Benefits of Dry Calibrating

Ultrasonic Gas Flow Meters

AGA Operations Conference

Westin Hotel
Seattle, Washington

May 17-19, 1998

John Lansing
Instromet Ultrasonic Technologies, Inc.
Houston, Texas
and
Geeuwke de Boer
Instromet Ultrasonics B.V.
Dordrecht, The Netherlands
 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

A.G.A. Operations Conference Page 2

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

ABSTRACT
At present, in North America, South America and Europe, it is customary that ultrasonic
gas flow meters, used in fiscal metering or custody transfer metering applications, be
calibrated in a test facility by comparison to standards or reference devices. Calibration of
smaller meters (12 inch and less) is relatively practical with today’s facilities. As larger
meters become more commonplace, the process of flow calibrating these can be quite
expensive and time consuming. Because of the size, cost and operational considerations,
currently very few facilities are capable of calibrating meters larger than 12 inch. Thus, it
would be advantageous if another method of meter verification, such as a dry calibration,
could be developed and shown to accurately predict a meter’s performance.

The practice of dry calibration is well established for orifice metering. That is, meter
verification is based upon examination of the geometry, installation of the orifice plate, and
an accuracy test of instrumentation devices. Orifice meters are inferential in operation (no
moving parts), and it is assumed meter performance remains unchanged if all dimensional
and ancillary components are checked and verified. Since ultrasonic meters also have no
moving parts, it may be possible to imply a meter’s performance by dimensional and
operational checks.

The purpose of this paper is to briefly review the operation of an ultrasonic meter, discuss
the dry calibration procedure, discuss the sensitivity of the relevant variables, and present
results from more than 40 recently flow calibrated Q.Sonic multipath ultrasonic meters.
After a comprehensive dry calibration has been performed, flow calibration data for meters
ranging from 8 inch to 24 inch will show the linearity and reproducibility of meters available
with today’s technology.

INTRODUCTION
Multipath ultrasonic flow meters are becoming more accepted for custody transfer
applications. For these applications a calibration or verification of the metering device is
often a legal requirement, or a requirement based on a contract between the buying and
the selling partner. Ideally such a calibration or verification is performed by comparison of
the meter with a standard or reference. Traceability of standards used to the national or
international standards is a prerequisite.

Currently there is only one facility where large gas flows are available for calibrating 16 inch
(and larger) ultrasonic meters at full capacity. For the operator it is a burden when a meter
is be taken out of service and sent to a calibration facility for a calibration or verification.
Costs for the calibration itself are high, and in addition to this, the operator must face the
cost for removing the meter, transportation and possibly having the production facility shut
down. An additional restriction for the large meter sizes is that the test facilities are limited
to only a few months of the year for the high flow rates.

A.G.A. Operations Conference Page 3

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

One advantage of a meter being calibrated is that the owner will get a certificate with a
precise statement about the accuracy of the meter. Thus, it permits adjusting the meter in
order to minimize any measuring error or bias relative to the standards used. However,
considering costs and operational disadvantages, the concept of a dry calibration or
verification that does not require the meter be sent to a calibration facility is an attractive
idea. Similar practices are well established for orifice meters and have gained world-wide
acceptance.

This paper discusses the benefits achieved by dry calibrating Instromet Q.Sonic multipath
ultrasonic meters during the past year. More than 40 recent flow calibrations, obtained
from three flow laboratories located in Europe and the U.S., are used to demonstrate the
results of this procedure.

PRINCIPLE OF ULTRASONIC FLOW M ETERS

The principle of operation for an ultrasonic flow meter is illustrated in Figure 1. For the
sake of simplicity, only a single path meter is shown. However, the same principle applies
to other designs, regardless of the number of paths employed in the meter.

Figure 1: Principle of an Ultrasonic Flow Meter

Two transducers capable of transmitting and receiving ultrasonic sound pulses are
installed in the flow line in such a way that the ultrasonic sound pulses emitted from one
transducer can be received by the other transducer, thus creating an acoustic path. The
transducers alternately transmit and receive pulses within a few milliseconds.

