Air flow sensor

ASF

Lifetime Supporting Member
Join Date
Jun 2012
Location
Australia
Posts
3,907
Hi all,


I have an application where I have to measure air flow through a HVAC duct, as we have to ultimately maintain a certain rate of air "changes" in a room of known volume in a given time (e.g. the room is say 20m3, and we have to completely replace the air in the room at a constant rate of once per hour, so we need to maintain an air flow rate of 20m3/hr). I've not used air flow sensor before so I'm delving into this world for the first time, and most of the options I'm finding are air velocity sensors.

Some googling tells me that this is generally the way air flow is measured - by measuring the air velocity and calculating the flow. Velocity multiplied by cross-sectional duct area equals air flow. That seems straightforward enough, but I'm just curious about the range of different options for air velocity sensors. There are sensor like this one that basically become part of the duct, and then others like this one that just stick a small probe into an existing duct. I'm assuming that the first option which becomes part of the duct and contains multiple sampling points across the duct would give me a more accurate reading than the single-point measurement of the second probe, but how much difference are we talking about? Does anyone have any hands-on experience in this field to offer?
 
That's a little up in the air at the moment. Part of the problem is that the client wants (slash needs) to squeeze this HVAC system into quite a small space. This means that we're unlikely to have a lot of nice long straight sections to measure our flow, which means that since we won't have a nice uniform flow across the duct our accuracy will be reduced. Right now the conversation is along the lines of "if you really must squeeze it into that space, it will come at the expense of accuracy" and "if you spend more money on the full-duct sensor you'll get better accuracy, especially given the space constraints", but I'd like to be able to give them some idea of scale when I say "more accurate" and "less accurate".
 
You are correct in saying that almost all flow sensors measure velocity. About the only exception is the positive displacement meters used for liquids in custody transfer - for example, the water meter in your basement.

The accuracy improvement from an averaging type unit, like the grid, is hard to quantify, because the inaccuracy of the single point instrument is unknown. It depends on the duct conditions. If you are immediately downstream of a damper or elbow the single point measurement is probably garbage. If you ten or fifteen "diameters" of straight duct it might be very good and the grid won't give you much improvement.

In other words, "it depends". Sorry.

This may be a lower cost altenate: http://www.dwyer-inst.com/Product/AirQuality/PitotTubes/Series160S
 
Of the two options that you posted, I would go with the averaging pitot tube one (the VGS Velogrid). It doesn't require a long lead-in or lead-out, and other than getting clogged, tends to be pretty accurate and foolproof.

We've used them in the past (and other than paint inexplicably making it all of the way up to them and clogging their ports) and they did just fine.

I don't have any hard and fast performance numbers, that's best left to the salespeople. ;>
 
Thanks guys. At the moment we're looking at these units due to long lead time on the velogrids. They look similar to the ones Tom suggested, and they involve two sensors diagonally across the duct. So arguably better than a single-point measurement, but I still can't see them being super accurate unless we can get some decent straight lengths of duct to work with.
 
The AFG and the VGS Velocogrid both use averaging pitot tubes for volumetric air flow measurement (CFM). I have seen the use of two crossed lots of times, I was unaware of the VGS matrix until you posted it. I would consider the AFG style as accepted in the industrial world. The DP transmitter necessary to read the pressure drop and convert it to a flow rate can run from relatively inexpensive (Dwyer) to relatively expensive (any of the industrial DP transmitters).

The thermal dispersion units (EE650) actually provide mass air flow measurement (SCFM or pounds of air/time unit) but it's the air flow at the measurement point as opposed to the averaged pressure across the 'diameter' of a duct.

The Dwyer Pitot tube (160S) is a point measurement, too, not an averaging measurement.

The averaging pitot tube drops very little pressure, so make sure that your DP unit is sufficiently sensitive and has a low range that can handle the low drop.
 
Thanks Dan, that's really helpful information.


The velogrids only require 1.2 duct diameters upstream/downstream, and you can go even less if you install a "flow straightener" before it. Whereas the "crossed over" sensors say they need 3.5D to get +/-15% accuracy and 10D to get +/-10% accuracy. I don't think we'll get 3.5D, let alone 10.
 
Good luck finding published information on 'upstream disturbance error' for any flow device.


There's an IEEE paper, but one pays to read it.


https://ieeexplore.ieee.org/document/7053301/n


No manufacturer ever has published installation non-compliance error data, presumably for fear that a customer might try to hold them to published inaccuracy specs.


in a similar vein, for years, magmeters were spec'd for 5 diameters upstream, 2 or 3 diameters down. The sky would fall if that straight were not maintained.



The Europeans accept a ±2% of max flow rate accuracy for public water distribution metering. Siemens had their meter tested with upstream and downstream disturbances - single or dual elbows in plane and out-of-plane and the magmeters passed the ±2% test.

What a revelation that was to me, because I had no idea nor did my customers. Most industrial customers will live with ±2% error if they know that's what the maximum error is, but prior to Siemens publishing the test results, no one knew what the magnitude of upstream disturbance error was. One can't test flow like one can test pressure (deadweight) or temperature (furnace, dry block).

It appears that most magmeters have the same performance characteristics.


So there's a published inaccuracy value for magnetic flowmeters, but no such (openly) published data for averaging pitot tubes.
 
You can use the fan curve for your selected fan and feedback from the inverter to get air flow.
Are you expecting a temperature difference between air@fan and air@room?
 
You can use the fan curve for your selected fan and feedback from the inverter to get air flow.
Are you expecting a temperature difference between air@fan and air@room?


It's a little more complex than just a fan circulating through a duct. There are multiple supply and return air fans, motorised dampers, various routes through heat exchangers/heating elements, and potential for closed loop circulation, or drawing in fresh air from multiple sources/temperatures.
 

Similar Topics

Do you know any supplier for air flow monitor (vent captor), used inside dc motors for cooling protection? It should be a small plate switching a...
Replies
1
Views
2,141
Hello ! I need help to understand these arrows on this solenoid this solenoid is activated electrically by 110 VAC and it is being used to...
Replies
7
Views
1,881
On site we have Yokogawa EJA110E differential pressure transmitter connected to process air line. I'm not sure if this can be used to measure air...
Replies
9
Views
4,357
Hello, I'm building a system for my professor and for controls I will be using a PLC from automationdirect as a way of getting into the field and...
Replies
4
Views
1,696
Hi guys, We want to be able to test air flow in capillary tubes, roughly .1" in diameter, and 3' long for the tube. Air will be the medium for...
Replies
7
Views
1,966
Back
Top Bottom