Hello to all,

We 're dealing with an application where flow measurement in DN200 round pipe in needed.

The thing is that at times flow drops below 1 inch for a period of time. 1 inch is the lowest possible value that most of the Area Velocity flowmeters with submerged sensors are able to measure fluid velocity and thus flow. But a 2 hours operation under the above condition, e.g. Flow slightly below 1 inch, could give a big inaccuracy in our measurement.

Could anybody help in how we can deal with this?
We need to measure flow in as low as 1 cm fluid level. ISCO's signature flow meter has the option of a bubbler module (similar to AV sensor) for measuring level in depths as low as 0.003 m, but is it possible to convert level to flow??

Many thanks!
Michael.
p.s. sorry, my knowledge of Fluid Mechanics is not very high.

 

it is some times possible to use an internal weir, but only in vented rather large piping, and as with all flow meters the accuracy suffers at low flows.

the usual approach involves piping layout to insure that you always have a flooded meter even under non-flowing conditions, but to do that requires more details about your process, as it is not recommended where sedimentation and or regular cleanout is required such as would be the case in wastewater processes.

whatever solution you choose will not be cheap

 

If you are asking for accuracy when the depth is only 1 cm. and a range that includes when the pipe is full (20 cm. diameter), you are asking quite a lot. The flow is approximately the fluid velocity times the cross sectional area of the fluid flow. When the fluid is only 1 cm. deep, it has a cross sectional area of about 2% of the cross sectional area of a full pipe. Even if the fluid velocity were the same in both cases, you are already asking for accuracy at 2% of full scale. But you quite likely have a much faster fluid velocity when the pipe is full. So the situation is even worse. This range to resolution ratio is a lot to ask of any single measurement technology.

One very awkward solution for the low range end of the problem would be to accumulate fixed volumes of fluid that are emptied when it reaches a threshold. Then just count the number of emptying events. That will give good accuracy right down to zero flow. But you will still have to employ another technology for higher ranges.

 

How to do this properly depends on the answer to this question. Do you have submerged or surcharged flow ever? By that I mean is the pipe full, and does the pipe ever experience reverse flow?

If neither of those conditions exist ever, you can use a primary device. I would use a Palmer Bowlus flume with a bubbler transmitter. Using the built-in algorithm in the ISCO device, you can convert the level behind the flume to flow with some significant accuracy. The ISCO Flow Handbook can give you an idea of the potential precision of measurement, and the lowest flow you can reliably measure.

Since you have a relatively small pipe (200 mm is about 8 inch diameter), you can also handle surcharge and submerged conditions with a Datagator, by Renaissance Instruments of Austin Texas. This device is based somewhat on a Palmer Bowlus flume, but with three sensors and a top body to measure surcharge and reverse flow.

You will get something like 5% of indicated flow rate, or perhaps worse.

You CAN try a magnetic flow meter that is designed for partial fill. I have not seen many working, although several vendors make them, including ABB, Krohne, Endress+Hauser and others.

I do not recommend an area/velocity meter for this application. They are highly inaccurate devices at the best of times.

Best,

Walt Boyes, Life Fellow, ISA; Fellow InstMC; Chartered M&C Technologist
Spitzer and Boyes LLC

 

Walter, this was helpful!

To answer the question, there's no submerged or surcharged flow in the pipe. 95% of the time that flow exists, the pipe remains partially full and also no reverse flow occurs.

The tricky point or the imposed constraint if you prefer, is that there are times where flow could be shallow enough so that AV sensors can't measure it(practically < 1 inch)and also times where flow stops. So, we are asking ourselves how to handle this?

Many thanks!
Michael Tsakiris

p.s. Interesting that you mention that AV sensors have poor accuracy, does this comes from your experience? I was almost convinced that it would be the most close choice to meeting the requirements, considering installation, partially full flow measurement etch.