The ultrasonic sound pulses travel, with respect to the gas, at the speed of sound. The
speed of a sound pulse along the acoustic path traveling downstream is increased with the
projection of the gas velocity onto the acoustic path. The speed of the sound pulse
traveling upstream along the acoustic path is decreased with a projection of the gas
velocity onto the acoustic path. This results in different travel times for the upstream and
downstream directions. These travel times can be determined from the following
equations:

A.G.A. Operations Conference Page 4

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

L
t down = Equation 1
C + Vm cosϕ
L
t up = Equation 2
C − Vm cosϕ
where
L : Length of the acoustic path
C : Speed of sound in the medium (gas)
Vm : Velocity of the moving medium (gas)
ϕ : Angle between acoustic path and a vector representing the direction in which the
medium moves

The following equation for the measured gas velocity can be derived form equations 1 and
2:

L  1 1
Vm =  −  Equation 3
2 cosϕ  t down t up 

It is important to notice that the speed of sound in the gas is eliminated in this last equation.
This means that measurement of gas velocity is independent of the gas properties such as
pressure, temperature and gas composition.

MEASURING DIFFERENTIAL TRAVEL TIME

As with any ultrasonic meter, determining the difference between upstream and
downstream transit times is critical to insure measurement accuracy. Transit times are
measured to a resolution of 10 nanoseconds (10-8 seconds). Thus, small differences in
transit times due to cable capacitance, solid-state component delay times, and quartz
crystals could become significant. These must be checked and verified not to exceed a
prescribed limit. Otherwise the meter may register flow when there is none. This is
discussed and referred to as “zero flow verification” in the recent release of A.G.A. Report
No. 9, Measurement of Gas by Multipath Ultrasonic Meters.

FLOW MEASUREMENT - SINGLE PATH METER

To measure the gas volume flow, gas velocity must be multiplied by the meter’s cross-
sectional area. When the gas velocity is equal over the whole cross-section, that is, it has
a uniform flow profile, calculating the volume would easy and more exacting. As fluids do
not behave this way, we need to correct the average gas velocity with a K factor that
compensates for the flow profile. Figure 2 (following) provides a pictorial representation of
this velocity profile.

A.G.A. Operations Conference Page 5

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

Figure 2: Flow Profile Correction Factor

As shown in Figure 2, Vm represents the average gas velocity as perceived by the

ultrasonic flow meter. This is the linear-weighted gas velocity averaged along the acoustic
path. This results in the following equation for the gas volume flow:

L  1 1
Q= ⋅ A ⋅ K ⋅  −  Equation 4
2 cosϕ  t down t up 
where
A : cross section of the pipe
K : flow profile correction factor

From studies, literature and our own research, Instromet established a relationship
between the Reynolds number and the flow profile correction factor K (also referred to as
Reynolds factor), which is shown below in Figure 3.

Theoretical Axial Path Velocity Profile Correction Factor

0.970

0.965

0.960
Correction Factor

0.955

0.950

0.945

0.940

0.935
1.0E+05 1.0E+06 1.0E+07 1.0E+08
Reynolds Number

Figure 3: Graph of Profile Correction vs. Reynolds Number

A.G.A. Operations Conference Page 6

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

This relationship is applicable to a single path meter with an acoustic path through the
center of a circular pipe. Dependent on actual circumstances, the flow profile may show
some variation resulting in an uncertainty in the flow profile correction factor K. This
uncertainty can be estimated based on the residual errors as observed in numerous tests.
From these results it appears a realistic estimate for the uncertainty in the Reynolds
correction factor for a single path meter would be approximately 1%.

FLOW MEASUREMENT - MULTIPATH METER

Custody transfer applications require the use of multipath meters. The reason is the
uncertainty of the profile correction factor for a single path meter is not accurate enough to
be acceptable for custody transfer applications. Multipath ultrasonic flow meters, by
implementation of integration techniques, allow using data from multiple acoustic paths to
improve the accuracy of the flow profile correction. Formally this can be represented with
the following equation:

 L  1 1  
Q= ⋅ A ⋅ K ⋅  −   ⋅ F Equation 5
 2 cosϕ  t down t up  

This equation is identical to that of a single path meter except that the part between
brackets   represents the integration using all acoustic paths.

This equation also includes a multiplier, F, which represents a correction factor, and is also
known as an “adjust factor.” This correction factor is a direct multiplier to the meter’s
output, and can be thought of as “electronic change gear.” Typically it has a default value of
1.0000 and can be adjusted based on a flow calibration of the meter in order to minimize
the meter error.

