Big story, hope you will be patient! I am trying to give the whole story all at once.

I have been called in as a consultant to help a friendly competitor with a problem he has been struggling with for over a year. My findings have been knocked by the manufacturer, who has a completely different view. No work has been done yet. The unit was installed a little over a year ago.

The complaints; Condensate is not draining from pan. The fan compartment is filling with water on floor at least 1/2 inch deep. The coil is freezing up. Insufficent air supply to halls. Filters being sucked out of holders.

It is a gas fired makeup air unit, with an induced draft blower fan, with an ac coil placed in the inlet opening, and a bypass damper placed above the ac coil.
The coil is coupled with a 70 ton Carrier condensor, 2 compressor circuits, each compressor feeding 2- 18 ton tx valves operating on R22. Each compressor has an unloader valve on the head, unloading 2 cylinders. The first stage has a HG bypass.

The fan is setup, from factory with pulleys sized to deliver 33,620 cfm of air into the halls of a condominium apartment building.
My first thought when I put my hand in front of the discharge air vent on the top floor is that the filters are plugged. Insufficient air is being delivered. I dont have the design specs on exactly what cfm should be delivered, so didnt measure this, just didnt feel right.
Examining the unit, the inlet opening is 5 ft high, and 10 feet across. The ac coil occupies the bottom 3 ft, with the bypass damper across the top 2 feet. The bypass damper is closed, as this is in summer mode, and the previous tech felt it should only bypass in winter mode.
The fan compartment is full of water, the water is not draining from the pan. A quick inspection of the pan reveals 2 hacksaw cuts at each end, in the side of the pan. I would believe that someone cut them, to try and drain the water out of the fan compartment back into the condensate pan, and down the condensate drain.
A quick calc shows me that the air velocity across the coil...33,620cfm/3x10 sq ft of opening=1000+ ft/min. I am assuming that the condensate is being sucked off the coil. I believe that this happens above 500 ft/min velocity, but cannot verify this, or remember where I got this number from.
The filters (cardboard frame pleated 2in thick) are also bent in the middle and sucked from the frame. Supplier advises me that they are also designed for 500 ft/min max.
I measure air flow at a number of points across the evap coil. A close measurement (2inches away) gives a variety of readings, from 200 fpm, to 2000 fpm.
I deduce that the coil is partly plugged. A measurement taken 18 inches from the coil, gives an inconsistent reading varying from 600-1200 fpm. There is a lot of turbulence in the air.

At this point I am deducing that the air speed is too high! The manufacturer advises me that the design is for the bypass dampers to be permanently open, so that 22,000 cfm are to go through the dampers, and 11,000 cfm are to go through the coil.
I still see that this air speed is too high, and that the air filters will not survive this install...my math is 33,620 cfm/5x10=672.4 fpm. I am thinking that they need a bigger inlet opening to reduce the velocity of the inlet air, reducing the static.
The static in the fan compartment is 1.2 inWC. I think that this is preventing the water from draining out of the fan compartment.
I have not been back to site to test the manufacturer's advice to open the bypass dampers fully.

I am thinking that the air will not get a lot of cooling, when 2/3 of it is bypassed around the coil. The plan according to one engineer is normally to bypass 1/3, but this unit is designed to bypass 2/3.

I am having concerns about the coil. It will only see 11,000cfm. 30 ft sq of surface, means 366 fpm, a bit slow. I saw some freezing on the edges of the coil when all 33, 620 fpm (84F 40RH) were going through it, have big doubts about what will happen when we bypass 2/3 of the air.
(Am questioning all my assumptions at this point...so,,,anyone know what happens to the amount of heat transfer when the air is going to fast? Am I seeing poor heat absorption from this fast air, but perhaps I will see better heat transfer when I alter the air flow as recommended, to the 366fpm).
This coil is hooked to two 35 ton compressors, each one with 2 18 ton tx valves. I am viewing this as way too much mechanical cooling equipment for that much air flow. The rule of thumb that I have grown accustomed to is 400 cfm/ton for AC.

The engineers claim that this unit is designed for 90 deg F ambient air entering, and will deliver a 33 deg F TD for a discharge air temp of 57 deg F.

If I use their number of 57 deg off the coil, and mix that 11,000 cfm with 22,000 cfm of 90 deg F air I get (1x57+ 2x90)/3=79 degF. I dont see that as adequate cooling for supplying common areas in an apartment building.

When I saw the unit it was only operating on one compressor (the second stage has plugged LLFD) with entering air at 84 deg F and 40 %RH. The back pressure was sitting at 65 psi. There was frosting on the ends of the coil (on the fins at each end and return bends, also on distributor tubes) I see this as insufficent load on this unit.
The compressor was unloaded, and the HG bypass would have been bypassing but I shut off the isolation valve.

MY math is 11000 cfm/450 cfm/ton=25tons, or 11000/400=27.5.


My conclusions:

Inlet air opening is too small, needs to be increased from 50 sq ft, to 75 sq ft, to allow 33,620 cfm at 450 fpm air.
My suggestions;
-is to cut a side out of the mixed air chamber, and put filters in it with balancing dampers in it, c/w filter frames.
-remove the 2x10 bypass damper and install a evap coil in this opening above the existing coil, to give a full surface area of 5x10 evap. My math is 450 fpmx 50 ft sq=22,500 cfm and 22,500 cfm/400cfm/ton=56.25 tons, or 22,500/450=50 tons. Use the balancing dampers in the newly installed opening to ensure 2/3 air through coil.