THE EFFECTS OF AERATION ON
CORIOLIS MASS FLOW METERS
Guidelines for Installing Meters to Avoid Problems Related to Air Entrainment
(and how to diagnose and resolve problems if they do occur)
Table of Contents
Topic
Page
1. Purpose .............................................................................................. 2
2. Definitions ........................................................................................... 2
3.
4.
5.
6.
7.
8.
9.
Background ......................................................................................... 3
Meter Sizing and Selection ................................................................. 6
Aeration Testing .................................................................................. 8
System Design to Reduce the Effects of Air Entrainment .................. 9
Piping and Process Design to Minimize Air Entrainment .................. 11
Diagnostic / Troubleshooting Techniques ......................................... 16
Micro Motion Transmitter Parameters ............................................... 18
10. Meter Zeroing.................................................................................... 18
11. Recommendations ............................................................................ 19
J. Reizner
Page 1 of 19
Date: 4/16/2001
Acknowledgements
I would like to acknowledge the input and critique that was so generously given to me by Bob
Catlin of P&G. I would also like to thank Micro Motion and their people who contributed many
comments and ideas to this document.
Purpose
The purpose of this document is to explain the effects of aeration of liquids on Coriolis mass flow
meters and how to deal with these issues. This document will provide information regarding:
1. The affect of entrained air on Coriolis flow meters
2. Concepts to minimize the effects of entrained air on Coriolis meters:
a. Meter sizing and selection
b. System design
c. Meter installation
3. How to recognize the symptoms of entrained air on existing systems, and how to resolve
the problems when they exist.
Similar issues may be experienced when powders are entrained in liquids that are metered with
Coriolis technology.
In this document the terms air entrainment and gas entrainment are used interchangeably.
Definitions
Drive gain:
A relative indication of the amount of power the transmitter is delivering to drive
the sensor. When drive gain is saturated the transmitter is delivering the
maximum amount of power it has available.
Amplitude:
An AC voltage measured at one of the sensors pickoffs (PO), directly
proportional to the sensor tube displacement. Amplitude is often referred to as
PO amplitude; furthermore, there are two pickoffs on Micro Motion Coriolis
meters, referred to as left-pickoff (LPO) and right-pickoff (RPO). The LPO and
RPO have the same amplitude.
J. Reizner
Page 2 of 19
Date: 4/16/2001
Background
For a background on the operating principals of Coriolis mass flow meters refer to the Micro
TM
Motion software program Tutor . This program is available from your local Fisher-Rosemount /
Micro Motion representative, or from the Micro Motion web site at: www.micromotion.com
To generate accurate flow and density signals the Coriolis meter has a drive coil that vibrates the
flow tube to a specified displacement (see Figure 1). This displacement is large enough to
produce high-resolution signals at the sensors pickoffs. The transmitter provides the power
required to produce this vibration. Direct feedback from the pickoffs is used to automatically
adjust the amount of power that the transmitter provides and to maintain the proper tube
displacement. Thus if the transmitter detects the PO amplitude decreasing, it will increase the
drive gain and deliver more power to the sensor, allowing the sensor to drive to the correct
displacement.
Figure 1
(Copyright Micro Motion, Inc.)
Under nominal conditions the sensors material properties, tube geometry and size determine the
power that is required to drive it to the specified displacement. In general, a larger sensor
requires more power than a smaller one, and a straight tube sensor requires more power than a
similarly sized curved tube sensor.
The amount of energy required by the sensor to maintain nominal vibration displacement varies
from sensor size to sensor size. The primary reason for this difference in required energy is that
the force required to deflect the tubes varies with the stiffness of the flow tube (and the stiffness
of the flow tube varies with size). In general, the energy required to drive the Coriolis meter
increases with the size of the Coriolis meter, e.g. a 2 inch Micro Motion ELITE meter requires
more energy than a 1 inch Micro Motion ELITE meter. Another general statement is that a
straight-tube Coriolis meter is stiffer (and requires more power) than a like size bent-tube Coriolis,
e.g. a 1 inch straight-tube meter requires more power than a 1 inch bent-tube meter. However, it
should also be noted that the power required by a straight tube is not always greater than that
required by a bent tube, i.e. smaller straight-tube meters can require less power than larger benttube meters.