DRY CALIBRATION OVERVIEW

A dry calibration of a flow meter is not the same as a flow calibration. That is, it is not a
check of the meter’s result (measured gas volume / reference flow) with a standard or
reference device. If only dimensions were checked, and electronics and transducers
tested for proper operation, it may be more suitable to call this procedure a verification.
However, since the meter’s configuration is also adjusted as a result of this verification, the
procedure currently employed by Instromet may more appropriately be called a dry
calibration.

A dry calibration of an Instromet ultrasonic gas flow meter is performed in several major
steps. These are detailed in an Instromet procedure titled Dry Calibration of Q.Sonic
Meters, and is part of the documentation provided with all meters. This is a
comprehensive procedure that discusses all steps required to perform this task. A brief
summary is provided in the following section.

A.G.A. Operations Conference Page 7

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

DRY CALIBRATION PROCEDURE

The various steps of a dry calibration must be followed to insure meter performance and
provide an audit trail of the work performed. This section discusses the essentials of the
dry calibration procedure. An uncertainty analysis on these measurements, and their
associated impact on meter accuracy, is also presented in the following section.

VERIFICATION OF METER GEOMETRY

The relevant geometry parameters for an ultrasonic gas flow meter must be measured to
determine the acoustic path angles and path lengths. Reviewing Equation 5 one can see
these two are very important inputs for maintaining accuracy and repeatability of adjust
factor from meter to meter. This verification is critical to insuring meter performance.

Precision manufacturing of large meter bodies (some weighing in excess of 5,000 lbs.) is
complex and subject to various tolerances. If minor errors in fabrication can be
accommodated for, improved accuracy will result. Accurate verification of dimensional
tolerances is a very important aspect of the dry calibration.

ELECTRONICS AND TRANSDUCER FUNCTION TEST

After the electronics and ultrasonic transducers are installed on the meter body, a function
test is performed. This test insures the meter electronics package, also known as a signal
processing unit (SPU), and all transducers are operating properly. The meter is also
checked for leakage at this step.

It should be noted that by using state-of-the-art electronics and high quality quartz
oscillators, time measurements can be performed with excellent accuracy and stability
(better than 0.01% for both thermal and long-term). Insuring these components are all
operating correctly is part of the dry calibration procedure.

CONFIGURATION OF ELECTRONICS
Once correct dimensional measurements are obtained, an ultrasonic meter is configured
to accommodate any minor variations in fabrication. As was discussed earlier,
dimensional measurements are very important in determining path angle and path length.
These are the only configuration items which are adjusted for a specific meter (they are
derived from a Dimensional Inspection Report). All other configuration items are generic
for a given meter size and do not vary from meter to meter.

Minor manufacturing variations are taken into consideration by adjustments in path length
and path angle. Generally the path angle for the ultrasonic signal is 60 degrees relative to
the gas flow. If the mechanical dimensions are not accurate to within a few thousandths of
an inch, this angle will change. Measuring the “as built” dimension and computing the
actual angles and path lengths improves meter accuracy.

A.G.A. Operations Conference Page 8

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

ZERO FLOW CHECK

A zero flow check is a test to insure the meter does not indicate flow when none exists.
This is referred to as “false registration,” and can occur if there are timing errors with
electronics and transducers. In other words, even though there is no gas flow, a difference
in transit times between transducers can be detected. This difference will result in the
meter registering velocity.

In order to perform this test, the meter is first fitted with blind flanges and pressurized with
nitrogen. The assembly must be located in a thermally stable environment. Due to the high
resolution of the ultrasonic meter, even air blowing from an air conditioning system will
impact results. Since there is no gas flow (movement) through the meter (after thermal
equilibrium has been achieved), gas velocity observed on all acoustic paths should be
zero.

To insure there is thermal stability, the speed of sound is monitored as an indication of

temperature change. The speed of sound in gas is very sensitive to temperature. Once
the temperature has stabilized, 600 seconds of meter gas velocity data is collected. An
average is computed using an Excel spreadsheet. For example, if the average velocity for
a 12 inch meter is less than 0.01 fps, it is accepted. Typically the average velocity
recorded for this meter size is around 0.005 fps.

The maximum permissible zero flow velocity changes with meter size. Larger meters
having a proportionately lower value since they have a longer path length, and the offset
due to timing errors becomes less significant as path length increases.

