moonshineJ
Member
- Location
- USA
Devin, is it be possible that the remaining process fluid within the Coriolis meter flex tubing (while no ACTUAL flow present) would create a false indication of SOME flow?
Problem regulating the Flow Control Valve at 0-2% open position
- Thread starter moonshineJ
- Start date
- Status
StarCat
Industrial Engineering Tech
- Location
- Moab, UT USA
- Occupation
- Imdustrial Engineering Technician - HVACR Electrical and Mechanical Systems
This is a correct premise as seen from one angle. " Hunting " is a known characteristic of an oversized control valve.
Why are you speaking in BAR units?
Devin Hanes
Member
- Location
- United States
yes, if it's not setup correctly it can do that. There may be some applications I'm not familiar with where that's typical but not in my limited experience. If it's a newer micro-motion, some models come with a free trial of smart meter verification you can run once as well, the first thing I would do is look at it compared to the other lines that are working as desired and what I suggested in the previous post. Someone may have mistakenly done a zero calibration during commissioning, should only do zero verification if it's a choice, factory zero is more accurate.
I would also double check there is actually no-flow as thought to be, that the valve isn't leaking by. If there's a local block valve in the line you can close it and see if it changes the flow measurement. Might also be worth comparing low-flow-cutoffs on the meters.
Last edited:
synchro
Senior Member
- Location
- Chicago, IL
- Occupation
- EE
Bubbles trapped within the meter may be causing issues at low flow rates:
RumRunner
Senior Member
- Location
- SCV Ca, USA
- Occupation
- Retired EE
A lot these “precision processes” deploy valves that are “BLOCK and BLEED”.
This approach eliminate the aforementioned scenario.
It gets rid of whatever remains in the system after the valve is closed by bleeding the lines.
Where I used to work (years ago)--dealing with chemical processing, the plant was designed for the piping to handle different chemicals to convey one particular product to different packaging or filling machines.
During the changeover—the system is purged--so no mingling of different chemicals is apparent.
After purging the lines--purified water is run through the system and the wash water is monitored by spectrometer to check for product/bacterial contamination etc.
In case one wonders what Spectrometers are:
From WIKI
******Spectrometers were developed in early studies of physics, astronomy, and chemistry. The capability of spectroscopy to determine chemical composition drove its advancement and continues to be one of its primary uses. ******
Nothing is left to chance and the place was constantly monitored by the FDA.
Do not use control valves for shutoff and vice versa. Getting bubble right seals needs a different valve.
Also just realized you are trying to balance/control two valves. This is alwAys trouble. The typical way of doing this is put them both on control loops. What happens is they oscillate because as one adjusts the other reacts to it but because of process delays it creates an oscillator. There are ways to do it (ratio controllers) but the easiest way is set the process gains of one at about 1/4 to 1/10th of the other. Or use two independent variables like say ratio and total flow. Also ratios get you in trouble because of the math being nonlinear.
There is a book series. Instrument Engineers Handbook by Liptak. Pretty much explains how to properly select valves, how to properly do blend/ratio controllers, all of this. These are all solved problems but you can’t just naively buy a couple control valves and expect it to work.
Also just realized you are trying to balance/control two valves. This is alwAys trouble. The typical way of doing this is put them both on control loops. What happens is they oscillate because as one adjusts the other reacts to it but because of process delays it creates an oscillator. There are ways to do it (ratio controllers) but the easiest way is set the process gains of one at about 1/4 to 1/10th of the other. Or use two independent variables like say ratio and total flow. Also ratios get you in trouble because of the math being nonlinear.
There is a book series. Instrument Engineers Handbook by Liptak. Pretty much explains how to properly select valves, how to properly do blend/ratio controllers, all of this. These are all solved problems but you can’t just naively buy a couple control valves and expect it to work.
moonshineJ
Member
- Location
- USA
Myspark, Paulengr, thank you for the insight! I like the idea of counteracting the oscillation problem with dissimilar gains or using separate reference variables.
- Status