J. Reizner
Page 3 of 19
Date: 4/16/2001
For various reasons (including the laws of physics, concern for tube and weld integrity and
intrinsic safety power limitations), the amount of power that a transmitter has available to drive a
sensor is limited. Figure 2 depicts the relationship between the energy required by a sensor and
the energy available from the transmitter to maintain nominal vibration displacement yielding an
accurate measurement in non-aerated water. Defined by this figure is the term energy overhead.
Energy overhead is typically expressed as the ratio of the maximum amount of energy supplied
by the transmitter to the energy required by the sensor in non-aerated water. It can be seen in
this figure that the energy overhead varies from sensor design to sensor design (by size and
sensor shape, metallurgy, etc.).
Figure 2
Energy Level
Maximum amount of energy
the transmitter can deliver
Energy Overhead
Increases with
meter size
Energy Required for accurate
measurement in non-Aerated Water
Increases for
straight tube
Time
When a fluid is aerated the entrained air (or gas) bubbles absorb some of the vibrational energy
from the sensors tubes that makes it more difficult to drive the sensor. This requires the
transmitter to provide more power to maintain the proper tube displacement, but the transmitter
has only a limited amount of power available. A relatively small percent volume of air entrained in
the fluid will cause the drive gain to become saturated. When the sensor requires more power
than the transmitter can deliver the tube amplitude will decrease below the desired value. This
can result in less accurate flow and density measurements. Since the percent volume flow of the
entrained air is likely to be erratic compared to the overall fluid flow, entrained air can also cause
unstable flow and density measurements.
J. Reizner
Page 4 of 19
Date: 4/16/2001
Figure 3 depicts the relationship between the energy required by a sensor monitoring an aerated
fluid and the energy available from the transmitter.
Figure 3
Energy Level
Energy Required for accurate for Accurate
Measurement in Highly Aerated Liquid
Energy Shortfall
Maximum amount of energy
delivered by the transmitter
Time
Accuracy and performance of a Coriolis meter will degrade as the volumetric quantity of entrained
air increases. For example, testing performed by Micro Motion using their ELITE meters with
water show that drive gain becomes saturated with the introduction less than 2% (by volume) of
entrained gas. However, as long as the pickoff amplitude (LPO or RPO) remains greater than 50
mVrms, an accuracy of much better than 1% can generally be attained. Once the pickoff
amplitude drops below 50 mV the measurement is considered to be unreliable. As with drive gain,
both pickoff amplitudes are available as diagnostic information from the transmitter. This
information can be accessed digitally with a Rosemount 275 Handheld Communicator or ProLink.
The specific levels of aeration that cause these problems are unknown. The aeration problem
appears to be exacerbated by viscous fluids, i.e., the same level of entrained air in a viscous fluid
appears to create more dampening effect than in a thinner fluid. The size of the bubbles also
appears to affect the meter performance. Very small bubbles may not effect the operation
whereas larger bubbles may. Another factor that influences the amount of aeration tolerated by
the sensor is the level of backpressure on the meter. Increasing backpressure tends to decrease
bubble size or, in some cases, eliminate the bubbles by keeping the air in solution. Thus, in
general, a sensor will be more tolerant of entrained air as the meter backpressure increases.
J. Reizner
Page 5 of 19
Date: 4/16/2001
Experienced-based rules of thumb for levels of entrained air are:
5% entrained gas by volume will generally be a problem
Recommend keeping entrained gas below 3% by volume
Micro Motion has laboratory test data on entrained air in water. Since their flow lab is limited to
centrifugal pumps, they have no laboratory test data on the effects of entrained air on viscous
fluids. MMI is in the initial stages of designing and building a flow lab to obtain data on the effects
of aeration on viscous fluids. MMIs limited data on entrained gas in viscous fluids is from
customer installations.