It might be noted that this test can also be performed with an ultrasonic gas flow meter
installed in a field application, provided the meter can be isolated. However, utmost care
should be taken to avoid misleading results since even the slightest isolation valve
leakage, or thermal gradients due to sun and wind, can effect the reliability of results.

SPEED OF SOUND CALCULATION

The reader may recall from a previous discussion that the speed of sound calculation is
always available from an ultrasonic meter. That is, due to the nature of an ultrasonic meter,
a value is always computed. This speed of sound value is not significantly influenced by
gas flow. In other words, the speed of sound the meter reports will remain relatively
unchanged, regardless of flow rate, assuming the gas temperature, pressure and
composition remain unchanged.

As was also discussed earlier, accurate path length and proper operation of transducer
and electronics is required for correct speed of sound computation. Thus, the meter can
not produce accurate speed of sound information if the path length is incorrect, or the
electronics and transducers are not performing properly. Performing the following
procedure is also very important in providing a basis for field maintenance after the meter
has been placed into service.

A.G.A. Operations Conference Page 9

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

After the mechanical measurements have been completed, path lengths and path angles
are computed and uploaded to the meter. As the speed of sound calculation is dependant
on the path length, any path length error will impact the meter’s accuracy.

As with all devices, small tolerances can contribute to detectable errors that may be difficult
to measure directly. This is especially true when trying to determine the exact path length
when the ultrasonic signal is being “bounced” twice across the spool piece, as occurs in
the Q.Sonic meter. However, if the gas composition, pressure and temperature are
known, the theoretical speed of sound can be computed very accurately (generally better
than 0.05%).

The computed theoretical speed of sound can then be compared to the meter’s output and
used to “adjust” the path length. This is also known as acoustic path length adjustment.
The result is the highest possible accuracy in speed of sound measurements by the meter.
This not only reduces measurement uncertainty, but it helps insure repeatable meter
performance (prior to any flow calibration) and provides an excellent baseline for field (or
laboratory) checks in the future.

Currently there are several methods available for computing speed of sound in different
gas compositions. One computer program developed by Lomic, Inc., with the support of
GRI, is called SonicWare. This program can compute gas speed of sound with various
gas compositions, including pure gases such as nitrogen. This software is used for all dry
calibrations performed by Instromet.

ACOUSTIC PATH LENGTH CALIBRATION

To obtain accurate speed of sound results, the meter must be pressurized with a gas of a
known composition. Typically ultra-pure nitrogen (99.995%) is used due to its inert nature,
low cost, and known speed of sound properties. Since speed of sound varies with
different gas mixtures, using ultra-pure nitrogen (in lieu of commercial grade nitrogen)
minimizes uncertainty. Generally this test is performed in conjunction with the zero flow
verification procedure. However, several steps must be completed prior to proceeding.

Prior to testing, both pressure and temperature transmitters are calibrated. This is
accomplished by comparing each with known standards. If the transmitter’s calibrations
are not correct, adjustments are made until they agree within 0.2 degrees F and 0.5 psig.

To compute the theoretical speed of sound using SonicWare, accurate pressure and
temperature data must be obtained. This data is acquired by using the meter’s electronics
to collect 600 seconds of data. Information is “logged” using UNIFORM, a software
package used to communicate with the meter. The 600 readings are averaged to produce
separate pressure and temperature values.

Since the meter is initially exposed to atmospheric air prior to the blind flanges being
attached, it must be evacuated to insure no contamination once the nitrogen is introduced.
An evacuation procedure is performed twice with 100 psig of nitrogen being introduced in

A.G.A. Operations Conference Page 10

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

between each vacuum operation. This vacuuming and purging operation reduces the
uncertainty of this procedure by minimizing contamination of the nitrogen.

The meter is pressurized to 200 psig and the gas temperature is allowed to stabilize
before proceeding. This generally takes from several minutes to more than one hour,
depending upon meter size. Without thermal stability the nitrogen inside will be in constant
motion. This will not impact the speed of sound accuracy, but will impact the zero flow
verification (being performed simultaneously).

The observed speed of sound, when compared to the expected SonicWare value, will
reveal any error in the acoustic path length or travel time measurement. Error in travel time
measurement is unlikely since the frequency of the quartz clock has been checked before
this test, and can be easily verified if any doubt exists.

