System Effect–Dealing with the Point Where the Fan Meets the Duct

I am going to take a break from my string on building a psych chart to do a post on System Effect in support of one of the classes Ryan Stroupe and I teach at the Pacific Energy Center.   If you are following the psych chart string, don’t despair;  the next post in the series is almost done.  I just needed to get this information out and using the blog as a mechanism to do that has a benefit in that it makes the information available to more people.   And judging from the number of view statistics for the posts I have done on turning vanes in elbows and related pressure drops, I think what follows will be of interest.

This is a fairly long post, so the links below will jump you to various topics of interest if you don’t want to have to read the entire thing to find something.

The Question

The driver behind this is a question that we asked our students in a quiz.  Specifically, we gave them the following image …


… and then asked them:

How does the fan and ductwork configuration shown above impact the system performance?

Identify two issues that exist with the existing configuration. Describe or draw at least two strategies for improving this configuration in the space provided. The air needs to be flowing to the left for your redesign.

(Back to Contents)

Existing Buildings vs. New Construction

This particular question is one from a quiz that we give in a year long, hands-on existing building commissioning class that Ryan and I teach at the Pacific Energy Center. (I am pleased to report that we are currently in our 10th year, meaning there seems to be some interest in the topic out there).

One of the points that I think is important to make beyond simply answering the question is that what you can do in an existing building commissioning situation is different from what you can do if you catch this problem during the design phase.  That, in turn, goes to a point I try to make about how the field lessons can come back  to inform the design process via design review. 

The other point is that once the system is installed, it can be quite challenging to figure out a way to make a modification to the system that can be paid for by the energy savings associated with it.    That’s not because there is not much energy to be saved;  there is.   Rather, it’s because the cost to make the necessary modifications is typically quite high.  You have already invested in sheet metal and equipment and are faced with changing them vs. optimizing how you used them in the first place.

(Back to Contents)

An Existing Building Example

For example, at a recent workshop Ryan and I did in an existing building, we encountered a system that had a poor fan discharge condition and a poor duct connection to the discharge plenum on top of that.  The poor duct connection to the discharge plenum represented a measured loss of 0.74 in.w.c. at design flow.  The system effect associated with the fan blasting into the plenum with no duct was estimated to be equivalent to another 0.25 to 0.50 in.w.c.  from the AMCA tables in AMCA 201 (ASHRAE has similar tables in the Fundamentals Handbook). This SketchUp model depicts the situation.


At design flow, the poor plenum connection and transition to the round duct would have used up over 3 bhp.  But, since the system was a Variable Air Volume system (VAV) and most of the time, operated at 50% of design flow or less, the loss associated with the condition, most of the time was around 0.22 in.w.c. (also measured).  

And since the system operated on a schedule, the total savings to be achieved was in the range of $200 – $300 per year.   This number would go up if:

  • The system operating hours were higher (24/7) and/or
  • The system spent more time at or near design flow or
  • The system was a constant volume system operating at the design condition. 

For instance, if this system ran 24/7 at the design flow rate, the savings potential associated with the fitting improvement would be in the range of $3,000 – $3,700 per year at electric rates commonly encountered in CA.

In contrast, the cost to try to mitigate the poor fitting using a design similar to what is shown below was in the $8,000 – $10,000 range. 


Doing something to improve the fan orientation and interface with the duct system (i.e. reduce the system effect) could easily cost 2-3 times that.

A lot of the cost associated with making the improvement was the cost to set up and clean up after doing the work; i.e. mobilization and demobilization costs.  The fitting itself represented about $1,500 – $2,000 of material and labor. 

Trying to do anything to improve the fan discharge condition (re-orient the fan, eliminate the discharge plenum and replace it with straight duct that then doubled back to the existing duct connection point, etc.) in addition to improving the fitting was a practical impossibility due to existing conditions and the cost associated with buying a totally new fan module or trying to reconfigure the fan in the field, which is no small trick.   

The bottom line is that for most Owners, spending $8,000 – $10,000 to save $3000 or so per year is a marginal investment.  Spending that much to save $200 – $300 per year is not even going to be considered unless they were a very long term owner with a lot of resources.   Most Owners could save a lot more energy a lot faster by spending that kind of money on a different project. 