Meter Sizing and Selection
High flowing velocities through the meter are generally advantageous to the free movement of
gas through the meter. A rule of thumb is that meter velocities should be kept above 3 feet per
second to aide gas removal in thin water-like fluids, and above 1 foot per second in higher
viscosity fluids (lower flow rates are required with higher viscosity fluids to keep meter pressure
drops reasonable. Generally these lower flow rates are sufficient to keep entrained gas flowing in
the more viscous products). Low velocities may more readily allow air to become trapped in the
meter, preventing successful meter operation. This must be balanced vs. meter pressure drop
considerations.
As noted previously, smaller meters require less power to drive the tubes. Given that the
available power from the transmitter is limited, smaller meters have more drive overhead.
Therefore, everything else being the same, smaller meters are a better choice for entrained gas
applications.
As noted above, the Micro Motion ELITE meter has the largest amount of headroom, and
therefore is generally the preferred choice for aerated fluids. Within the ELITE line there are two
styles: the Omega shape (Figure 4) which is used through 1 inch size, and the U shape
(Figure 5) which is used in sizes above 1 inch.
J. Reizner
Page 6 of 19
Date: 4/16/2001
Figure 4
Micro Motion ELITE Omega-Style Meters
(Copyright Micro Motion, Inc.)
J. Reizner
Page 7 of 19
Date: 4/16/2001
Figure 5
Micro Motion ELITE U-Style Meters
(Copyright Micro Motion, Inc.)
The U-style is easier to mount in a way to prevent air pockets from forming than is the Omega
style. Therefore the U-style is preferred for this reason but since the U-styles are larger they will
result in lower fluid velocities for a desired mass flow rate. Where such tradeoffs are presented, it
is suggested that the user select the smallest meter that meets the pressure drop requirements.
Additionally, if it is suspected the fluid aeration will be present in the process the user should
consider ways to increase the back-pressure on the sensor - e.g. sensor location (low-point in the
process), down stream valves and constrictions, etc.
Aeration Testing
Fluid aeration levels can be tested simply. First, extract a small sample of fluid from the process
at the point where the meter is currently installed or will be installed. Place approximately 200 mL
into an appropriate container.
(Note: There are many potential safety issues associated with obtaining a sample, including but
not limited to: high pressure, high / low temperature, hazardous fluid properties such as corrosive,
etc. The person obtaining and handling the sample must learn and follow all pertinent safety
standards.)
Put about 100 mL of this sample into a graduated cylinder. Record the time, along with the
starting sample volume and fluid temperature. Let the sample sit for a period of time sufficient to
insure that all entrained gas is removed. For viscous samples, this could be as long as 48 hours.
Higher temperatures generally reduce viscosity and speed up removal of entrained gas (but
remember that heating up the sample may be hazardous). After the sample has sat for a length
of time sufficient in time to remove all entrained air, record the ending volume of the sample.
J. Reizner
Page 8 of 19
Date: 4/16/2001
This sample volume must be temperature corrected, which may be done in one of two ways.
The easiest way is to record the ending volume, then using density vs. specific gravity
charts for the fluid make the temperature correction.
If such density correction charts are not available, temperature correction may be done
by heating the fluid sample so that it approximates the fluid starting temperature. Once
again, remember to consider any potential safety hazards associated with heating the
fluid.
Divide the temperature corrected ending volume by the temperature corrected starting volume.
Subtract this number from 1, and express this number as a percent. This is the percent of
entrained air in the sample.
System Design to Reduce the Adverse Effects of Air Entrainment
In systems where entrained gas is inevitable, several methods exist to minimize the potential
effects. Increasing the amount of back-pressure on the sensor will reduce the gas volume and
possibly keep the gas in solution. This higher backpressure may be accomplished with a
restriction device such as a pipe reducer, valve, orifice plate or equal vent device downstream of
the flow meter.
The preferred solution is to prevent air from entering the system in the first place. The following
offers details on good system design practices to prevent the introduction of air:
1. Supply Piping into Tanks
Generally the supply piping bringing fluids into tanks enters the tank from the top. In large
tanks the fluid can fall significant distances to the fluid level in the tank. This can cause a
large amount of air to be entrained into the fluid. Solutions to this problem include placing
the entrance pipe below the normal tank level, continuing the entrance pipe inside the
tank to a point below the tank low level, or running the tank at higher levels to make the
drop less severe. Make sure that the material entrance is kept as far away from the outlet
piping as possible to minimize the possibility of air-entrained product being sucked
directly into the pump. Do not direct the inlet pipe discharge towards the pump suction.