After the 200 psig test is complete, pressure and temperature values, recorded from the
meter’s log file, are used to determine the theoretical speed of sound. Before any path
length adjustments are performed (forcing the meter’s speed of sound value to agree with
SonicWare), a second test is performed at 500 psig. This second test helps insure no
computation errors have entered into the calculation, and also serves to show the meter is
performing properly at the higher pressure.

Once information has been obtained at both pressures, data reduction is performed. The
amount of path length change is based upon an average of the 200 and 500 psig data.
New computed path lengths are then uploaded to the meter. After re-configuration of the
path lengths, a third log file is obtained to insure the meter is computing the speed of
sound correctly. If the difference exceeds 0.1 feet per second, all test results are reviewed
and the source of error is located and corrected.

This test may be performed in practical applications as well, but it is critical to know the
exact gas composition at the time of the test. This may be difficult since gas composition
is constantly changing (during periods of flow), and it is often difficult to obtain up-to-the-
minute data. Of course, accurate temperature and pressure measurements are a
prerequisite.

Following are two photographs to provide the reader with examples of a dry calibration
procedure. Both are 30 inch multipath meters.

A.G.A. Operations Conference Page 11

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

Picture 1: Dry Calibration of 30 Inch Meter

Picture 2: Dry Calibration of 30 Inch Meter

A.G.A. Operations Conference Page 12

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

UNCERTAINTY OF A DRY CALIBRATION

As a general basis for the concept of dry calibration we will investigate the uncertainty in
the measured gas flow (volume or flow rate). Using Equation 5 we can calculate the
impact of the uncertainty of each individual parameter or measured value with respect to
the uncertainty of the measured gas flow. In this section we will examine the contribution of
these individual parameters and measured values.

The parameter F is a correction factor used when, based on a flow calibration, the reading
of the meter is adjusted. In case of a dry calibration this parameter is set to its default
value, based upon experience with numerous flow calibrations, and may not always be
1.0000. In case of a dry calibration this parameter is a constant, and is therefore not
associated with uncertainty. This may appear to be a contradiction since, based on a flow
calibration, this variable may be changed to a different value from the default. However, in
case of a flow calibration, this parameter corrects (compensates) the meter’s error due to
all other parameters and variables.

The data presented in this paper is based upon the results of more than 40 flow
calibrations of ultrasonic meters. As a result of these flow calibrations the default value of
the correction factor F was adjusted (if necessary). Since the variation in correction factors
reflect uncertainty due to the manufacturing geometry and dimensional parameters, as well
as other sources of uncertainty, the frequency distribution of this adjustment factor is a
good tool to verify the overall uncertainty for meters manufactured without flow calibration.

UNCERTAINTY OF PROFILE CORRECTION FACTOR

Based on the graph as presented in Figure 3, the Reynolds (profile) correction factor for a
single path meter is estimated to have an uncertainty of ±1.0%. Based upon Instromet’s
research and test results with Instromet’s multipath meters, and the path configuration of
the Q.Sonic meter, we estimate the uncertainty of the Reynolds (profile) correction factor to
be approximately 0.3% for a 5 path meter and approximately 0.4% for a 3 path meter.

UNCERTAINTY DUE TO METER BODY GEOMETRY

The relevant geometry and dimensional parameters of the meter body, which have an
impact with respect to the accuracy of an ultrasonic flow meter, can be seen in Equation 5.
They are as follows:

L : acoustic path length

ϕ : angle of acoustic path
A : cross section of the pipe

The acoustic path parameters are related to the location of the ultrasonic transducer’s
crystals, the surface that transmits and receives the ultrasound pulses. The transducer’s
position is determined by nozzles which are welded on the meter body. The transducers
are attached to the nozzles with seal housing which provide a gas-tight seal. The center
point of the face of the transducer is used as a reference point.

A.G.A. Operations Conference Page 13

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

Figure 4 is a typical spool piece drawing. The dimension between the transducer crystals
is represented by dimension “A” which is the center-to-center of the nozzles. The meter
diameter is shown as dimension “D.” The path angle, typically 60 degrees, is used to
establish the appropriate “A” dimension prior to manufacturing.

