I need flow meter for accurate and consistent measurement of water flow from Cooling tower to Condenser. Application details are given below:

Pipe Size =   60 in
Pipe material =    Carbon Steel
Pressure Class  =    125 psi
Pipe location   =    Underground, manhole is available to install a sensor.

Total Pipe straight run  =     100 ft underground from CW pumps (4 no.) discharge header to condenser.

u/s straight run from mounting point =    58 ft
d/s straight run from mounting point =    42 ft

Mounting orientation = Top of pipe, vertical

Flow Range  =     20,000 to 100,000 gpm,
normal flow is 50000 gpm

Transmitter type =     Loop Powered 24V, HART
Transmitter output signal =     4 to 20 mA (preferably with integral display in gpm)

I know about Verabar, Annubar etc. anybody suggest which is the best. I have only vertical mount choice while Verabar doesn't recommend it.

any other idea.

 

I would also consider; Insertion magmeter, insertion vortex, and insertion turbine meters. All of which have advantages in their own right. You might try VorTek Instruments (www.vortekinst.com) for the Vortex, and InFLOW, INC. (www.inflowinc.com), or EMCO, Flow Technology, Hoffer, etc. for the turbines.

John Catch
InFLOW, INC.
[email protected]

 

If I had my choice, I'd use a spool piece magnetic flowmeter for this application. Put it in and forget it. Accurate, reliable, repeatable under all flow regimes, and it will work until your children are retired.

But, if you are intending to use an insertion meter, I would use an averaging pitot tube such as annubar or verabar or other vendors (there are about 10, if memory serves). The reason for this is that in any 60 inch line from the cooling tower to a condenser that I've ever seen, there is not likely to be anywhere near the requisite amount of straight run to allow an insertion magnetic flowmeter or paddlewheel or insertion turbine meter to properly work. You will need 25 diameters of straight run from the nearest elbow or reducer...that's 125 feet of linear distance upstream, and 10 diameters downstream...another 50 feet downstream. I doubt that you have it, and I doubt that's enough, either.

The reason is that all single point velocity devices (insertion mag, paddlewheel, insertion turbine, pitot tube) measure the velocity at a single point, somewhere around 0.1 diameter inside the pipe wall. Different vendors use different insertion depths, and the absolute insertion depth changes with the pipe diameter, obviously.

When you have a long straight run, and the flow is either laminar or fully developed (Rn>10,000 or Rn<4000 roughly) then Vm=Va (the measured velocity is equal to the average velocity in the pipe. More or less.

When you have any perturbation of the velocity profile at all, this is no longer true.

The ability of the averaging pitot tube meters (and the averaging insertion magmeter by McCrometer/Marsh-McBirney) to essentially recover from those distortions of the velocity profile make them the correct choice here. Averaging pitot tube meters have been calibrated for 0.5% of measured flow rate inserted directly into an elbow, or after a butterfly valve. You cannot do that with a point-velocity device.

Multiple point pitot tubes (or the averaging insertion magmeter mentioned before-- which was designed to mimic the performance of the multiple port pitot tube) will give you the greatest accuracy, but more importantly, the highest repeatability of measurement. They can also be installed through a hot tap or "corporation stop" assembly.

It is a good idea not to use a completely vertical installation. Installing the sensor at the 10 o'clock or 2 o'clock position is preferred, but a straight up and down installation will still work.

The other choice, which necessitates cutting a very large hole in the pipe, is an insertion propeller meter, such as the Water Specialties or McCrometer brands sold by McCrometer division of Hach/Danaher. This will be quite accurate since the response of these meters in large diameter pipe has been thoroughly studied by the US Bureau of Reclamation for irrigation and water wholesaling service.

Good luck,

Walt Boyes
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com
Mailto:[email protected]
Read my blog SoundOFF!! At www.controlglobal.com/soundoff

 

Walt,

Good comments, however I think you'll find lots of disagreement with your 25D straight run requirement. A few years ago while working for one of the insertion turbine meter companies, we spent a rather large amount of money and time on velocity profile, turbulence, swirl, pulsation, and other anomalies of flow in large pipes. We did the test to prove that single point insertion meters were equal to averaging pitot tube meters (the main competition).

We developed an X-Y probe set of insertion turbine meters for pipes up to 10 feet in diameter. The probes were automated with stepper motors and a PC to do pipe traverses and then polar integrate the results to actually come up with the real velocity profile (profile factor).

Our results pointed to the following; if you are going to measure flow in large lines with an insertion style meter (regardless of sensor type) you are better off doing a traverse (maybe X & Y directions), integrating the result, and then determining an actual profile factor. Otherwise, for a relatively good installation (10-15D upstream, 5-10D downstream straight run) industry standard profile factors for single point meters or hole locations for averaging pitot tubes yield about the same accuracy. Since the tests, these results have been corraborated numerous times, both in the field and tests at various NIST labs.