To do so could result in cavitation, water hammer or other similar problems.
2. Tank Agitators
When tank agitators are operated with the agitator blades at the fluid level the blades will
entrain substantial quantities of air. A possible solution to this involves electrically
interlocking the tank agitator to the tank level so that the agitator does not run when its
blades are at the fluid-air interface.
J. Reizner
Page 9 of 19
Date: 4/16/2001
3. Tank Discharge Piping to Pump Suction
a. Dont pump when the fluid level is below the top of the pump suction line. To do
so will suck substantial quantities of air into the fluid. Interlocking with a tank
level-measuring instrument can automatically perform this function.
b. Insure that the pump suction line connection at the tank is located high enough
above the tank bottom so that tank sediment is not sucked into the fluid and
passed through the system.
c. Insure that there are no high spots in the pump suction line for air to accumulate.
If there are, this can cause slugging of the flow that will adversely affect meter
performance.
d. Do not put filters in the pump suction line. They belong in the pump discharge
line. An appropriately sized strainer may be located in the pump suction line to
remove large objects. Strainers are often required in the suction line to various
types of positive displacement pumps to prevent debris from mechanically
damaging the close mechanical passages of a positive displacement pump.
e. Make sure that pump suction piping is tight and leak-free. Since this suction line
may be under vacuum when the pump is operating, an ultrasonic vacuum leak
detector may be required to find these leaks. A vacuum leak on the pump suction
can suck in large amounts of air, and is generally non-observable.
4. Air Removal Devices
Devices are available to remove entrained air from fluids. The simplest device to
remove air is a tank of relatively large size (to give a long residence time for the
air bubbles to rise to the top). The tank can be MILDLY agitated to facilitate the
air bubbles coming to the surface. The tank should have as large a surface area
as possible and not be very deep. This will make it easier to remove the air
bubbles. The tank should be put under vacuum if this is possible and if this can
be done safely. The vacuum will speed up the process of air bubble removal.
Custom-built devices are available from various companies to remove entrained
gases from fluids. These devices are generally called deareators. These custommade devices are generally used for more viscous products and where higher
levels of air removal are required. Basically these devices cause the product to
fall on a rotating plate, where the product is wiped into a thin film by centrifugal
force and a squeegee. The device is under a vacuum. The large surface area,
thin surface film and high vacuum result in high amounts of deaeration for
viscous products.
5. Operating Parameters
Generally raising the system operating temperature will decrease the viscosity of the
fluid, and increase the ability of entrained air to be removed in any given time from the
fluid.
J. Reizner
Page 10 of 19
Date: 4/16/2001
Piping and Process Design to Minimize Air Entrainment
1. Pressure Piping
a. Dont create high spots in the piping system that can trap gases.
b. Slope horizontal lines up at a slight angle. The exact slope is a function of the
viscosity of the fluid. Thin fluids may only require a 2-degree slope; thick fluids
may require a 10-degree slope. This may be difficult for construction crews to
accomplish, so discuss this with them. The following sketch shows an easy way
to accomplish this.
Copyright Krohne, Inc.
c.
Use reducers correctly so that they do not trap air. Use concentric or eccentric
reducers as appropriate.
Concentric Reducer
Eccentric Reducer (shown with flat up)
When using eccentric reducers, install the flat so that air can most easily travel
through the system. The exact orientation of the flat of an eccentric reducer is a
function of the piping orientation, but should be obvious when one thinks about
minimizing points where air can be trapped.
J. Reizner
Page 11 of 19
Date: 4/16/2001
2. Meter Installation to Minimize the Effects of Entrained Gases
Accurate measurement with any type of flow meter requires that the meter be completely
full of process fluid at all times in order to provide an accurate flow measurement.
Therefore the following meter locations must be avoided:
a. Installation at high points in the piping system (this allows air accumulation)
b. Installation in a downward flowing line, especially where the line discharges
immediately downstream of the meter. The lack of backpressure on the line will
cause slug flow in the meter.
a. a
.
b.