(Back to Contents)

The Lost Opportunity

The really sad part of this situation in particular (and this sort of thing in general) is that if the system had been installed with an eye towards a good fan discharge condition from the start (which would look something like what is illustrated below), then:

  • The project would have saved first cost because you would not put in the discharge plenum module in the first place, which is worth about $4,000 – $5,000 in materials, and
  • The fan module cost itself would be virtually identical, as would the cost of the duct connection. 


What that means is that since the discharge condition is near ideal, the operating cost penalty that is now imposed for the life of the system would simply not exist. The bottom line is that getting things right from the start can save first cost (i.e. it’s a value engineering item that is also a value added item)in addition to saving operating cost over the life of the system.

So, that is a recent field example that puts some numbers to the concepts behind the quiz question.  

(Back to Contents)

The Quiz Question in the Context of an Existing Building

For the system in the quiz, if you were to encounter it in an existing mechanical room, it might look like something like this when you walked up to it.


At this point you would not be sure there was a problem since you can’t see the fan (yet). But the clues that there might be one are:

  • The “U turn” off what appears to be the discharge end of the AHU, and
  • The fact that the duct gets bigger after the turn, which implies that the air is being slowed down to typical duct velocities at that point (1,500 – 2,500 fpm). 

The transition that slows the air down is a little hard to see in the perspective above, but here is a view from overhead that illustrates it.  Note how the duct gets wider after the 2nd elbow.


Some of you might be thinking what if the connection is to a plenum;  then there would be no system effect due to the fittings since they are not directly connected to the fan.

That observation is valid, but, if the “U-turn” off of the end of the unit were a well-designed plenum connection, the transition would actually go the other way (big to small); i.e. you would try to exit the plenum at 500 – 800 fpm to minimize the entry loss and then narrow the duct down.  So there is another clue about a potential issue. 

(Back to Contents)

Digging a Little Deeper

If you were to talk up and look through the window …

Looking in Window

… you would see that there is a fan connected directly to the elbows; i.e. the problem in the quiz illustration. 

I will take the entire side panel off the fan section in this next image so you can clearly see the problem that was illustrated in the cross section shown on the quiz and also have the bigger perspective on the unit relative to the surroundings, which will be important moving forward.


Here is a close up of the fan section, pretty much the view in the quiz illustration.


(Back to Contents)

The Problem (and Potential Opportunity) (Or the Lost Opportunity)

From a technical standpoint, there are two problems associated with the installation.  One is that there are fittings located immediately on the discharge of the fan. I should note that the elbow on the fan discharge happens to be in the worst possible configuration in terms of imposing a system effect penalty.  But any fitting that close to the discharge of the fan would be a significant problem. The factors that drive how big the problem will be include the fitting design, the outlet velocity of the fan, and the blast to outlet area ratio of the fan.

That is the point of this slide, which comes from the annual class we do at the PEC on Fans, Ducts and Air Handling Systems.


I mentioned the blast to outlet area ratio of the fan.   If you look closely at a fan discharge, you will discover that there is often a piece of metal near the wheel that makes the area available for air to leave the wheel smaller than the actual outlet area of the fan at the point of connection.   The piece of metal is called the “cut-off” and the area at the plane of the cut-off is called the “blast area”.  The area at the plane of the discharge connection of the fan is called the “outlet area”.  This picture of a fan discharge illustrates those terms for your reference.


Returning to our discussion of the issues associated with the fan application, the second problem is that there is a second fitting immediately down-stream of the first fitting. Closely spaced fittings have significantly higher losses than you would anticipate from the loss coefficients due to the impact of the distorted flow profile from the upstream fitting.

AMCA documents this in Publication 200 and the slide that follows illustrates the effect based on the AMCA data for two elbows in series. 


There also is a third issue that feeds into both of the other issues.  Specifically, the duct size off the fan and through the elbows is at the fan outlet size.  Typically, that is not the size you would use for a duct run at a reasonable friction rate because the velocity off of the fan discharge is very high.  But, in the example above, the turn was made before slowing down the air to the desired duct velocity. That makes both the system effect and the losses through the elbows even worse than they would otherwise be.

(Back to Contents)

Developing an Existing Building Solution

A Theoretical Perspective

The theoretically correct way to solve the fan discharge problem is to provide enough duct after the fan and before a turn (or other significant fitting) to allow the non-uniform velocity profile off of the fan to re-establish itself as a uniform flow profile.  This slide, while focused on the definition of fan static pressure, illustrates what I mean.