Figure 4: Typical Spoolpiece Drawing

In order to assess the uncertainty of the acoustic path parameters we need to take a closer
look at the manufacturing process. For simplicity we will look at a single reflection path
and use the dimensions of a 16” size meter body as an example. Parameters applicable
to a 16” Q.Sonic are as follows:

Nominal bore: 15.000 inches

Nominal path angle: 60.00 degrees
Nominal path length: 34.641 inches

Based upon measurement of dimensions “A” and “D” from Figure 4, we estimate the value
of these parameters as realized and reported are subject to the following uncertainty
values:

Path Angle : ± 0.05 degree

Computed Path Length : ± 0.021 inches

For the inner diameter uncertainty we might assume a practical value of ± 0.005 inches,
although, depending on manufacturing technology, this can be improved when necessary.

The contribution of each parameter’s uncertainty can be calculated according to

Equation 5, and are as follows:

A.G.A. Operations Conference Page 14

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

path length L : ± 0.06 %

1/cos ϕ (path angle) : ± 0.15 %
cross sectional area A : ± 0.06 %

When all of these factors add up to a worst case situation, the uncertainty due to the meter
body geometry and dimensions would be ±0.27%. However, since these contributions are
due to independent sources of error, the expected magnitude of the error may more
appropriately be calculated according to the root mean square rule. This results in an
uncertainty of ±0.18%.

UNCERTAINTY DUE TO TIME MEASUREMENT

The uncertainty due to the measurement of the travel times can be assessed by
distinguishing (similar to DP transmitter practice) between the zero reading and gain
errors. The zero error is related to the resolution of the travel time measurements and
small offsets in the travel time measurement. This could introduce a travel time difference
being measured even when the gas flow velocity is zero. From equation 4 it can be derived
that the following equation is applicable:

C 2 ⋅ δt ⋅ tan(ϕ )
δV = Equation 6
4D

where
D : meter body inner diameter
C : speed of sound in the gas
ϕ : acoustic path angle
δt : error in differential time measurement
δV : error in measured gas flow

The uncertainty in the travel time measurement is at maximum 10 ns. From the previous
example at approximately 200 psig, using the 16” meter body would result in the following:

C = 1160 feet per second

D = 15.000 inches
ϕ = 60.00 degrees

The resulting uncertainty in time measurement translates into a gas velocity of

approximately 0.06 inches/sec.

Using the gas velocity range that can be handled with our ultrasonic flow meters (maximum
gas velocity 100 fps, or 30 m/s) this can be converted to a relative percentage value.
Figure 5 is a graph showing the relative error in percentage, based upon a 0.06 inches per
second error, as a function of the gas flow. This graph was based upon a 16 inch meter

A.G.A. Operations Conference Page 15

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

size. As the meter size is reduced, the offset will proportionately become a noticeably
larger error at the lower flow rates.
Figure 5: Absolute and Relative Error Due to Travel Time Uncertainty /Offset

Absolute Error From 0.06 Inches per Second Offset

2.0

1.5

1.0
Percent Error

0.5

0.0
0 10 20 30 40 50 60 70 80 90 100
-0.5

-1.0

-1.5

-2.0

Velocity

This graph shows clearly that the uncertainty of the travel time measurement is the
dominant factor at low flow rates. This imposes a limitation on the low side of the
operating range but has no significant effect in the normal operating range. Again, this
example is a worst-case scenario. For reference, typical offset errors for a 16 inch meter
range on the order of 0.08 inches per second, well below the maximum permitted offset of
0.48 inches per second from A.G.A. 9.

Travel time measurement is related to a clock signal. If the clock is fast or slow it would
have a proportional impact with respect to the measured gas flow. However, since this
clock is a high stability quartz clock (accuracy ±50 ppm or 0.005%), this can be ignored.

TOTAL METER UNCERTAINTY

In the previous sections it has been shown that the significant sources that contribute to the
total measurement uncertainty of an ultrasonic flow meter are:
Flow profile correction factor K ±0.3%
Meter body geometry (rms) ±0.18%.
Worst case, the combination of both sources of uncertainty would result in a total
uncertainty of ±0.48 %. Since these are independent sources of uncertainty it is justified to
estimate the total uncertainty using the square root rule to calculate total uncertainty as:

0.32 + 0.182 = 0.35%

This number is of the same order of magnitude as the uncertainty of the best “world class”
flow calibration facilities (0.25 – 0.3%).