Copyright Krohne, Inc.
Downward Piping
When the piping downstream of the meter is in a downward direction, follow the sketch
below. Install a goose neck immediately downstream of the meter, with a siphon air vent
valve located at the high point in the piping to break the vacuum. This installation will
prevent partial vacuum from occurring in the line.
Syphon
Copyright Krohne, Inc.
Open Discharge
Install the flow meter at a low point in the line for open discharge piping systems to insure
that the meter is always full of fluid.
Copyright Krohne, Inc.
Open Discharge
Control Valves
Always install control valves and shutoff valves downstream of any flow meter to prevent
vacuum from developing in the flow meter.
Copyright Krohne, Inc.
Pumps
Never install a flow meter on the suction side of a pump.
Copyright Krohne, Inc.
J. Reizner
Page 12 of 19
Date: 4/16/2001
Preferred Installation
a. ELITE U-Style
Concentric Reducers
(if required)
Flow
Flow
b. ELITE Omega-Style
ELITE OmegaStyle will allow air
to be trapped in
pocket
J. Reizner
Page 13 of 19
Date: 4/16/2001
2. T-Meter
Concentric Reducers
(if required)
Flow
J. Reizner
Page 14 of 19
Date: 4/16/2001
Alternate Orientation
a. U-Style and Omega-Style ELITE
Flow
Eccentric Reducers, flat
up on the inlet, flat down
on the outlet
(where required)
b. T-Meter
Flow
J. Reizner
Slope away from inlet
for air removal
Page 15 of 19
Date: 4/16/2001
Diagnostic/Troubleshooting Techniques
Due to the complexity of the variable interactions, currently-available Coriolis mass flow meters
cannot automatically diagnose an air entrainment problem. Fortunately the transmitters do
provide enough information for someone who understands the operation of the sensor and the
process to properly make that diagnosis.
Air entrained in a system generally manifests itself in one of two ways:
As a semi-steady stream of bubbles
As a large pocket of air that collects in a system high point. This will occasionally allow a
large bubble to burp through the sensor, typically called slug flow. When this occurs, the
transmitter will report a Slug Flow alarm and the transmitters outputs will go to fault
conditions until the condition clears, usually within a few seconds.
We will concentrate on the first item, the stream of bubbles.
Some key points:
Entrained air does not always result in a foamy, frothy fluid. While this is possible, in the
majority of cases no characteristic foam is present in the process fluid.
Bubble size does matter but no one is sure to what degree because lab testing has not
been done on this subject. It appears that larger bubbles may be worse than smaller
bubbles. Having said that, microscopic bubbles can also be very detrimental to sensor
performance. Smaller bubbles are more difficult to see when a sample is drawn. Be
prepared to allow a sample to sit for a period of time (as long as 48 hours) if visual
confirmation of entrained air is required. Entrained air shows up as a decrease in sample
volume over time, as described in more detail elsewhere in this document.
Entrained air causes the density measurements to be lower than expected in bent-tube
sensors (Micro Motion U-tube and Omega style). This makes sense, since fluid with
entrained air is less dense than the fluid itself.
Entrained air causes the density measurements to be higher than expected in straighttube sensors (Micro Motion T-Series). In straight-tube meters the entrained air makes it
difficult to drive the tube and therefore the tube vibrational frequency goes down,
resulting in a false higher density reading. This is counterintuitive. More air means less
mass in the tube, which suggests that the density reading should be lower - but the
difficulty in driving the tube is the overriding factor here.
If the sensor is operating under an entrained air condition the drive gain will be saturated.
J. Reizner
Page 16 of 19
Date: 4/16/2001
Typically, the first indications of an air entrainment problem are inaccurate flow and density
readings. Usually the sensor is assumed to be at fault, and a calibration check is called for. Prior
to recalibrating or replacing the sensor, perform the following:
1. Monitor drive gain and density under normal flowing conditions. If the drive gain
is saturated and the density reading is in error per the previous statements, air
entrainment is the most likely cause. The following steps can confirm this
diagnosis.
