Very high turndown flow meter

[mrsteam]

I have an application where I need a flowmeter capable of a turndown ratio of almost 70 (0.3 L/min to 20 L/min)... So far I have not found a single flow meter that would do the trick. Also, this meter will be measuring a controlled variable on a PID control loop, so decent accuracy is needed. Any suggestions? Thank you very much.

 

[jsummerfield]

What line size, fluid state, flow rate, AND ACCURACY, etc. Some Coriolis meters, turbine meters and other technologies can meet that. Perhaps an orifice plate and smart transmitter can exceed 10% full scale accuracy with 70/1 turndown. However, you are likely looking for 0.5% full scale with some low percentage of the actual flow.

John

 

[mrsteam]

Sorry, I guess I didn't give enought details.

1.5" line size, liquid flow, 0.3 L/min to 20 L/min variable flow rate depending on application, accuracy as good as we can get (ideally not more than 0.5% full scale). DP meters are out of the question, unless they can GUARANTEE accuracy over the ENTIRE range.

Thank you very much for your comments.

 

[FrancisL]

Basically it is not feasible with a regular flow meter.
Dual Transmitters across an orifice (or maybe a single smart DP being recalibrated on the fly) might approach it, but at the lowest flow rates even 0.5% of full scale is 35% of actual flow.
But do you really need that turndown? Exactly what is the process?

 

[schlebda]

Is it a conductive liquid, like water, or is it oil? Did you look at magmeters?

 

[mrsteam]

OK, a little more info on accuracy needed... We could actually go to 0.5% of measured value at the high end (20 L/min), and could go up to 1.5% of measured value at the lower end (0.3 L/min)...

I have been suggested a coriolis flowmeter... As a matter of fact, a special model could give 0.8% of measured value at the low end, but it's expensive.

I don't know if a mag meter could provide such a turndown ratio? Besides, I do not know if chemical additives would be conductive...

Turndown is needed because we are trying to design a "one size fits all" unit here... Different chemicals being added to another fluid flowing at various different rates.

Once again, thank you much for the valuable input.

 

[mrsteam]

OK, I am told that viscosity of the fluid is a big factor for sizing and spec, at least for coriolis meters. Would this be the case for all types of flowmeters? Or could we find one for which viscosity would not be much of a problem?

 

[CMfgE1]

You didnt mention the material type. No matter though. A coriolis meter will be fine. Magmeters are limited by the conductivity of the material, A Differential transmitter (DP) would also work in this application. Remember, your flow rate is only as accurate as the tuning on the valve controlling the rate. The meter signal will be picked up by your control system and an output will be sent to the valve. So the whole system will only be as good as the valve anyways. What is the material and what is the application?

KB

Quote: "Its not what you know, its who you know"
Everythings a learning experience-Everything

 

[FrancisL]

CMfgE1
Sorry but you are wrong about the valve. The valve will not be the determining factor for accuracy, the flow meter will. The valve is just part of the controlled process.
The valve characteristics however will be a factor - the valve should have a high turndown too.
And yes,it looks suitable for a coriolis meter.
Francis

 

[mrsteam]

Thank you all for your valuable input. And once again I apologize for the lack of info: the controlled device is not a valve, but actually a progressive cavity pump with a variable speed hydraulic motor... OK, I guess the controlled device will be a valve, in this case controlling the hydraulic fluid going into the hydraulic motor which in turn controlls the speed of the pump, and thus the flowrate. So far I have been told that the worst-case-scenario viscosity will be similar to that of whole milk.

I hope this helps, and yes, I believe the best choice here would be the coriolis... I was just hoping someone would come up with a less expensive option.

 

[mrsteam]

Regarding the DP transmitter, in order to work with such a high turndown ratio it would have to be used with some kind of variable area flowmeter, wouldn't it? An orifice plate wouldn't cut it.

The problem with this is that you require some amount of pipelength before and after, and space is a major constraint in our design.

 

[JimCasey]

If you're careful with the sizing of a positive displacement flowmeter you can do this. Something like an oval-gear meter with a frequency output would be worth considering. Oval gear meters are very accurate but are subject to damage from overspeeding, so you want to make very sure there is no way to get an air bubble into one. An expanding air bubble can over-rev the OGM and then you get to replace the bearings. Aside from that they only get more accurate as viscosity increases.

 

[hydrae]

Why do you need to meter the material twice, the progressive cavity pump moves X amount of material each revolution just count the rpm and verify the pressure relief valve does not open
Hydrae

 

[jmw]

The problem is slip flow.
In a pump the higher the pressure drop across it the greater the slip flow i.e. the fluid that slips through the working tolerances.
In a PD meter the pressure drop across the meter is usually very low so the slip flow will be much less and can be calibrated out (especially if you have a high resolution pulse output...)
But a PD meter could be a very good choice at this line size.
However, now we need to know the pressure.
If this is high pressure then you may need to select a twin case PD meter. For Coriolis meters high pressure caused more stiffness in the tubes and pulsating pressure, such as you can get from some pumps, can cause even more problems. You need to talk to the mass meter manufacturers to discover the current state of play. Of course, if the pressure is low, then you are fine.

JMW

http://www.ViscoAnalyser.com

 

[djs]

You would need some kind of a switch, but another possibility is multiple DP transmitters and an orfice or venturi. We once put in an application with a venturi and three DP transmitters for extended range.

 

[jsummerfield]

What is the viscosity? For most flow measurement the viscosity is not a significant factor. Use the Coriolis meter calculation software and enter the details for a tpical fluid. Multiply the viscosity by 10 and run the same calculation. The same Coriolis meter applies to the rate for low or higher viscosities. The software predicts the accuracy and pressure drop. Expect slightly higher pressure drop with the increased viscosity. As you move toward highly viscous or non-newtonian fluids - say peanut butter or concrete, everything gets problematic with flow measurement. The difference between water and crude oil are measureable in terms of pressure drop and not functionality. However, any vapor or flashing in the fluid is an issue for Coriolis and many other flow meters.

John

 

[ScottyUK]

Hi John,

What do you make of Foxboro's claim that their latest Coriolis meters handle mixed-phase flow without loss of accuracy? Some real-world experience is useful from time to time - cuts through the salesworm's spiel and reflects the truth instead. Micromotion seemed pretty adamant that their instrument wouldn't handle it.


----------------------------------

One day my ship will come in.
But with my luck, I'll be at the airport!