Air heater underperformance

[hollerg]

I have a parallel 27"x27" steam heated finned air coils. One underperforms when the flow rate is shifted to that unit. It may be inerts or poor quality (wet) steam. However my duct work may be skewing the flow profile too

Short of CFD modeling, anyone know of a paper on duct flow misdistribution into or a paper or method to compute/test for the skewed flow & impact on the heat transfer? My west unit max's out @175 F but the east unit hits 200 F no problem. (configuration: asymmetrically mounted bullnose tee feeding mitered elbows w/o gore)

thks
Gary

 

[MintJulep]

CFD? Really?

This sounds like a simple case of "one of these things is not like the other."

Go get your hands dirty and find the differences.

Pitot traverses. Smoke. Thermal imaging or IR thermometer.

 

[willard3]

Simple field testing is what you need here like MintJulep said....don't over complicate a simple problem.

 

[walkes]

Air balancing check? do the airflows match?
are the coils clean?

Are you getting proper steam flow?

 

[IRstuff]

Scaling or just blockages are distinct possibilities.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[hollerg]

Gee wiz guys, my first time working with square duct, which has a lot more varation than round stuff.

Have already done the following:

Trap sizing check
New trap
Disassembled unit and inspected air side for damage and fouling
Flushed tube side
Air flow rate confirmed - FCI meter
(isolated one & ran the other)

Already getting ready to do:
thermal imaging of each

Production demands to keep running or I would have:
skipped asking about literature
swapped steam supplies
modify duct and do velocity & temperature traverse

Right now it seems likely that it is air side maldistribution, but of course the "designer" thinks not, and down time needs to be kept to a minimum.

The duct is discharges to the heaters as follows:

bullnose tee 18x18 in 14.5x18 out branches
asymmetrically mounted between two square corner mitered elbows (2 ft & 4ft from centerline of tee). In parallel you find:

Elbow close cpld
--> to eccentric 20 deg expansion flat side down, close cpld to
--> to heater 12" thick, 27"hx27"w close cpld to
--> to outlet ecc reducer flat side down w/ flat end
--> flat end connected to 12h x 24w oval duct 2 ft long
--> 12" oval damper,
--> oval duct, 1 ft to dryer wall.

No gores in any fitting

Obstructions in the area prevented laying it out per ASHRAE,and I thought a paper might cover the issue and provide details to show me specifically how to improve the flow in the ducting I have. Common sense, to me, does not make it common sense to others.

 

[Drazen]

There is reason why collegaues propose that you ned to simplify the issue.

Real aerodynamic picture can be very complex though you possibly look at it as sheet metal pile. That is why we use customary, proven simplificatios. CFD analisys could cost more than your whole system.

Looking at those customary requirements, your ductwork seems to be example of what should not be bad practice.

For instance:

- to avoid "system effect" which can cause really high pressure drop, you should have straight piece of at least 3 x haydraulic diameter at unit fan exhaust
- sharp bends needs deflectors. Bands close to each other can cause high pressure drops. It is recommendable that any fitting is 5 x hydraulic diamter away from the other whenever possible, and for sharp fittings things are much worse.

You can use pitot tube to measure pressure downstream of your fittings to confirm those well proven points, or pay say $25k to try to confirm it with CFD, but in any case you will likely need to rebuild part of your ductwork, where good fitter is faimiliar with most of mentioned good practice.

 

[MintJulep]

Total airflow through each may be the same, but the flow profile over the coil face area could be very different.

You could have a stagnant area, with all the flow going through only part of the coil at very high velocity.

In fact, from your description of the duct I think you almost certainly do. As Drazen notes, the design sounds sub-optimal.

 

[berkshire]

As Drazen says Use a pitot tube to measure pressure downstream of your fittings,
At the same time ,take a combination Pitot static rake and run it across the entire downstream side of the coils to get a velocity profile. This will confirm or deny Mint Julep's comments.
Any way you slice it, it sounds as though you are going to have to rebuild at least part of that duct and add some turn vanes.
B.E.

The good engineer does not need to memorize every formula; he just needs to know where he can find them when he needs them. Old professor

 

[MechEngNCPE]

Sounds like bad inlet conditions like mentioned before.