The “generic” answer to the question How much duct does it take to re-establish the uniform flow profile is 5 to 10 equivalent duct diameters of straight duct before any changes in direction.  If you expand the duct using an angle of 15° or less, then that is O.K.  but you want to go 5-10 equivalent duct diameters before you do anything else, like a turn for instance. 

There is technical definition of an equivalent duct diameter, but in general terms, it is around duct with about the same cross-sectional area.  In our case, that is about 43”, so 10 equivalent duct lengths gets pretty long.


As you can see, you would be through the wall of the equipment room and then have to double back again.  Here is what the entire run would look like, just to give you that perspective.


Even getting 5 equivalent duct diameters can be tricky in most mechanical rooms.


(Back to Contents)

A Real World Perspective

The bottom line is that I most situations, you would be constrained by the physical space available, either due to walls and other structure or other equipment, pipes, sprinklers, conduits, etc.  So you would do what you could do in terms of a retrofit, (assuming you could afford it).  In this particular situation, that might look like this.


And truth be told, the long duct runs I illustrated above are the extreme associated with really high velocity situations.  I used them in an effort to make a visual point about the fact that it really does take some space to achieve theoretical perfection;  space that we usually don’t have in most of our mechanical rooms.

The good news is that at the velocities we encounter in typical HVAC systems 2 to 3 equivalent duct diameters (which is about 8 to 12 feet of straight duct for our example) will work wonders and capture the bulk of the savings if not all of it.  And, if we are creative about it, we often can find that much space, especially if we are starting with a clean sheet of paper.

In the equipment room in our example, we have that much space available.  But, if you modify the system to take advantage of it in terms of improving the fan discharge condition, your retrofit budget should include a line item for a couple of hard hats to hang outside the equipment room door because you have just created a “head banger”.  In other words, the duct now extends across the entry path to the mechanical room and at a level lower than your head.  This picture,which I took looking straight ahead (my eye level is about an inch below the bottom of the duct, meaning my forehead is above the bottom of the duct), says it all.

Bang Forehead Here

(Back to Contents)

Implementation Cost Considerations

In terms of what is new duct and what could be reused, you might be able to save some retrofit cost by relocating the transition piece from the existing installation to be directly on the fan discharge (green highlight below)  But the old elbows are too small so you would need to make new ones (blue highlight) in addition to making some new duct sections (yellow highlight).


(Back to Contents)

Looking Beyond System Effect

Up to this point, all we have done is address the system effect;  i.e. the penalty associated with having a fitting located on the discharge of the fan.  We have not addressed the closely connected elbows. Compared to the system effect, that issue may be insignificant, especially since in their new location, the velocity through the elbows is lower (they are now after the transition to the larger duct size vs. ahead of it) and they do not see and exacerbate the distorted flow profile that is present right on the discharge of the fan. 

But since you are making new elbows anyway, you might want to make an elbow with less loss than a mitered elbow.  Typically a swept, smooth radius elbow will have a lower loss than a mitered elbow with turning vanes.  If you want an example of this, visit my blog and read the two posts turning vanes in elbows.  In any case, if you used the smooth radius elbows instead of mitered elbows, then you modification would look like this.


As was the case for getting theoretical perfection with regard to the discharge duct length, getting theoretical perfection in terms of the distance between the elbows is probably not practical, especially if you want to stay friends with the folks upstairs.


That said, if the space above the existing unit (but below the floor above) was clear (or maybe only had a light or two or small conduits or a sprinkler that were in the way), it may be practical to elevate the existing duct to the extent possible to minimize the elbow interaction, especially if the losses where significant and/or the unit ran a lot of hours per year at relatively high flow rates.


If you have an experienced and creative tinner doing the work for you, this may not cost as much as you would expect since you are just moving existing duct (yellow highlight below) rather than installing new duct.   


If there literally was nothing above the unit (or if anything there could easily be relocated), you could literally accomplish what I am proposing by jacking the duct up via shortening the hangers a little at a time (after breaking the connections between the horizontal and vertical sections of course).

Meaning you would elevate the entire horizontal section gradually as one piece on the existing hangers vs. removing and replacing it.  And since the duct in this example is all the same size and the reduction in length required at the end where it turns up and goes into the chase is exactly the increase in length you need at the discharge end of the fan, you may even be able to reuse the duct you took out at the chase to fill in at the fan discharge (green highlight above).

(Back to Contents)

There Are Options

Having said all of that, there may be other ways to capture the savings, especially if you think outside the box a bit.  One of the most creative solutions to this problem is one that Steve James (a Marriott Director of Engineering) came up with, which is illustrated in the context of our example in the image below.