2. Stop flow. Allow time for any air bubbles to rise out of the tubes. (Note: a more
viscous fluid will require more time for the bubbles to rise out.) Monitor drive gain
and density. If there is no air in the sensor, the density reading should be
accurate and stable, and the drive gain should be at a normal value (not
saturated.)
3. If possible, under normal flowing conditions draw a sample of the process fluid.
Be sure to follow all necessary safety precautions. Visually check the sample for
air bubbles.
4. If doubt remains as to whether the sensor or system is the source of the problem,
perform a calibration check of the sensor using non-aerated water and a bucket
or drum on a scale. Note: If the calibration check is performed by just diverting
the normal process fluid to a scale, all that will be confirmed is that the sensor is
not accurately measuring the flow. This cannot determine if the problem
originates with sensor or the process.
As noted previously, density and drive gain voltage can be monitored using either a Rosemount
275 Handheld Communicator or ProLink. Pickoff amplitude below 50 mV is considered to be
unreliable.
J. Reizner
Page 17 of 19
Date: 4/16/2001
Micro Motion Transmitter Parameters
Micro Motion transmitters contain two parameters that can be adjusted to minimize the effects of
slug flow (gas slugs in a liquid flow stream).
The Slug Flow parameter allows the user to set high and low process fluid density limits at which
the meter will report an error signal. Above the high or below the low process density limits, the
transmitter output signal will go to its minimum value (zero for frequency, 4 mA for a 4 to 20 mA
signal), and an error message (SLUG FLOW) will be displayed on the transmitter.
The Slug Flow Duration allows the user to program a time between 0 and 60 seconds into the
transmitter. If the fluid density goes outside the high or low slug flow limits, the flow outputs will
hold their last measured value for the period of time established by this parameter. After that time,
if the product density is still out of limits, the outputs go to their minimum value as described in
Slug Flow above.
Meter Zeroing
Capturing a good zero is essential for optimum sensor performance. It is best to zero the meter at
conditions approximating actual fluid operating conditions. Fluid temperature and pressure should
be representative of actual process operating conditions for an accurate zero and accurate flow
measurement.
Check the stability of the zero capture by performing the zeroing procedure a couple of times to
check the consistency of the zero captured.
J. Reizner
Page 18 of 19
Date: 4/16/2001
Recommendations
Following are a list of recommendations on how to deal with entrained air:
1. Use the best Coriolis meter possible, which is generally the Micro Motion ELITE. In
general do not use the T-Meter, due to its lower headroom as discussed previously.
2. Size the meter correctly as discussed previously.
3. If at all possible, determine the percent of entrained air expected (use the procedure
discussed previously). Discuss this data with Micro Motion.
4. Experience is that, depending on fluid viscosity, bubble size, etc. entrained air of 5% by
volume or more can be a problem for Coriolis mass flow meters. It is recommended that
entrained air be kept below 3% by volume. Recognize this is simply a rule-of-thumb and
is not a precise value for all situations. Knowledge and testing does not exist to give more
specific direction.
5. When possible, lower the amount of entrained air using techniques noted above.
6. Design the system, piping and installation to minimize possible problems with entrained
air. See elsewhere in this document for more details.
7. If the amount of entrained air is more than a Coriolis meter can deal with, consider
alternate flow meter technologies such as magnetic flow meters, vortex meters or positive
displacement flow meters. Recognize that each meter type has its drawbacks, and that
none can deal with an unlimited amount of entrained air. Slug flow also creates a problem
with most meter types. For more assistance with alternate meter selection refer to the
author.
8. If the entrained air cannot be dealt with, and no flow metering technology is appropriate,
load cell tank weighing systems may be the only option. The advantage of load cell tank
weighing systems is that they will read mass regardless of quantity of entrained air. The
disadvantages of load cell tank weighing systems are that they are very expensive and
are difficult to design, install, calibrate and maintain. They also generally require that the
process be designed as a batch process rather than as a continuous process, so
retrofitting load cell tank scales into existing continuous processes can be extremely
expensive.
J. Reizner
Page 19 of 19
Date: 4/16/2001
