Inlet Guide Vanes (IGVs) and Variable Speed Drives(VSDs)

First, I’ll apologize for my really low rate of posting over the past couple of months.  Its been a busy summer, both personally and professionally and with the blog being something that I generally do in my spare time as more of a hobby than not, it has suffered in terms of the frequency of posting.

The reality is that the line for me between work and hobby and work and spare time is a bit fuzzy.  Frequently, after a class or out in the field, I am asked a question that turns into a fairly long, technical discussion, often in the form of a follow-up e-mail, usually with illustrations of some sort.  When I’m done with it, I’m never sure if I just finished work, just finished playing or both.  Probably the latter in most cases;  I always learn something when I do that and I enjoy learning.

More than once, somebody, having read through an answering e-mail has responding saying “that would make a good blog post”.  And after having had that happen the other day, I thought maybe I should start doing just that;  post my answers to technical questions if the topic seems worthy of publication.  If nothing else, it will likely increase the frequency at which I post since I usually do a long technical answer every week or two, just like the one that I did last week on inlet vanes, which is exactly where this post came from.

Specifically, one of the students in the Existing Building Commissioning Workshop Series that I teach with Ryan Stroupe and others at the Pacific Energy Center e-mailed to ask if, when retrofitting a fan with inlet vanes to a variable speed drive, you should remove the inlet vanes.   On the one hand, they reasoned that the inlet vanes represented an unnecessary pressure drop.  But on the other hand, they felt that they may provide some benefit by means of directing the air flow into the eye of the fan.   

It was a great question accompanied by some great observations along the lines of why  you might want to be cautions before removing the IGVs.  And like most HVAC stuff, the answer is “it depends”, at least as far as I have been able to determine.  Before I get into my answer, I thought it might be helpful to show a couple of pictures of inlet vanes on a fan for those who may not have had the opportunity to see them before.  Here is a picture of a set of inlet vanes on the drive side of a double width, double inlet fan (DWDI) …

… and here is what they look like at the other end of the fan shaft.

Here are close-ups of the linkage system at the hub …

… and the perimeter.

When the vanes actuate, each individual blade rotates around the shaft running from the hub out through the inlet cone, with all blades moving in a synchronized manner.  If you think about that for a minute, you can probably see that this would tend to direct the air flow into the fan wheel. In other words, the potential impact goes beyond a simple restriction to flow.

Incidentally, I believe these pictures were actualy taken by a student in one of the first EBCx workshops we did at the PEC.  My recollection is they brought them in to ask a question regarding the installation.  Sadly, I did not made a note of who it was;  I simply saved the pictures to the technical pictures folder on my hard drive.  But they were much clearer than the ones I have of a similar installation, so in the interest of education, I have used them here.  My thanks to whomever provided them;  if you recognize your handi-work, please let me know so I can give credit where credit is due.

Here is a slide I use in class when I talk about this and the impact inlet vanes have on fan performance and the fan curve.

And here is the slide I use to illustrate the effect of varying fan speed, to do the same thing,  just as a frame of reference.

Note that in both of these illustrations, the dashed horsepower curve only applies to the full speed, non-actuated inlet vane fan curve. 

The chief advantage of the variable speed approach when contrasted to inlet vanes is it does a better job of preserving the fan efficiency.  Note that the inlet vanes cause the fan curve to “droop”.  Changing the speed creates a little rainbow of curves similar to the base curve.  If you were to plot constant efficiency lines on the curves, you would discover that they follow the reduced speed curves where as the “drooping” curves created by the inlet vanes tend to cut across the constant efficiency curves.

If you study the pictures a bit, you may conclude that a variable speed drive is also a lot less complex mechanically than the inlet vane system (but not necessarily technically when you think about what is going on inside a variable frequency drive, one of the common ways to vary fan speed, but not the only way).