Steve’s creativity is the example I used in the PEC RCx101 class when I touch on this topic.  Steve, had a roof top air handling unit that was configured with two elbows right on the fan discharge, just like our example and just like the quiz question.  The system served their spa and underperformed round the clock.  As Steve put it:

We struggled to maintain a uniform level of dissatisfaction in the treatment rooms

Not exactly a ringing endorsement for the spa.  In fact, Steve made the change to improve guest satisfaction in the spa, which, of course, is what they were selling.  The fact that it saved energy was just icing on the cake.

Steve made the decision about what to do simply based on logic;  i.e. he did not know about AMCA system effect factors, etc.  But his engineering background and the logic associated with it led him to the conclusion that the two elbows on the discharge of the fan had to be a bad thing and a restriction to flow. 

So, after thinking about it a bit, Steve had a contractor:

  • Cut out the top of the existing fan housing,
  • Build a new fan housing on top of the existing fan housing
  • Remove the fan and re-install it in the new fan housing facing the other way.
  • Re-connect the fan discharge to the supply duct, eliminating any offsets and system effect,

Here is a closer view of how that would look for our example system with a perspective from slightly above the AHU.


The air still makes the 90° turn, but the trick is that it does it at 500 fpm instead of 2,500 fpm plus.  As a result, the losses associated with making the turn are inconsequential.  And, because the discharge duct leaving the fan is straight, the system effect has been eliminated. 

Pretty cool.

(Back to Contents)

Carrying Field Lessons Back to the Drawing Board

If you have analyzed and tried to improve existing fan systems with poor discharge conditions a few times, you will likely conclude that the best solution is to avoid this problem in the first place, at least that is how I feel about it.  That means delivering better designs.  I have come to think that one of the most powerful things folks working with existing buildings can do is take the lessons learned from the field back to the drawing board/design process. 

As a designer, I was blessed with that opportunity because the firm I worked for sent you out into the field to help start up your stuff and make it work.  That was both fun and sobering.  And it made me a better designer because I learned (sometimes via the school of hard knocks) what worked and what didn’t work.  And, because I was curios and had good mentors, I also usually learned why things worked or didn’t work, which is a very important part of making sure you don’t make the same mistake again.

From what I can tell, in the world we live in today, many designers are not provided with the opportunity to work with their designs after they have been installed.  But folks working with existing buildings and doing new construction commissioning are blessed with that opportunity every day.  So if we want to help make the world a better place, we need to take what we have learned back to the drawing board.  But I personally think that we need to do that in a detached, non-judgmental manner. 

(Back to Contents)

Keeping the Focus on the Problem, Not the Person

I don’t want to divert our discussion too much, but I think this is an important point in the context of doing design review, which is where I am headed.  When I say in a detached, non-judgmental manner, what I mean is that in my experience so far, you will get farther along in solving a problem if you simply focus on the problem vs. the person who you feel is behind the problem.  Meaning that saying:

I think your design is flawed due to the thoughtless way you have oriented the supply fan in your system.

will likely be heard and reacted to differently from the response you might get if you say something like

I had the opportunity last month to do some testing on an AHU with a fan that had a very similar orientation to the one on this project and was astounded at the 0.75 in.w.c.  pressure drop I measured  from the discharge plenum into the supply duct.  I realized the equipment room on this project is very tight, and you may have already thought about this a lot and there are no other options.  But after my experience last month and given the LEED goals of this project, I just wanted to be sure there was not something we could do to improve the fan discharge condition.

As field folks, we have a pretty powerful message we can deliver.  Instead of saying “I read in a book a while back that a square edged orifice has a high pressure drop”, we can say “I ran into a system with a square edged orifice where the supply duct connected to the plenum and measured a pressure loss of 0.75 in.w.c. across it and the fitting after it at design flow”. 

What that means is that the folks out there working with operating systems on a day to day basis are in a position to inform the design process based on our field experience and the hindsight we have gained from it. At the same time, by interacting with the design team, we can develop a deeper understanding of the design intent for our current project and the engineering principles behind it.  That will give us a bit of foresight for the current effort as sell as future efforts. 

So, to my way of thinking at least, everyone wins. But I think you will be more successful at it by staying positive vs. being negative.

