Venturi flowmeter - without upstream tapping 1

[ATPD]

Hi.
I am reviewing a vendor's design for a burner air supply which uses a venturi flowmeter to measure the combustion air flowing into the combustion air fan. The flow will be around 15,000 m³/h. So the air flow will be from atmosphere, through an inlet silencer, through the venturi and into the fan. So far so good, I think. But the vendor is saying that our upstream tapping, the one I see is usually in the duct a set distance upstream of the venturi throat, can be just in the atmosphere rather than in the duct. I'm a little uneasy about this. His justification is that the flow measurement does not need to be super accurate, just reliable and consistent, because he will "set up in commissioning" to ensure that the burner air/gas ratio is correct. We are on a tight budget, so I'm wondering if it is worth paying extra to get a more standard system, or whether this will work just as well.
Aidan

 

[danw2]

A venturi is designed to provide a volumetric flow measurement. The measurement uses differential pressure and a math conversion to get a volumetric flow rate; in the US, in units of CFM or CFH if a conventional DP transmitter is used, or an inferred mass flow rate in standardized flow units (SCFM or SCFH) if a multivariable transmitter with static absolute pressure and temperature compensation is used.

A pressure measurement with a pitot tube can provide a velocity measurement, but your venturi is not a pitot tube and even the pitot tubes measures DP with the static pressure as a reference. I mention this because the vendor's approach sounds somewhat like an HVAC guy's approach.

A tap on the low side port of the venturi is a static pressure measurement at that point, it is not the differential measurement that the factory's sizing sheet assumes for that venturi's DP-to-flow volumetric calc.

If this venturi flow measurement is being made so that it can be used as a flow rate signal in a control system for air/fuel ratio control, then by all means, insist that the high side port be used (as you understand it), because without tapping that high side point, the measurement is not a volumetric flow rate, it is merely a static pressure at that point in the venturi; which is nice, but so what?

If the measurement is just being made for kicks and your vendor is making mechanical air/fuel ratio adjustments by tweaking the air/fuel valve linkage, then what does it matter what the Dwyer gauge on the venturi reports? Does it matter whether it says it's a volumetric flow rate or a static pressure at the venturi?

As a point of interest, what is saved by not connecting to the conventional upstream tapping point and getting the volumetric measurement you paid for when you bought the venturi? Several feet of steel tubing?

 

[georgeverghese]

The downstream pressure measurement would have accounted for the kinetic energy term V2/2g, while the atmospheric measurement wouldnt, so the reading would be in error by some amount which is relative to the actual design dp across the venturi.

 

[ATPD]

Thanks danw2. I don't know what he thinks he is saving - it is probably less than 1m of impulse tubing, though at this stage of the project it is also some engineering & drawing changes, still no big deal. There's also a very short inlet duct length, so I'm pretty sure the flow will not be straight by the time it reaches the flowmeter.

 

[ATPD]

Thanks George. I'm OK if the error is consistent. It means our flow reading will be out but as long as there is enough oxygen in the off-gas we can work with it. I think I'll push for a change - I'd rather have an accurate reading I can use in a heat and mass balance later on.

 

[georgeverghese]

The V2/2g term discrepancy between the 2 taps will vary depending on combustion air flow.

 

[ccfowler]

ATPD,

You said the magic words indicating that you have ductwork between the atmospheric air supply and the Venturi flow meter. Thus, you have an unknown restriction between the atmospheric pressure and the true upstream pressure needed for accurate and repeatable meter performance. The inlet silencer further complicates the situation since some variation in fouling effects are an operating certainty that will introduce progressive drifting in performance and control. To me, this amounts to a "no-brainer" that a proper upstream trap should be required even if some significant cost is involved. Why bother with some trivial initial cost savings only to risk enduring needless start-up adjustment troubles followed by continuing operating troubles. From experience in cleaning up similar messes, the initial savings afforded by this seemingly minor shortcut will be squandered many times over due to operating problems followed by a more costly installation of the needed upstream tap and finally doing the proper calibration and adjustment work that should have been done at the beginning.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

 

[georgeverghese]

Whether this error is significant enough to be a concern may be checked by comparing this V2/2 term ( in consistent pressure units at design flow ) against the design case dp across this venturi meter to verify what the vendor is saying. Am assuming there is no bug screen or mesh on the air inlet that could get partially blocked and induce some dp also. Also check if the combustion air duct work is free draining everywhere - there should be no low points where heavy rain could collect / pool up.

The last time I worked on a forced draft fired heater, air to fuel ramp up / ramp down flow ratio controls to Shell standards worked okay, with HP side of venturi FT tap located on the ductwork - the high frequency noise issue from the FD fan / high velocity through ductwork was still on the TO DO punchlist even after this unit was handed over to Operations.

 

[ccfowler]

Perhaps I should clarify my perspective and concerns. Getting a system working seemingly nicely during startup makes everything nice for the designers and installers, and it may make operations happy for a while. It gets the construction and startup bills paid. Later, there is no small risk that operating characteristics will start to drift or operating costs (usually some form of energy consumption) will start to seem a bit high or increasing. At that happy time someone else (typically plant engineering or technical staff) will have the pleasure of trying to sort out the mess while trying to minimize disruption of operations. Almost always, retrofitting hardware costs several times what an original installation would have cost. If the subject system is as I am understanding it, the actual additional initial cost may well be on the order of $100 to $200, but getting it installed later for under $2000 may be close to a miracle without taking into account potentially tens or hundreds of thousands of dollars in wasted energy, defective product, and otherwise needless process problems or adjustments.

Sorting out a problem that actually originates with a shortcut such as this is often complicated by the likely presumption that the problem must due to something other than this shortcut because it will have the apparent "blessing" bestowed upon it by being part of the original design and installation. Most likely, significant effort will be dedicated to other potential sources of the seemingly unclear problem before the real problem is even effectively considered.

In this case, the fact that a Venturi meter is being used as the primary flow element rather than some cheaper and less certain one such as an orifice or even some less well documented element is very likely to postpone looking at this potential problem. The presumption is likely to be that with such an excellent meter being used that it and its associated parts are likely to be blameless until other other possible problems have been cleared.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.