This figure, scanned, from my copy of the Buffalo Fan Engineering Book, illustrates the impact inlet guide vanes have on the flow of air through a fan wheel.

If you are interested in the details of what the Buffalo Fan Engineering Book has to say about this, I have uploaded a .pdf scan of this section of the section of the manual that discusses inlet vanes into my Google Documents account.

But even with out that, I think the vector diagram gives you some insight into the fact that the inlet vanes will impact the details of the flow through the wheel.  And it’s not a big stretch to imagine that there might be a specific vane position – which translates to a specific resultant velocity vector through the fan wheel – that could optimize performance for a give wheel at a given flow rate.

In general terms, the discussion in the Buffalo book indicates that there are trade-offs related to fan efficiency, immunity to impacts from upstream disturbances (i.e. the vanes act as flow straighteners), fan wheel blade fabrication complexity (a flat plate vs. a plate with a curve at the heel), and mechanical strength of the wheel that come into play when a fan designer considers whether or not to apply inlet vanes.  Many of them are things you probably should consider if you are trying to decide whether or not to remove the inlet vanes when you make the conversion to variable speed operation.   This seems to be especially true for backward inclined fans and radial blade fans. 

What I have taken that discussion to mean is that you should be cautious with regard to removing inlet vanes on a retrofit. If the specific fan design required fixed inlet vanes to achieve peak efficiency and the fixed vanes were made variable to accommodate that in addition to allowing the fan performance to be varied, then in a retrofit, you might want to retain the inlet vanes and lock them to the position associated with peak efficiency.

If, on the other hand, the vanes were an after-market addition, or even a factory option but installed purely to facilitate modulating the fan capacity, then they are not necessarily associated with maintaining the peak efficiency of the wheel and they may represent a pressure loss or system effect impact that is not required or desired if the fan is retrofitted with a VFD.

That said:

• Since the idea of fixed vanes improving wheel efficiency is related to manipulating the velocity vector relationship for the air entering the wheel, and
•  Since, if the wheel size and inlet size are fixed, the velocity will vary with flow,

I think it is possible that a fan that is operating over a range of flows due to variable speed operation may in fact have a different “ideal” fixed position for each speed.   I say this as my logical extrapolation of the other information I mention above.  Others who know more about centrifugal machine design than I do may be able to verify or refute me (and they should, especially if I am wrong). 

As an aside, another reason I think the preceding may be true is that I think it is part of what goes on with the control algorithms that are running on VFD equipped centrifugal chillers – which are centrifugal machines, just like fans.  The ones I have watched seem to adjust inlet vane position along with drive speed to optimize the compressor performance.

Returning to our inlet vane discussion, Dave Brown and I had an opportunity to do a crude science experiment in this regard when we first started our retrocommissioning efforts at the facility he was working in at the time. Specifically, there were 4 air handling systems serving meeting rooms in a high-rise tower where inlet vane equipped fans that had been retrofitted to VFDs with the vanes abandon in place.  As is often the case, the abandonment was not robust; in one instance, bailing wire (literally bailing wire) was holding the linkage for the vanes to keep them open. 

Now, before you start thinking unkind thoughts about facility operators tolerating, or maybe even implementing this this sort of arrangement, I would like to digress for a moment and point out:

• The situation is more common that you think (or would like to believe ) (or would like to hope for).  Sometimes, the retainer is a tie-wrap instead of bailing wire.  Sometimes it’s a bolt or screw through the hole in the crank arm where the actuator clevis used to connect (like the picture below)  Sometimes it’s simply the (hopefully) normally open actuator abandon in place with no signal to it.  Seldom is it the image whomever had in mind when (if) they said that the work associated with the retrofit included disabling the inlet guide vanes as a part of the VSD retrofit.