 

Area/Velocity sensors have been found to have poor accuracy in many studies. In one, they found that A/V sensors were off by as much as 40-50% of measured flow rate. Marsh-McBirney (now McCrometer) makes one that is radar-based that is the most accurate of the lot of them. It is also the most expensive.

That¹s why I suggest a Palmer-Bowlus with a bubbler flow meter. It is the most accurate, most capable of low flow measurement in wastewater (a 22.5 degree v-notch weir will measure accurately lower, but this is wastewater. It will clog with solids and crud very quickly) and a bubbler is pretty easy to use and maintain.

Since you have a small pipe, you could use the low alarm contact on the bubbler to run a gate valve to stop the flow downstream of the Palmer-Bowlus Flume and open it back up when the height behind the flume is in a readable flow measurement range. Or you can just expect that the PB Flume¹s zero is as close to zero as you can get. This is an active zero, not dependent on the pressure transmitter, which will always be submerged, whether it is a solid state or bubbler transmitter. This is quite different than the Area Velocity meter¹s pressure sensor, which is sitting in the invert of the pipe. There is always fluid behind the measurement one of the flume. If this doesn’t make sense, let me know and we can talk about it.

Yes, my opinion of area/velocity meters is from personal experience. I was Global Sales Manager for Montedoro-Whitney A/V meters (sued to death by Marsh-McBirney) and was one of the development team for the Datagator product that I mentioned in my earlier email.

Walt

Walt Boyes, Life Fellow, ISA; Fellow InstMC
Chartered Measurement and Control Technologist
Spitzer and Boyes LLC

 

Dear Walt (not Walter..i'm sorry),

You seem to know the subject really well!
So, your posts are helpfull to me (by the way i'm an Electrical engineer dealing primarily with control systems).

So, if i use a Palmer-Bowlus Flume as a primary measuring device, i need also a level sensor due to the predetermined level to flow relation in the Flume, isn't that the common solution?
This level to flow relation is specified by themanufacturer, do i guess right?

Or i could use a flowmeter and

"...use the low alarm contact on the bubbler to run a gate valve to stop the flow downstream of the Palmer-Bowlus Flume and open it back up when the height behind the flume is in a readable flow measurement range"

as you say, but this is a little more complicated.

Beyond all these, could you say if it is relatively easy to install a Palmer-Bowlus Flume to a pipe?

Thanks!
Michael Tsakiris

 

You are right. The Palmer-Bowlus Flume is the primary device, but unless you want to stand there with a ruler and a flashlight, you will need a level gauge to measure the head height behind the flume. When you buy a bubbler or ultrasonic or direct pressure transducer flowmeter from, say, ISCO, the electronics have the equation for height to flow conversion built in.

How easy it is to install a Palmer-Bowlus flume in an 8² pipe is somewhat dependent on the physical geometry of the pipe. If it is above ground, you can cut and install a section with the flume mounted inside. The best known manufacturer of the flume is http://www.plasti-fab.com/wastewater-products/palmer-bowlus-flumes. They¹ll have data on how to do this better than I can dredge up from my memory.

Good luck,
Walt

 

Many thanks for the help Walt!
Take care!!

Michael.

 

Hi Mike,

I am the President of Eastech Flow Controls out of Tulsa Oklahoma. Please go to our website; www.eastechflow.com and click on the Accuron Cartridge Meter. The Accuron (permanent version) or the FlowScope (portable version) can measure open channel flows accurately down to 2GPM. Any questions, please feel free to contact me directly via my email or on my cell: 201-406-0099.

Looking forward to hearing from you.

Regards,

Frank Sinclair
Eastech Flow Controls

 

What Frank said. This might be your best bet.

Walt

Walt Boyes, Life Fellow, ISA; Fellow InstMC
Chartered Measurement and Control Technologist
Spitzer and Boyes LLC

**Spitzer and Boyes LLC publishes the Industrial Instrumentation INSIDER (www.iainsider.com)**