JackR,
The Emerson 7835/45EGA variants of the tube density meters solve the problem of entrained air by operating in a different harmonic mode where the VOS effects are considered much less.
The trouble with going to low frequencies is interference; i.e. sensitivity to pipe born noise and to crosscoupling with other mass meters.
Originally operating at low frequencies, coriolis meters needed to be detuned if there were to be operated within close proximity.
In general the ideal is to operate at higher frequencies because most high frequency pipe borne noise is very quickly attenuated close to the source and doesn't propagate to the sensor with any significant amplitude.
So dealing with entrained air means balancing conflicting requirements if the only approach is to operate at low frequencies.
The standard tube density meters operating at 800-1200Hz and could operate next to each other and not interfere with each other.
It also takes very high amplitude noise to interfere with them or unsettle them.
Air creates a number of problems.
Problems of lost lock and having to hunt again for the resonant frequency have been largely overcome by better amplifiers, PLL drive and more power. Some design changes also have helped the 7835 and 7845 be better sensors than the 7830 and 7840 sensors (as a part of the development of these into the MassMaster series of single straight tube coriolis meters (now discontinued, wrongly in my view).
One mechanism is where the air bubbles migrate to the tube walls so when the tube walls vibrate the tube moves and the air flows around the liquid column so the liquid doesn't get displaced.
But to what extent can they manage entrained air and not suffer serious loss of accuracy?
How "organised" or chaotic can the entrained air be?
Is it the mass flow that is least affected or the density?
What about air pockets?
Hswang2 mentions start and end conditions.
It is common to run dry to dry where the lines may be charged with air initially and they flooded with liquid. Most usually, with low viscosity fluids, one can use air eliminators to void most of this before a meter. Or you can detect the air and inhibit registration on meters where false registration can occur.
The question I have about twin tube coriolis is if you have high volume air pockets, is there a risk of unbalancing the tubes if the air preferentially flows through one tube stalling liquid in the other tube?
What about low viscosity fluids?
Bent tube designs will cause a degree of conglomeration of the bubbles together on the inside of the bends...
Air bubbles are much more manageable in high viscosity fluids because they tend not to be so mobile.
It was
my supposition that the entrained gas coriolis were initially developed for crude oil well head metering and have only subsequently been marketed for heavy fuel oils.
This is latter something I don't fully understand (marketing wise) as I don't see this as the best solution.
Entrained gas in well head metering is an unavoidable integral part of the fluid stream and so some form of multi-phase metering is essential, especially as you will want to totalise both phases.
In HFO the vapour phase flow is air which does not need to be accounted for but it is not an integral or desirable part of the flow stream.
Any sensor that responds dramatically to entrained air can be used to detect and inhibit flow when excessive air is present.
If suppliers can't supply fuel when it has excessive air they will stop trying and revert to supplying fuel which is air free.
HFO from the refinery or terminal is usually bubble free. The reason air entrainment is an issue is because the industry persists with old fashioned methods such as tank dipping with tapes.
Thus it is an easy fraud to blow air into the fuel which, because it is black and viscous, is not often evident as having large volumes of entrained air pumped into the fuel and the air will tend to stay in the fuel for a very long time.
The operator then converts the volume (determined from tank tables) to mass using the density the original fuel supplier quotes.
In these applications the cost of mass flow meters is significant because the meters have to be oversized because of their headloss and the pumps have to be upgraded simply to stand to be able to deliver the same rates as without mass meters.
However, the industry requires higher lifts, larger stems and faster turnarounds, all of which imposes even greater headloss burden and there is a move to even more viscous fuels. This all makes the Coriolis meters even less suitable.
The pressing need to minimise headloss will, I think, see coriolis as a short lived solution and the other solutions will come to the fore.
JMW
