Why does airflow continue to be a problem with residential and commercial installations?

Getting the system to deliver something close to the design airflow (measured in cfm) continues to be a challenge with many residential and commercial installations. Sometimes it's the corner office or a home's second floor-east side that is "starved" for air. A worse case: when the problem is getting the entire system to see enough airflow to stop freezing the coil.

Of course we know it is the boss who occupies that corner office, and we know who uses the master bedroom-bath area in that home.

After 25 years of providing design-build commercial duct systems to installing HVAC contractors, I believe many people misunderstand the nature of duct design/sizing. If you ask most HVAC technicians what is the key word used with duct design, I'll bet most will say velocity. You'll hear them continue with, "You have got to keep the velocity up to deliver the air."

Well, that is partially true; if you don't move the air along at a reasonable rate, say 900 fpm for residential and 1,000 fpm for commercial, then even with insulated duct there is the real possibility and problem of delivering air with a 3° to 5° temperature difference. While just a nuisance with fossil fuel, warm air heat is a real problem with the less than warm air from a heat pump or when trying to cool the space. Chances are good that the callback is because of inadequate airflow. This is where the original designer/installer and the service tech need to be thinking pressure.

I have explained the basic operating principles of a duct system to my apprentices by suggesting we are using a pump called a fan to pump a number of pounds of that other fluid, air, into pipes we call ducts. We put the system under pressure and cut holes into it, and let the air squirt out. But even when they understand it is a "pressure thing" there are still problems.

I think, as creatures of habit, we continue to rely on those longstanding rules of thumb and assumptions we heard from the people with which we apprenticed.

How often have you heard "No problem, I use 1 per 100 feet for commercial?" Or, "I always use no more than .08 per 100 feet for residential on my ductulator?" Even if the cut sheets for the unit are available, and we know how much external static pressure (ESP) is available, how many people will sketch a simple duct plan and not just assume that a 6-inch-hit [flex section] for 100 cfm's should be just fine?


This brings us to the next mistake of believing that external static pressure is a reflection of the entire supply and return duct, plus a filter that is a little dirty. But almost everyone forgets to first deduct the "other stuff" hanging off the system like the return grille, the supply take-off, the flex, and the register or diffuser. How many people have forgotten to deduct the pressure drop of a wet coil when they have added a split to a home furnace? Of all that "stuff," the flex is the biggest question mark.

I remember in the mid-1960s when most of the diffusers were "cut in hard." But by the late '60s, it became clear that flex was the greatest thing since sliced bread. Making that final connection with a flex hit [section] saved a whole lot of time and money, and did a good job. Unfortunately, like a number of good things used to excess, there are some problems.

First, because it was meant to be the final, short, connection from the hard duct to the diffuser with a neck velocity that rarely went above 500 fpm, the higher pressure drop we see with the internal "slinky" was not a concern at these low velocities.

Now that it is commonly being used as a duct with higher velocities and longer length, there are some concerns about not following the flex manufacturer's installation requirements. The pressure drops for flex are based on it being in the relaxed state. If you take the product out of the carton, stretch it and release it on the floor, it is relaxed. The problem comes when your installer has a 20 feet run and, "oh what the heck, he paid for it" and he installs the whole 25 feet in the 20 foot run. It is now in a compressed state with a much higher-pressure drop.

Also, the manufacturers publish the need for hangers on 5 foot centers with no more than a ½ inch sag. We need to hang the product, especially in residential attics, because if we were to allow the insulated flex, or any insulated duct, to rest on (or worse) be covered by loose fill insulation, the surface of the duct's vapor barrier could very easily be below dew point. This is very well understood in Florida, where the conditions are usually present for "sweating" ducts, but often overlooked in other parts of the country like our Philadelphia area.

The idea that a steady source of water could wash out the chemical treatment in cellulose insulation and leave behind organic food, water, and darkness that could encourage green or worse black stuff to appear is a thought that makes one pause. Suffice it to say, remember to deduct the other "stuff" hanging off the system and take a close look at how much flex you use, and that it is properly installed.


Another concept that seems to be overlooked is design leg. The design leg is the most difficult path that the air needs to travel. Many times it is not the longest measured length until you have included the fitting equivalent lengths to your sketch. We need to size the duct so that we can get air from the most difficult supply diffuser to the most difficult return grille. Any other supply "hits" need to be dampered.

This explains why it is sometimes a corner office that is the problem that can't easily be fixed. Once we know how much ESP is left for the duct, we can do the math and pick a friction rate that will allow enough pressure to move the air from supply diffuser to return grille. This is especially important with residential, direct-drive equipment where there is no flexibility once you exceed the high speed ESP value.

We should not expect the manufacturers to add extra fan horsepower as a safety factor for a badly sized job. Commercial units can be ordered with a larger motor, if you know you have a job with limited space and smaller ducts with more friction. Ordering a larger motor should be the exception and not the rule. Your customer will appreciate the choice every time they pay their electric bill.

This design leg concept also brings into question the best practices of using extensive panned joist and stud spaces for duct runs. A sketch of the system may very well indicate that there isn't enough fan ESP to support a fully ducted, supply and return duct system. We need the supply duct, but are those joist and stud spaces really helping us do a good job?

There are a number of articles that say otherwise. Haven't we all seen the comment of John Tooley, in Florida, who said these panned spaces are energy leaks that increase infiltration? Many of his comments about this and the real energy loss due to leakage that can't be overcome by increasing R factors, are worth reviewing. If you are concerned that the "air won't get back to the unit" from the closed door to the master bedroom and bath, then use one stud space for a transfer duct.


In my opinion, one, high, central, ducted return is the most effective design detail especially on multi-story homes. We all know that warm air rises with a stack effect that automatically makes the higher floors a little less than 1° warmer for every foot in height. By grabbing that warm air in the summer and cooling it, then delivering it through the whole house, we can greatly reduce the natural stratification before we have to figure how much more cool air we would need to deliver to the upper floor. In the winter, there is still efficiency in taking warm air, adding heat, and then delivering it to the rest of the house.

Extensive return duct creates problems on commercial systems. This is especially true of rooftops with economizers.

The idea of getting free cooling in the spring and fall is great, if we can get the economizer to work. The problem is when there is an extensive, high resistance return duct hanging off the system, and the outside air can't force its way into the building unless it can first exhaust some of the air inside the building. The standard barometric dampers were supposed to react to the pressure difference and relieve the internal pressure difference created by the rooftop. But barometric dampers don't work if the air has to first travel through an extensive return duct system.

Manufacturers saw the problem, then began to offer rooftops with power exhaust fans to do what the barometric damper wasn't able to do. But a power exhaust fan is not a return fan! In the 1960s, extensive return duct systems would have their own fan, sized to move just the return air, because the old dead guys understood that to get outside air into the building they would need to exhaust the return air. Sometimes we need to rediscover things.

Well, the conclusions we can leave with after this walk down memory lane include: Take a long, hard look at the assumptions we use in our work. Remember that ESP is not available for duct. Get the equipment cut sheet, make a sketch, and do the math for duct design. Keep the return short and high, and hard connections are best for returns.

Mike Bergen is president of Air Handling Services, a company that specializes in design-build commercial duct fabrication since 1978. Bergen has extensive experience with industry training having assisted with NCCER Sheet Metal, NAIMA, ACCA, and ARI educational programs.

Publication date: 06/05/2006