A.G.A. Operations Conference Page 16

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

FLOW CALIBRATION RESULTS

This section presents data on more than 40 flow calibrations during the past year. This
data was obtained from the following three different facilities: SwRI in San Antonio, Texas,
NMi in Westerbork, The Netherlands, and Pigsar in Dorsten, Germany. This data has not
been linearized, adjusted, or manipulated in any way. The only meter configuration change
was adjustment of the multiplier F (also called the adjust factor).

This first series of graphs show linearity results based on meter size. Each graph shows
data for a given meter size. Regardless of where the meter was calibrated, linearity results
were plotted together. These results were not linearized in any way. The only adjustment
available on a Q.Sonic meter is the final factor (F). In other words, the meter’s linearity
remains unchanged from the dry calibration, and would be the same whether or not the
meter was flow calibrated.

Three photographs of 10 and 20 inch meters being calibrated have been included to give
the reader a mental picture of two commonly used calibration facilities.

Picture 3: 12 Inch Meter Calibration, SwRI, San Antonio, Texas

A.G.A. Operations Conference Page 17

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

Picture 4: 20 Inch Meter Calibration at NMi, Westerbork, The Netherlands

Picture 5: 20 Inch Meter Calibration at NMi, Westerbork, The Netherlands

A.G.A. Operations Conference Page 18

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

0.7 0.7

0.5 0.5

0.3 0.3

Percent Error
Percent Error

0.1 0.1

-0.1 0 10 20 30 40 50 60 70 80 90 100 -0.1 0 10 20 30 40 50 60 70 80 90 100

-0.3 -0.3

-0.5 -0.5

-0.7 -0.7

Velocity (Feet per Second)

Velocity (Feet per Second)

0.7 0.7

0.5 0.5

0.3 0.3
Percent Error
Percent Error

0.1 0.1

-0.1 0 10 20 30 40 50 60 70 80 90 100 -0.1 0 10 20 30 40 50 60 70 80 90 100

-0.3 -0.3

-0.5 -0.5

-0.7 -0.7

Velocity (Feet per Second) Velocity (Feet per Second)

0.7 Five 10 Inch Meters 0.7

Ten 8 Inch Meters
0.5 0.5

0.3 0.3
Percent Error

Percent Error

0.1 0.1

-0.1 0 10 20 30 40 50 60 70 80 90 100
-0.1 0 10 20 30 40 50 60 70 80 90 100

-0.3
-0.3

-0.5
-0.5

-0.7
-0.7
Velocity (Feet per Second) Velocity (Feet per Second)

A.G.A. Operations Conference Page 19

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

The next series of graphical results show the reproducibility obtainable with today’s
technology. Meter data for a given meter size is presented on one graph. However, since
different flow facilities may provide different “F” adjustment factors, differences plotted
were developed based on the same size meters calibrated at the same facility. For
example, deviations between all 12 inch meters, calibrated at one site, were compared to
the average from that location only, not an average of all calibration facilities. This is a
better representation of meter reproducibility as all results are compared to the same
measurement standards, rather than introducing any uncertainty that may exist between
different laboratories.

0.3 24 Inch Meters 0.3 20 Inch Meters

% Difference From Average

% Difference From Average

0.2 0.2

0.1 0.1

0.0 0.0
1 2 3 4 1 2 3 4
-0.1 -0.1

-0.2 -0.2

-0.3 -0.3

Individual Meter Individual Meter

0.3 16 Inch Meters 0.3 12 Inch Meters

0.2 0.2
% Difference from Average

% Difference from Average

0.1 0.1

0.0 0.0
1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

-0.1 -0.1

-0.2 -0.2

-0.3 -0.3
Individual Meter
Individual Meter

0.3 10 Inch Meters 0.4 8 Inch Meters

0.3
0.2
% Difference From Average

% Difference From Average

0.2
0.1
0.1

0.0 0.0
1 2 3 4 1 2 3 4 5 6 7 8 9 10
-0.1 -0.1

-0.2
-0.2
-0.3

-0.3 -0.4
Individual Meter Individual Meter

A.G.A. Operations Conference Page 20

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

The following figure is a summary of all the previous graph’s data. It includes meter sizes
ranging from 8 inch to 24 inch. This graph clearly shows the reproducibility of all the
calibrated meters is within ±0.3%. In other words, the adjust factor F is predictable to
within ±0.3% when compared to a flow calibration.