One other result of these tests was the confirmation of the effectiveness of Flow Conditioning Plates. The flow plates are great at stabilizing flow conditions so even if high accuracy can't be obtained, a good repeatable measurement can. I once had a job of spaghetti shaped 24" pipe feeding natural gas into a heat treating furnace. After the averaging pitot folks gave up, we went in with a Mitsubishi type flow plate, and an insertion turbine meter and got good repeatable flow measurement for the control system.

I wasn't sure what you were getting at with the laminar flow comment. In this application, even at 0.25 Ft/sec the pipe Reynolds number is ~115000. The application also had about 20D straight run to work with.

If I had my choice I would put in a Venturi or ASME flow nozzle. Of course this is expensive, but the results can be verified in the field and these units will last until the next ice age. In lieu of Venturis or Nozzles, an insertion meter especially with a flow plate will likely work very nicely in this application, and is a good economical choice.

John Catch
InFLOW, INC.
[email protected]

 

My choice would be a prop. meter from Water Specialties, these are quite inexpensive and easy to retrofit into an existing line. An insertion mag should work but I have had several bad experiences with these.

If you use an averaging pitot tube you will need to figure out how to get rid of entrained air which will migrate up the impulse tubing and give you a zero offset(I have a modified design which takes care of that).

I agree with Walt, a spool type mag meter would be the best solution but I assume you want something you can install into an existing manhole.

Regards
Roy

 

I'd like to see your data, John. My research (and I've helped design and test insertion mags, insertion ultrasonics, propeller, paddlewheel and insertion turbine flowmeters) indicates that what I said is conservative and accurate.

I don't have a problem using a venturi or a flow tube.

Except...most cooling water applications I've ever seen have really bad (for flow measurement) piping layouts with lots of elbows and turns in three dimensions, which is known to cause spiraling of the flow.

The reason I suggested a magnetic flowmeter is that, absent really bad spiraling, it handles very poor flow conditions upstream better than any meter in the size range (you can't make a swirlmeter that big...or at least nobody does).

Any insertion flow meter is a bundle of compromises, where a full pipe flow meter is less so.

Walt Boyes
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com

Mailto:[email protected]
Read my blog SoundOFF!! At www.controlglobal.com/soundoff

 

We think a lot alike. Now that the second half of my post has appeared, you can see that I also suggested a weld-in Water Specialties or McCrometer (same vendor, different trade name) meter.

Walt Boyes
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com

Mailto:[email protected]
Read my blog SoundOFF!! At www.controlglobal.com/soundoff

 

Thank you Walt and John for your nice suggestions and comments.

I want to tell you that my predecessor bought and installed TX201B/FT420 (mfr: Seametrics) at this application. Didn't work at all, gave extremely erroneous measurements right after installation, high fluctuations and failed completely within just 2 months, maybe some bearing problems.

I took it out, serviced and reinstalled but didn't get stable readings. so, I am absolutely against any kind of turbine meter here. I think they all got some bearing problems (you know this one is not a very clean water application, source is underground wells but with sandy environment). I also tried to raise and lower the insertion length, measurement varied drastically. definitely I don't know where I'll get the average velocity point so it is just I think, guess work. I am also not sure any average velocity point retains the location.

I checked McCrometer website, Ultramag and propeller types. Former is not suitable for my line size. if there is one, it will be I think very costly. latter is not easy to install and again is a rotary type which I don't like.

so I am not yet convinced to have any other choice besides the averaging pitot tube. so I wanted to know which is better annubar, verabar or any other. I am also restricted at present to the vertical mount position. if it works, I can schedule a regular PM to get the entrained air out. Please share your thoughts on performance & reliability of annubar, verabar or any other averaging pitot tube meter.

Regards,
James

 

I have to confess that I helped to design and test and sell that Seametrics flowmeter when I was Director of Marketing and Sales for Seametrics many years ago, and sand and grit will be a definite application limiter. It is/was intended to be used in clean fluids only. The advantage it has over point magnetic flow meters is that it has a 2 inch diameter rotor, which increases the "sweet spot" it measures the point velocity in. If you have that much sand, it won't ever work right.

There are two components to selection of a multiport pitot tube flow meter--the pitot tube itself, and the transmitter.

Emerson's Annubar and Veris' Verabar, and Racine Federated's Preso designs are well known to be excellent. There are others, like Mid-West Instruments. Get AND CHECK customer references on similar applications to yours.

Make sure the transmitter to be supplied is one you want and can live with. Annubars are commonly supplied with Rosemount (duh!) transmitters. Others can often be supplied with the transmitter of your choice. If you like the new ABB dP transmitter better, or Yokogawa's or whoever's, you ought to be able to get that with the pitot tube.

I hope this helps.

Walt Boyes, ISA Fellow
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com

Mailto:[email protected]
Read my blog SoundOFF!! At www.controlglobal.com/soundoff

 

As I said earlier the problem with pitot tube mounted from the top of pipe is entrained air will enter one or the other side and give you a false DP reading.