• My guess on why this happens is that many of the retrofits were likely the result of a utility incentive program.  Since the retrofit installed a VFD, the program incented the Owner to hire an electrician to install the drive.  The electrician may or may not have understood inlet vanes let alone come prepared to disable them in a robust manner.
• Incentive programs often use “cookie cutter” pricing, which basically means the “lucky” installer has a fixed (usually very tight) budget to accomplish the work.  It’s highly likely that the budget was developed by someone who did not think about there being a need or a cost associated with disabling the inlet vanes.  Then, the work probably was bid competitively. If you have been involved in that process, then you probably know that many times, the reason you are low bidder is that you missed something (and sometimes the reason you are high is you know too much from having already worked on the facility or system).  So, in the case of the low bidder, you may already be “in the hole” and even less inclined to spend a lot of time and money disabling inlet vanes that you did not know existed in the first place.
• As you can see from the pictures previously, the mechanism associated with making the inlet vanes work is located on the fan housing.  For most air handling systems, the fan housing is contained inside a casing, and for many systems, the casing was furnished with out an access door.  (Incidentally, noting the absence of such is a good design review target, both during the construction document phase and the shop drawing review phase of a commissioning process.)  That means that the operator may not even know the inlet vanes were not secured until the first time they blow closed.  At that point, after obtaining access via  the removal of  innumerable bolts/zip screws or finding a reciprocating saw  in the instances where an access door has not been provided (probably 50% or more of the instances)   you make a temporary repair with bailing wire or a tie-wrap because that is what you have in your pocket at the time.   And, you fully intend to improve upon that later in the day after you put out the other “fires” that have occurred.  (All of this happening with your cell phone ringing out of its holster due to people calling to say they are hot). 
• You remember that you intended to improve on the bailing wire installation as you pull into the driveway around 2 am after dealing with a domestic water line that broke around 4 pm.  So, you make a note to rectify that first thing when you get back to the building (at 6 am, which is when you start). 
• Walking through the door at 6 am the following day, you run into a very irate tenant who came into work early and discovered their carpet was soaked with water (see previous bullet).

I know these sorts of things happen because I have “been there and done that” as in I have not include obvious (in hindsight) requirements in scopes of work, been the low bidder, and been the guy getting home at 2 am who has to be back to work at 6 am. 

So, hold your tongue lest you be smitten with an inaccessible inlet guide vane system concurrently with the activation of a sprinkler system caused by somebody hanging their freshly ironed shirt on a sprinkler head.

Sadly, there are reasons people place these little signs and symbols around the equipment in their facilities.

Returning to Dave Brown, me, the start of the a retrocommissioning process and the AHUs with abandon inlet guide vanes, occasionally, the retention system would fail, the vanes would blow closed, rooms would overheat, complaints would roll in, etc.

On one system, someone had upgraded from the bailing wire to a zip screw driven into the inlet cone that the blade ran into if it tried to closed.

 But sadly, the screw was not stainless, so while more durable than the bailing wire, eventually, it would probably fail, especially given the marine environment at the location.

So, Dave, having lived with the problem for a while, and having gained some insight into system effect, pressure drops, and the relationship between fan power and static pressure was thinking great, we can cut those suckers out of there and solve the nuisance problem and get some energy savings for the RCx project to boot and was all ready to do just that when I brought up the discussion about the potential for the vanes to have an impact on fan efficiency.

So, he checked that out and determined that the inlet vanes were not an integral part of the fan design. Based on his research, we decided to cut them out. But, before we did that, we set the fan up for a flow rate based on a traverse of the mixed air plenum at the filter bank with a velgrid and locked the speed down and documented the fan power based on the drive kW.

Then we got in there with Alberto and a reciprocating saw and cut out the vanes.

If you have never done something like this, it may seem like no big deal.  But the reality is its pretty gutsy.  Basically, you are taking something that was working, albeit not as efficiently as possible, and removed what may be an important part in a manner that is very difficult to undo.  If you are wrong, you could be in “big doo-doo”.