0.3 Summary of 48 Meters

0.2
% Difference from Average

0.1

0.0
0 10 20 30 40 50

-0.1

-0.2

-0.3
Individual Meter Results

Figure 6: Summary of 48 Meter Calibrations

COST OF CALIBRATION
Calibration of an ultrasonic meter can be a significant part of the overall purchase price.
Following is a table with an approximate cost for calibrating several different meter sizes.
Costs estimates include appropriate shipping and customs charges from Houston to
Europe. A typical list price for a meter is shown for reference in column 2. The third column
is for a meter without a flow conditioner. The fourth column is for a meter including flow
conditioner and associated piping (typically 2 or 3 spool pieces).
Additional charges have been included for bolting the multiple spool piece assembly
together and for shipping the additional weight. These costs are typical for calibration in
Europe. A significant portion of these cost can be attributed to shipping via air freight.

A.G.A. Operations Conference Page 21

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

Meter Information Approximate Meter Calibration

NPS Size Typical No Flow With Flow
24 $82,000 $30,000 $45,000
20 $68,000 $25,500 $39,000
16 $58,000 $18,000 $27,000
12 $49,000 $12,000 $23,000
10 $38,000 $8,000 $20,000

Table 1: Meter Calibration Costs by Size

This summary shows the cost of calibration can be significant when compared to the
meter’s purchase price. It’s also apparent that larger meters have a higher percentage for
calibration cost when compared to smaller meters. The costs listed are for calibration in
one direction only. An additional charge approximately equal to the “No Flow Conditioner”
value will be added to the bi-directional calibration costs will add cost.

CONCLUSION
Ultrasonic meters have gained widespread acceptance for applications in the natural gas
industry during the past few years. This is primarily due to several cost-effective benefits.
These include low maintenance, lower initial cost for most applications, bi-directional (with
equal accuracy in both directions), large rangeability, no pressure drop, no moving parts,
and a high degree of accuracy and linearity. However, one aspect of purchasing ultrasonic
meters which can add significant cost to the purchase price is flow calibration. This is
especially true for the larger meters.

There are many applications which may not require accuracy equal to the best “world class”
flow calibration laboratories. These applications include pipeline balancing, lost and
unaccounted for (LUAF), underground storage, odorant injection, and non-critical custody
transfer applications. Thus, if ultrasonic meters are reproduced to within ±0.3%, and linear
within ±0.1%, when compared to “world class” flow calibration facilities, significant cost
savings can be realized for many customer applications.

Once an ultrasonic meter has been flow calibrated, its accuracy becomes equal to that of
the flow laboratory. That is, the meter can not be any more accurate than the absolute
accuracy of the facility. In other words, flow calibrating reduces measurement uncertainty to
that of the facility, plus repeatability and linearity.

As has been shown from the graphs on page 16, linearity is generally better than ±0.1%,
and often within ±0.05%. These numbers are probably within the stated linearity of most
facilities. Thus, one might state the meter is accurate to within ±0.1%, assuming the flow
calibration facility is correct.

A.G.A. Operations Conference Page 22

 Benefits of Dry Calibrating Ultrasonic Gas Flow Meters

It is important to remember that the linearity shown in these graphs is not the result of a
multi-point calibration. The Q.Sonic meter design produces a linear performance curve
which does not require linearization. The only adjustment applied to these results was to
change the F factor (also known as adjust factor). Deviations in F factors for a given meter
size were shown on page 17, with a summary of some 48 meters on page 18. It clearly
shows all meters were within ±0.3%, and most were within ±0.2%. Without linearity by
design, dry calibration results could not be used to imply this level of installed accuracy.

As presented in this paper, the concept of dry calibration can be applied to Instromet
Q.Sonic ultrasonic gas flow meters in very much the same way as with dry calibration of
orifice meters. Thus, we consider it justified to state that dry calibration techniques, when
applied to Instromet ultrasonic gas flow meters, will assure accuracy is within stated
specifications, plus far exceed the requirements of A.G.A. 9.

Instromet’s continuing commitment to research will further improve the meter’s dry and flow
calibration results. This will be reflected by improved accuracy, linearity and low-flow
performance. It’s our goal to continue reducing the uncertainty of the meter’s performance,
based upon a comprehensive dry calibration, making the ultrasonic meter an even more
cost-effective solution for today’s competitive marketplace.

A.G.A. Operations Conference Page 23