(Back to Contents)

Considering Our Example from a New Construction Perspective

Rotating the Air Handling Unit 180°

In the case of the example we have been using, if you go back to the original “as found” existing building arrangement but consider it as an “as found” condition that you encounter during a design review process (i.e., it only exists on paper), one of the things that seems to jump out to you (or at least to me) is that the air handling unit is pointed in the wrong direction.  Here is our original condition again to refresh your memory.


What I am saying is that if you rotated the unit 180°,  and purchased it with a fan that had an up-blast discharge instead of a horizontal discharge, you could eliminate both the system effect and the issue associated with the two fittings in close proximity to each other.  That may not be too hard to do on paper, especially if it is early in the design process.  Here is what that would look like.


In terms of the amount of duct you need to make, the supply duct is shortened and the outdoor air duct got longer.  This system includes an air side economizer cycle, meaning that at 100% outdoor air, the outdoor air duct needs to handle the total supply flow.  As a result, the supply and outdoor air ducts are both sized for the same flow and will be approximately the same size.  So there is likely no significant cost addition (or reduction) associated with reorienting the unit from that standpoint. 

From the supply fan static standpoint, a similar thing happens; i.e. the supply fan has to provide the energy to move the outdoor air into the system.  So, it needs to provide the energy to move the air through the longer outdoor air duct.  But that is offset by the reduction in energy required to move air through the supply duct because it got shorter.  And, the new orientation compounds the benefit because things like the closely spaced elbows at the fan discharge and system effect have been eliminated. 

That said, you probably have observed that the outdoor air duct now has two pairs of closely connected elbows (one pair at the bottom of the shaft and one pair at the connection to the mixing box).  But, neither pair is in a high velocity fan discharge.  And their impact could be further mitigated by perhaps making the outdoor air duct a bit larger or improving other elements of the design. 

For example, if you compare the connection to the outdoor air plenum on the back of the unit in the original design to the revised design, you will notice that duct expands between the elbow and the plenum in the new arrangement.  This slows the air down before it exits the duct and enters the plenum, reducing the loss associated with that transition.  For that to work, the expansion angle needs to be gradual; 15° or less is a good rule of thumb.  Otherwise the air will not be able to “follow” the wall of the duct and the expanding fitting will not have the best possible loss due to the turbulence associated with the air flow separating from the duct wall.

Because  of the orientation of the fan and its location directly below the supply duct heading up the shaft, the system effect and closely connected elbows in a high velocity fan discharge on the supply side have been eliminated.  Here is a closer view of the fan and discharge to illustrate that. (I took away a couple sections of the outdoor air duct so you could see the fan discharge more clearly).


(Back to Contents)

Reorienting the Fan

The preceding, of course, is only one option available to you during a design process. Most design processes are really a string of engineering compromises that attempt to identify the best over-all solution given the constraints of time, budget, existing conditions, the building structure, other systems and equipment, etc. 

For example, maybe there are constraints that simply will not allow you to re-orient the air handling unit.  In that case, you could still reduce the impact of system effect and the two elbows in series on the fan discharge by purchasing the unit with the fan in a different orientation.   One option would be an up blast configuration like this.


Note that there is a subtle but important difference between the image above and the image below.

New Construction Option 1 v2 - Fan Flipped

Here are closer views side by side.  The fan from the bottom picture is on the left and the fan from the upper picture is on the right.

New Construction Option 1 v2 - Fan Flipped Fan close-up  New Construction Option 1 v2

Both fans are in an up blast configuration, but the right image is termed “Counter Clockwise Up Blast” while the left image is termed “Clockwise Up Blast”.  If the duct on the discharge is long enough for a fully developed profile to evolve, either up blast configuration would be satisfactory. 

But, as the elbow moves from the fully developed flow profile position towards the fan, it will have a bigger negative impact on the fan in the left image as compared to the fan in the right image (see the first comment associated with this post for an explanation of the blue italic font). So bottom line, for our particular situation, we would want the up blast configuration shown on the right to minimize the system effect.

Since we are now talking about a design process vs. a retrofit process, it will likely be easier to run the supply duct as high as possible, maximizing the distance between the supply fan discharge and the elbow, which minimizes the system effect.  I say it will be easier because in a design scenario, missing things like ducts, light fixtures, sprinklers, pipe, etc. is a function of coordinating with the other design team members vs. the need to deal with things that are physically already in place, which is the challenge in a retrofit situation.