After we cut the vanes out, we restarted the fan, set it back up for the flow we tested at previously, and discovered that, as near as we could tell, the operating point had not changed.  In other words, we were moving the same amount of air at the same fan speed with the same kW drawn by the drive.

So, to us, that said that for this particular case, the wide open vanes had no measurable effect, positive or negative, on the fan performance. Given that the fan was in a casing, with dimensions meeting the rules of thumb for spacing between the fan wheel and casing wall and the fan wheel and coil, the vanes probably did not provide any flow straightening either. And the fan bearing was supported by other structure not related to the inlet vane system. So, bottom line was that the only reason to remove them would be to save energy and when we did that, it had no impact.

We also surmised that the same may not be true in the general case. For instance, this was a modular package unit, meaning that the fan section was designed to accommodate a range of flows and a number of different fan wheel sizes and designs. And, in fact, most fans, while they have a very specific operating condition associated with peak efficiency, in our industry, will be applied over a range of conditions as an off the shelf solution to the requirements of a particular project. I think this means that:

• For other fan wheel designs, the vanes may have been required for peak efficiency, specifically a BI fan based on the Buffalo Fan Engineering Book information (this might also be true for a radial blade fan but those typically are used for exhaust, usually dirty exhaust and I don’t seem them much in HVAC applications). So, you can’t assume that you can just cut them out carte’ blanche.
• If the fan happened to be applied at the low end of its range, in terms of flow, the velocity into the eye of the wheel would be low relative to the velocity you would see at the high end of its application range. So, at higher velocities, the impact of the vanes might be more significant in terms of pressure drop or system effect, especially given the square relationship between velocity and velocity pressure and the relationship between velocity pressure and loss coefficient.

Related to all of this, Bill Michell, Director of Engineering at a high end hotel in New York discovered that you have to be careful when you instruct contractors to remove inlet vanes as they can get carried away with things. He inherited an under performing system when he became DOE at the facility and when he decided they needed to go look at the fan (as in get inside the casing and look vs. imply things from measurements), he discovered the inlet cone had been removed along with the inlet vanes. Here is a slide I use when this discussion comes up which is built from a picture he generously shared with me.

Obviously, a missing inlet cone has a huge impact on fan efficiency since there is basically an open path from the high pressure discharge portion of the wheel to the low pressure inlet portion of the wheel.

So, my bottom lines on this issue at this point in my career are (some of these are empirical based on my field experience or analysis based on basic physics):

• Don’t just assume you should remove the inlet vanes, especially with a BI fan with flat plate blades or a radial fan.
• If you can’t find out if fixed vanes are in integral part of the fan design:
1. Removing them is less likely to be an issue for fans that are not BI or radial blade.
2. You may be able to test in the field to see if they have an impact if you have accurate enough instruments to measure power and flow consistently.  To do it, you would need to hold a fixed flow, vary the vane position, and see if you can measure an improvement in performance via a reduction in power.
3. If you abandon them in place, even if they do impact fan efficiency, it is likely that the best position is nearly wide open.  But, for a VAV system that is varying flow with a speed change, this position may vary with flow.
4. Leaving them in place and locked open is less likely to be an issue for fans that have low velocities through the eye of the wheel vs. high velocities. 
5. For a VAV fan that has high velocities through the eye of the wheel, these velocities (and the related potential impact) will only be an issue at times of peak flow and load. That means that unless the load profile is flat with a lot of time at or near peak flow, the impact of the vanes left in place is likely minimal.
• Galvanized zip screws blocking vane closure are more persistent than bailing wire. Aircraft grade stainless steel safety wire or stainless zip screws will likely show better persistence than either of the zip screws or bailing wire, especially in a marine environment. Bolting the actuating lever on the linkage to something solid may be the best of all options if you are going to abandon the vanes in place in the open position.
• If you remove the vanes, don’t remove the inlet cone.

David Sellers
Senior Engineer – Facility Dynamics Engineering

Click here for an index to previous posts