary.

[WAYNE3298]

Some server rooms have outside air and some do not. The ones that do usually have just enough OA to slightly pressure the space. Fresh air requirement isn't much of an issue because most are only occasionally occupied. Economizing is not normally done.
I have never seen a server room supported by one large AHU. If that is what you have there is no redundancy which is normally a major design consideration.

[BBeerme]

For the last of the three photo's, I'm standing on top of the second unit.

120 tons of A/C for 2,000 square feet.

[BennyD]

Makes sense. With a return restriction due to the lack of return grills, the unit was seeing a good negative return pressure. This would make the unit pull even more air through the OA dampers.

I've had a similar issue on some 15 ton units. I used sheet metal and foam tape to make a tight seal against the damper when it was fully closed. Did help.

[hvacker]

Saw the pictures. What a crappy job. Isn't exposed duct required to be insulated in CA?
There is so much outside duct an access should be cut so the duct can be sealed from the inside unless the existing sealant is better than it looks.
I expect ducts to leak 10% or more on older installs. Duct pressure testing can be a real eye opener for those that think sheet metal joints are air tight.

I've seen exposed glue lams for architectural details and I've had a piece outside for 15 years or so and it's surviving so far. That doesn't mean it's approved. It would still seem to need a preservative I would think. Maybe it has one. All the good preservatives have been pulled from the market thanks to CA. Stuff like Penta preservative is missed.


Hopefully you won't need to go to this extreme but with 120 tons on 2000 sq ft you might. On one space I had with oversized equipment that resembled a wind tunnel I disabled one of two compressors, blocked half the evap coil and slowed the fan to match the new capacity. Like said...extreme.

A server room really doesn't want uncontrolled infiltration with all the crap blowing around.
Blocking off the OSA dampers and doing any necessary duct sealing and insulating. Adding a damper for any OSA necessary. You still might need to cut capacity because of short cycling.
Or your sales person might make a call to replace the whole mess with a couple of rational replacement units. Operating costs won't get cheaper.

[BBeerme]

Normal around here to insulate the exterior ducting on the interior of the duct. Didn't really inspect for that, but pretty sure it's insulated, I remember looking up into the return when I was in the room. Think I would've noticed if I was looking at sheet metal.

But yeah, pretty crappy install. Those ducts span something like eight feet. Unsupported. And with metal too thin. You need a real tall standing seam, supports inside the ducting, or external supports. The duct sealant applied to the exterior is so thin it didn't even do its job.

Since the tenant is just taking occupancy, seriously doubt they're going to be re-doing anything in the near future.

For now, they're happy they can open the door much more easily.

Saw the pictures. What a crappy job. Isn't exposed duct required to be insulated in CA?
There is so much outside duct an access should be cut so the duct can be sealed from the inside unless the existing sealant is better than it looks.
I expect ducts to leak 10% or more on older installs. Duct pressure testing can be a real eye opener for those that think sheet metal joints are air tight.

I've seen exposed glue lams for architectural details and I've had a piece outside for 15 years or so and it's surviving so far. That doesn't mean it's approved. It would still seem to need a preservative I would think. Maybe it has one. All the good preservatives have been pulled from the market thanks to CA. Stuff like Penta preservative is missed.


Hopefully you won't need to go to this extreme but with 120 tons on 2000 sq ft you might. On one space I had with oversized equipment that resembled a wind tunnel I disabled one of two compressors, blocked half the evap coil and slowed the fan to match the new capacity. Like said...extreme.

A server room really doesn't want uncontrolled infiltration with all the crap blowing around.
Blocking off the OSA dampers and doing any necessary duct sealing and insulating. Adding a damper for any OSA necessary. You still might need to cut capacity because of short cycling.
Or your sales person might make a call to replace the whole mess with a couple of rational replacement units. Operating costs won't get cheaper.

[hvacker]

The stuff inside is probably just sound deadening material. The black bats glued to the metal. It has some minor insulating qualities but that's not what it's for. It's to quiet air noise.
If I remember it's usually 1" r1 or r2. With that much metal exposed the skin transfer will be high. I used to have some charts giving the loss/gain but their misplaced.
My guess is no permit was pulled.

[BBeerme]

I didn't think I needed to ask. The install speaks for itself.

The stuff inside is probably just sound deadening material. The black bats glued to the metal. It has some minor insulating qualities but that's not what it's for. It's to quiet air noise.
If I remember it's usually 1" r1 or r2. With that much metal exposed the skin transfer will be high. I used to have some charts giving the loss/gain but their misplaced.
My guess is no permit was pulled.

[Redhau5]

If you look at it terms of air changes you could have as much as 1440 ACH of supply air in the space. If you only had 2% leakage that would still be about 29 ACH of outside air. To put that into perspective, operating rooms requires a minimum 20 ACH and 4 ACH of outside air. In fact, your ACH is acutally higher in theory because the servers might take up a significant amount of volume in the space.

The solution: Either re-evaluate the need for (2) 60 ton units or add an exhaust fan.

[BBeerme]

Tightly closing the outside air dampers has greatly improved the situation.

If you look at it terms of air changes you could have as much as 1440 ACH of supply air in the space. If you only had 2% leakage that would still be about 29 ACH of outside air. To put that into perspective, operating rooms requires a minimum 20 ACH and 4 ACH of outside air. In fact, your ACH is acutally higher in theory because the servers might take up a significant amount of volume in the space.

The solution: Either re-evaluate the need for (2) 60 ton units or add an exhaust fan.

[Poodle Head Mikey]

Huh; so you were adding something to the space all along, eh? <g>

I might gasket some sheet metal over the OA intake.

I guess now you'll want to start checking that return duct leak rate. <g>

PHM
--------

Tightly closing the outside air dampers has greatly improved the situation.

[BBeerme]

I think if you will re-read the third paragraph in the very first post on this thread, that is what I said.

To eliminate further high static in the room, they need more return grills for the common attic return.

Since the people there were convinced it was a balance issue, I popped a bunch of ceiling tiles to check for dampers. I didn't unwrap any insulation, just pressed on it trying to feel for the damper handles. But I left all of the ceiling tiles out as I went.

When I was done, I checked for static at the door. Since this is the barometer that everyone was using. And virtually no static. Teeny tiny bit, yes. But the door opened and closed normal.

Remember, in the second paragraph in the very first post on this thread, I mentioned that small people there could not even open the door. That's how bad the problem was.


All that said, yup, next time it rains, the return will likely suck in water. There is little doubt there.

Huh; so you were adding something to the space all along, eh? <g>

I might gasket some sheet metal over the OA intake.

I guess now you'll want to start checking that return duct leak rate. <g>

PHM
--------

[WAYNE3298]

Check out the thread about the negative pressure in the utility room. The off the wall, shoot from the hip advice concerning why the room is too negative that the inexperienced tech has been getting should not happen. Don't most people here know how to troubleshoot? At least the guy asking the question realizes he doesn't have all the answers.