In developing our answer key, Ryan suggested an even more elegant solution that would provide a 40% increase in the distance between the fan discharge and the first elbow relative to the options shown above.  Specifically, you could purchase the unit with a fan that was configured for a top angle up blast discharge as illustrated below.


Note also that this solution would also minimize the elbow loss since the loss through a 45° elbow will be less than the loss through a 90° elbow.

The fan orientation is a standard AMCA arrangement, so it is nothing out of the ordinary to ask for it when you are starting with a clean sheet of paper.  But taking a fan that was purchased in a different configuration and re-orienting it in the field is a totally different thing.  And while it is probably not impossible, it would take some engineering and a team of field professionals with very specialized skills to accomplish it.  Thus, its typically not an option in a retrofit situation. 

(Back to Contents)

Using Plenum Fans

The two options discussed up to this point have both taken advantage of the relatively tall ceiling height in the mechanical room.  Unfortunately, we are not always blessed with this much vertical space in an equipment room.  And, if rotating the AHU, which was illustrated as the first possibility, is simply not an option, then it may be worth considering plenum fans, as illustrated below (note the lower ceiling height compared to the previous examples).


(Back to Contents)

A Bit about Fan Efficiency

Contrasting Plenum Fans  with Fans that have Enclosed Fan Wheels

Plenum fans are single width fan wheels that do not have a housing or enclosure (a.k.a the fan scroll) around the wheel.  Here is a plenum fan next to a Double Width Double Inlet (DWDI) fan with the same wheel size so you can compare them. 

Fans side by side v2

Notice how the wheel (the round silver part with blades) and the inlet cone (the funnel shaped part directing air into the wheel) on the DWDI fan on the left is very similar to what you would get if you took two of the wheels and inlet cones from the plenum fan on the right and mounted the wheels back to back.  The scroll is the housing that surrounds the wheel and inlet cone assembly on the DWDI fan.

(Back to Contents)

How Fans Work;  An Analogy

By virtue of physics, the wheels of both fans will fling air in all directions from their perimeter.  If you visualize kids on a merry-go-round you will get the idea; i.e.,  put a kid in the middle, spin the wheel, and the kid gets flung to the side, usually with much giggling being triggered (don’t do it too much or they might barf).

Merry-go-round 06

With no scroll surrounding the fan wheel, plenum fans simply “fling” the air that comes through them in all directions into the compartment where the wheel is located (a.k.a the plenum).  So, bottom line, they move air into the plenum and pressurize it.  The geometry and configuration problems associated with connecting a duct to the fan discharge go away simply because there is not a discharge to connect to.  Thus, you get a vary compact arrangement, which can be a benefit in a tight mechanical room, especially if you have to connect multiple ducts to the fan.

(Back to Contents)

Plenum Fans;  Compact, But At a Price

But, this benefit typically comes at a price.  Specifically, the scroll on the DWDI fan makes a significant contribution to the fan’s efficiency because it is shaped to help direct the air to the discharge and convert some of the kinetic energy imparted to the air by the fan wheel (the energy of motion;  i.e. the energy associated with the velocity of the air coming off the wheel) to potential energy, which shows up as a static pressure, which is what we want from the fan.

That means that most of the time, all other things being equal, a fan wheel with a housing, like a DWDI fan will be several percent more efficient than a plenum fan moving the same amount of air at the same static pressure.  For example, below are fan selections for the AHU in the example that I made using my Twin City Blower selection program

DWDI fan wheels are the first table and plenum fan wheels are the second table.  Note how efficiency varies both with the fan type and the wheel size.  (the S.E. column is the Static Efficiency of the wheel and the M.E. column is the Mechanical Efficiency).



So, if you can use a housed fan wheel and provide a good discharge condition, you probably are better off in terms of efficiency.  But, if there is no way to provide a good discharge condition for the fan, you may come out ahead by using a plenum fan if the added efficiency associated with a housed wheel will be squandered in a poor discharge connection.  The plenum also will allow multiple ducts heading in multiple directions to be connected directly to the AHU.

(Back to Contents)

Plenum Fans Can Be Constrained Too

In the case of our example, the size of the AHU casing further constrained the fan selection;  i.e. a single plenum fan sized for the total flow would not fit in the fan module. So, two fans, each sized for 50% of the total flow are provided.   This provides some measure of redundancy (if backdraft dampers are installed) but probably increase the first cost if for no other reason than the added electrical service for the second motor.   And, it increases the complexity of the system due to things like the back draft requirements and the related interlocks and control logic required to sequence the fans. 