I had to replace a couple of insertion mags once in a hurry. It was a similar situation so we built a pitot tube in the shop with impulse tubes that went down to the tip before coming back up, that solved the air entrainment problem. We were able to calibrate the transmitters against change in tank level, to the best of my knowledge they are still giving good service 15 years later.

I still think one of the large diameter saddle style prop meters from Water Specialties would be a good fit.

Roy

 

...and they are fairly cheap, too.

I agree.

Walt Boyes
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com

Mailto:[email protected]
Read my blog SoundOFF!! At www.controlglobal.com/soundoff

 

I wanted to add just a couple comments regarding your apparent problems. Please don't think that all turbine meters are created the same. There are a number of different bearing styles and materials available. Rolling type bearings such as ball or close fit sleeve typically only work great in clean fluids. Pivot and shielded styles can definitely work in less than ideal fluids. Carbide works well in some applications, while man-made Sapphire is better in others, etc. In addition, the actual turbine hydraulic design has a major impact on performance.

The EMCO turbine meter has a good reputation for working in steam even with moderate carryover (dirty liquid). EMCO (and InFLOW) have been making turbines that work well (with routine maintenance) in Blast Furnace Gas. This stuff plugs averaging pitot tubes quickly and coats thermal meters. We routinely use turbines in steel mill water with little problems other than the occasional cleaning for string, and other debris. So apparently, your existing turbine meter is just a design that is not well suited for this particular application. More than likely it would work fine in other applications.

You mentioned that by raising and lowering the probe the results varied drastically. This is not what one would expect if you have good flow profile, and no swirl in such a large pipe. Typically on a large pipe you should only see mild changes for let's say +/-1/4" insertion change if you are near the average flow point (generally about the ID/8 position). More than this suggests abnormal flow conditions probably exist. That said, I wouldn't agree that an averaging pitot tube is going to give the correct result since it has fixed pick up points which may or may not be in the correct locations. As has been mentioned by others, a more full bore type meter may be more appropriate, or at least add a flow conditioning plate.

If you want to stay with the averaging pitot tube meter, my choice is the Verabar. The engineers that came up with that design all had years of previous averaging pitot experience to work from. The roughened upstream surface allows for better low flow performance (more stable signal at low DP values). I also know that a great deal of flow calibration work was done including that in large pipes.

One other comment on the averaging pitot tube meters; make sure they are installed correctly as the fluidic forces on such a large unit are quite significant.

John Catch
InFLOW, INC.
[email protected]

 

I just checked from the Plant Performance Test carried out in 1998, our condenser is designed for cooling water flow of 75000 gpm, while max flow is 82,500 gpm. So I correct myself here. I am not interested below 45000 gpm, accept my apologies for the max and min flows given in my original post.

I am also in contact with Veris separately, Veris sales recommends a partial insert sensor at this application. the full insert may not survive because of high velocity perhaps beyond the strength of sensor.

Veris has quoted V400SPI-15-H-R-B8CF15-C2NS-XX. They also suggested installing two pitot tubes at 45 degrees for better results.

I am still in doubt and looking forward at more opinions.

 

Hi James,

Have you thought of Ultrasonic Flowmeters as they are very accurate at these velocities 0.85m/s to 3,6ms as per your figures. the cost would be a big saver as well. In the UK certain meters have been approved for MCERTS which are approved to Environment Agency standards and most of the big power stations have fitted these (cost being the main criteria). If you need any further info please contact me and we can discuss it further

Regards
Kenny McEwan

 

VERABAR Partial design is a great design over very high velocity applications. Its not that you can't install it from the top & if its is completely filled pipe I can assure you partial works perfectly though there could be air entrapment issues which can be easily avoided by practicing recommended instrument tubing installation procedure. The advice to install at 45 Deg. C is to avoid air entrapment issues. Further note that few technologies being discussed (Propeller, Turbine etc) will have problems with dirt in the application whereas VERABAR's bullet shape has highest resistance to dirt.

I have recommended partial design sensor to customers on similar application & they are very happy to use it.

I see Full Bore Electromagnetic as an alternative choice but if you are looking for reliability at cheap price VERABAR is a solution. Also look at the 5-Year performance warranty on VERABAR.

 

I'm sure that Rohit means to say that you should look at several different manufacturers of multiple port pitot tube insertion flow meters, not just his. There are several others, including products from Rosemount (Annubar, the original), Midwest Instrument, Racine Federated (Preso) and others. I'm not saying that Verabar isn't a high quality instrument, because it certainly is-- just want to give the end user the most complete choices possible.

Walt Boyes
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com

Mailto:[email protected]
Read my blog SoundOFF!! At www.controlglobal.com/soundoff

 

As there was a specific mention of Veris VERABAR, I further clarified on the same but as rightly said by Mr. Walt there are other manufacturers. Partial Sensor is something which is unique to Veris.

 

non contact type flow meter is better .