[Bldg-sim] air flow sensors

[Paul Carey]

Dear Chris and all

I had mentioned to Chris outside the forum (sorry should have copied all in) that it might be useful carrying out a tracer gas decay test. As many of you note the hot wire anenometer will not sense direction and the use of pitot tubes or a flow grid may not be suitable for such low pressure ranges. This is the reason that I suggest a decay method may be the best way to analyse the overall flow rate through the room. This is the way I measured flow rates in both model environments and test rooms when I did my PhD. The main reason for this is the unsteady nature of ventilation flows themselves. Ventilation by its nature is not unidirectional across an opening it is an unsteady I.e. varying and fluctuating pressure across the opening that if you measure it with a transducer will have positive and negative amplitude all the time. It is the differences in the combination of buoyancy and wind driven pressures around the building that cause a variance in the mean pressure differences that drive the flow in one or other direction overall. Obviously this combines with other factors such as opening position, size, etc. The tracer gas test will get around this unsteady issue as you are then looking at a time decay method that will give you the decay rate and therefore a direct relationship to the flow rate into a room. Fairly easy to test, a relevant gas sensor, datalogger (though many sensors now have them built in) and a can of trace gas like SF6 or preferably something less ozone depleting but easy to test for.

I recommend that anyone studying ventilation flow reads the book on ventilation theory and measurement by David Etheridge and Matts Sandberg. It is an invaluable resource and will answer many questions people may have regarding these sorts of studies.

Regards
Paul

Dr Paul Carey


Sent from my HTC

----- Reply message -----
From: "Chris Jones" <cj at enersave.ca>
Date: Wed, Jul 4, 2012 14:46
Subject: [Bldg-sim] air flow sensors
To: "Ken Kolkebeck" <ken at firstfuel.com>, <bldg-sim at lists.onebuilding.org>

Thanks Ken and everyone.  It sounds like a big challenge.  As it is a spare time research project, I am going to keep expenses as low as possible.

One issue I realize will be a big problem is spot measuring all open windows at once - in the time it takes to move from a ground floor window to the second floor, the conditions can change significantly.

The windows have insect screens which may eliminate some of the cross currents outside the window.

Once I obtain the measuring instruments, I will be back for more advise.



At 04:30 PM 03/07/2012, Ken Kolkebeck wrote:
>I'm not 100% sure I understand the application but I will assume it is for
>some type of test. What you are trying to measure (natural draft or
>induced draft) will be important because that will dictate the velocity
>you are dealing with. The higher the velocity the easier to measure. The
>lower the more difficult and the more subject to disturbances on either
>side of the aperture.
>
>The problem is somewhat similar to that of measuring the velocity across a
>fume hood where measureable velocities run in the 50 to 150 feet per
>minute range. These are of course induced drafts which while low in
>velocity are high enough to be measured. The fact that modern fume hoods
>have somewhat aerodynamic entry which reduces turbulence enhances the
>ability to get clean measurements. Even when using the most accurate
>device, traverse technique and cross currents are the biggest causes of
>errors which can run 25%.
>
>Two methods are commonly used here-
>Hot (or Warm) wire anemometers as suggested by another responder. TSI
>being a common brand. The issue here is that these do not sense direction
>and on the face of an open window are going to be impacted by cross
>currents caused by wind outside the building. The other thing is you
>really need a grid of them across such a wide area because the gradient is
>likely to be large.
>
>Shortridge "Velgrid" which is in essence a pitot type device however the
>key to the way it works is with a very sophisticated pressure sensor which
>has all kinds of error corrections built in. These are not your standard
>HVAC pressure transducers as they are designed to operate in the
>differential pressure ranges well below 0.01 inches water column. The
>transducer zeros itself before each reading to null out errors associated
>with hysteresis and thermal effects.
>
>The pitot is actually a one by one grid of high and low sensors and has a
>multiplier effect so the velocity is going to be about 2000 times the
>square root of the differential in inches of water. 100 fpm therefore gets
>you about 0.0025 inches; 50 fpm gets you a quarter of that. For fume hoods
>we always mounted the grid on a stand because it is difficult to hold
>steady and the movement induces errors. Unfortunately, this device too is
>subject to cross current errors.
>
>A flow hood, also made by Shortridge, should work provided there are not a
>lot of other windows open which will be the paths of least resistance and
>will impact the reading because the hood induces a back pressure.
>
>Good luck and have fun. Sounds like a tough one.
>
>Regards,
>Ken Kolkebeck
>Certified Automation Professional, ISA
>Certified Building Commissioning Professional, AEE
>NJ Office: 201-784-8700
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/bldg-sim-onebuilding.org/attachments/20120704/44305d2c/attachment-0002.htm>