Finally, all other things being equal, smaller fan wheels and smaller motors will be less efficient than larger fan wheels and larger motors.  So there would probably be several percentage points of efficiency benefit to be achieved by using one larger fan and motor vs. two smaller ones along with some other benefits.

(Back to Contents)

The Trick to Applying Plenum Fans

All of that said, the trick to using a plenum fan is to ensure that the connection you make to the plenum with the duct (or ducts) that it serves is a low pressure loss connection.  Typically, that means the opening in the plenum will be significantly larger than the duct size.  Generally, designers target 500 – 800 fpm face velocity into the opening and then gradually narrow the duct down (at an angle no greater than 15°) to accelerate the air into the supply duct.   Here is a close-up of the transition fitting from the plenum to the supply duct in the example.


The reason for targeting a low entry velocity into the supply duct is that the loss through the entry is a function of the velocity pressure at the plane of the entry.   The velocity pressure associated with 500 fpm is 0.0156 in.w.c., meaning that with a loss coefficient of .95, the pressure loss associated with the entry would be 0.0148 in.w.c.;  a very small loss.  The fan energy at the fan shaft (before drive and motor efficiency losses) associated with this loss for the plenum fans that were selected (74.8% efficiency) at the design flow rate of 18,000 cfm is .056 bhp or about 42 watts.

In contrast the velocity in the supply duct is 1,846 fpm and the velocity pressure associated with it 0.2125 in.wc.  So, if the duct had been connected directly to the plenum with no transition, assuming the same loss coefficient of .95, then the pressure drop associated with getting air into the duct would have been in the range of .2019 in.w.c.   If you convert that to fan power, it is in the range of .764 bhp or 570 watts (over half a kW and more than an order of magnitude higher than the transition type connection).  So bottom line, the details of the duct connection to the plenum make a big difference.

(Back to Contents)

Fan Wheel Designs

In closing, it is worth noting that in addition to the different housing and discharge orientation options we have been discussing, there are different fan wheel designs available.  All of these wheel designs can be used in a plenum fan or in a housed fan.  Common designs are forward curved, backward inclined, and airfoil.  Here are a few slides  that compare them.




Incidentally, if you want more information, these slides are from the class we do at the PEC on Fans, Ducts and Air Handling systems, which will be offered live and as a webinar on November 4th this year (2014)  You can check out the class schedule from a link on the right side of the blog home page under Commissioning Resources (blue arrows below).  We usually do it about once a year.   You can also link to the latest set of class materials from a bit further down the page under Materials from Classes and Presentations  (the red arrows below).


Anyway, all other things being equal, of the three wheel designs, the airfoil wheel will be the most efficient while the forward curved wheel will be the least.  But the forward curved wheel will generally cost less, meaning that in a first cost driven environment, a performance spec for a fan that lacks any guidance with regard to efficiency will likely be met with a forward curved wheel, assuming some of the other characteristics like the double bump in the performance curve don’t rule it out.

Here is the forward curved fan selection options for the AHU in the example.


If you compare this table with the previous table, you can see that even the best forward curved fan selection sacrifices significant efficiency points compared to some of the other options. 

In my experience, the added cost associated with the more efficient fan wheel will just about always pay for itself in 2 to 3 years or less if the fan runs 2,000 or so hours per year or more.  As the operating hours increase, the payback gets better.  All in all, it’s a pretty good deal when you consider that most fans have a life expectancy of 20-30 years.  In fact, I just ran into one that was my age and I am going on 60. 

But, the trick here is that you only need to justify the incremental cost difference associated with the more efficient selection if you make the change before the fan is purchased; i.e. during the design process.  If you already have the fan installed, then the savings will have to justify the entire cost of replacement, including the a new fan, installing the new fan, and removing the old one, a number that gets very big, very fast.

So, if you are doing design review, another thing to check is the type of fan wheel;  making an improvement in that regard during the design process will capture significant savings for the life of the system, savings that will likely become cost prohibitive once the fan is in place. 

(Back to Contents)

The Bottom Lines (and a Few Ripple Effects)

So hopefully, at this point you realize that there are savings to be achieved by paying attention to the way we select and apply fans and fittings in our air distribution systems.  And while it may be possible to capture those savings via a retrofit in an existing building, the real opportunity lies in capturing them from the start, during the design process.

Folks involved with the day to day operation of systems, like commissioning providers, retrocommissioning providers, and facility engineers and operators have a valuable role to play in the design process by bringing the lessons from the field back to the drawing board during design review.   In addition to saving energy, there may also be first cost savings to be achieved by optimizing the equipment selections, minimizing the amount of sheet metal required, and reducing the support required from the utility systems serving the AHU, like the the power demanded from the electrical system and fan heat imposed on the chilled water system.

Its important to recognize these savings because they will compound themselves back through the utility systems serving the AHU.  On the electrical side, they will be compounded by the reduction in power that needs to be delivered through the grid, the building  distribution system and the drive systems, none of which are 100% efficient.


On the chilled water side, the reduced fan power shows up as a reduction in fan heat which is a cooling load on the system.  If the belts and motors are in the air stream, then the fact that they will need to develop and transmit less power means that the heat associated with their inefficiencies will impose less of a load on the air stream.  Those benefits will ripple all the way back through the central plant to the cooling tower. 


And remember, all of those central plant motors are typically supplied from the grid and will see compounding benefits on the electrical side as well. 

In the even bigger picture, reducing the demand for electricity from a system that on a national scale, is driven largely by the combustion of fossil fuels will reduce the emissions we reject to the atmosphere from our power plants.


I know that is a touchy subject, but if we really understood how the atmosphere worked, the weather forecast would be right every single day.  Since we are not at that point yet, it seems to me that logically, we should be careful how we interact with something as complex as the atmosphere especially given that it is critical to supporting life here on earth.

I won’t make a long post any longer by elaborating on all of these compounding benefits, but they are important and significant, which is why I mention them.  The slides I used to illustrate them are also from the set for the PEC class and you will find additional information there if you want to learn more.  And, a while back, I did a post on site vs. source energy, which is part of the reason for all of the compounding benefits, so you may want to take a look at that if I tweaked your interest.

(Back to Contents)


David Sellers
Senior Engineer – Facility Dynamics Engineering

Click here for a recent index to previous posts

This entry was posted in Air Handling Systems. Bookmark the permalink.

One Response to System Effect–Dealing with the Point Where the Fan Meets the Duct

  1. Friend, fellow RCx consultant, and alert reader Sabastian St. John e-mailed me yesterday to ask a couple of good questions/make a couple of good observations about this post.

    One was related to the close up of the two fans side by side in the section where I am talking about reorienting the fan as a solution for a new construction project. Specifically, Sabastian referenced the second paragraph after the side by side image and observed:

    “I would think that the configuration on the left would experience a bigger negative impact as the elbow moves closer to the fan and that is why we would prefer the configuration on the right. Is this not the case?”

    He is absolutely correct and the earlier version of the post had a typo, which I have corrected now with the change in blue italic text. All I can say in my defense is that I have no instinctive sense of right or left.

    Meaning that in this case, clearly, I forgot to visualize Mrs. Moore, my kindergarten teacher, standing with her back to me holding out one arm and saying “this is my left” … and then lowering that arm and raising the other, saying …” and this is my right”. For whatever reason, if I don’t go through that process or something like it, there is a 50/50 chance I will get it wrong, as I did in the original post. Kathy (my bride) is quite accustomed to saying “No, the other left” (or right) as a result.

    Sabastian also noted that the shape of the performance curves for air foil and backward inclined fans appeared to be virtually identical. That is true; to some extent, an AirFoil fan (AF) is just a Backward Inclined fan (BI) with an streamlined shape given to the fan blade. The blade is literally a flat plate of metal in a BI fan but looks like the cross section of a wing in an AF fan.

    As a result, the only real difference in the performance curves, all other things being equal, is an AF fan will be more efficient than a BI fan. I try to convey this by having the peaks on the curve for the BI fan example a little lower than the AF example. But the grid you would use to reference that does not show up very well in the screen capture I used to get the image into the post.

    But if you go and pull the slides back from the class they are from (its the second model titled “02 – Fans v5.pdf” in the Fans, Ducts and Air Handling Systems class materials, the grid lines show up much better there.

    Because of the more complex blade shape, all other things being equal, and AF fan will be more expensive than a BI fan. Both fans will be more expensive than a Forward Curved (FC) fan, but also more efficient. While a FC wheel has a lot more blades than a BI or AF wheel and those blades are curved, they wheel is typically made by stamping an bending metal. In contrast, BI and AF fans are typically fabricated from metal plate and extrusions that are welded together; much heavier metal gauges than the FC wheel and a much more time consuming manufacturing process.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s