Académique Documents
Professionnel Documents
Culture Documents
2008.10.08
1
10/14/2008
Types of Gas Meters
Positive Inferential
Displacement Meters
Meters
2
10/14/2008
Diaphragm Rotary
Meters Meters
3
10/14/2008
Inferential Meters
4
10/14/2008
Q = V x A
Where: Q = Flow Rate in CFH
V G V l it
V = Gas Velocity
A = Flow Area
Inferred Flow Rate = A flow rate derived indirectly from evidence
(e.g. velocity through a known area)
5
10/14/2008
Turbine Meters
Advantages Disadvantages
• Good Rangeability • Limited Low Flow
• Compact, Easy to Install • Susceptible to mechanical
• Direct Volume Readout wear
• No Pressure Pulsations • Affected by pulsating flow
• Wide Variety of Readouts
Wide Variety of Readouts
• Will not shut off gas flow
6
10/14/2008
Error The different between a measurement and its true value.
K‐factor A number by which the meter's output pulses are multiplied to
determine the flow volume through the meter.
Meter factor A number by which the result of a measurement is multiplied to
compensate for systematic error.
MAOP Maximum allowable operating pressure
Maximum allowable operating pressure
Pressure drop The permanent loss of pressure across the meter
Qmax The maximum gas flow rate through the meter that can be
measured within the specified performance requirement.
Qmin The minimum gas flow rate through the meter that can be
measured within the specified performance requirement.
Rangeability The ratio of the maximum to minimum flow rates over which the
meter meets specified performance requirement. Rangeability is
also known as the turndown ratio.
7
10/14/2008
Base conditions is a set of given temperature and pressure which describes the
physical state of gas in flow measurement.
physical state of gas in flow measurement.
Base conditions are defined jurisdictionally:
8
10/14/2008
The Ideal Gas Law
Conversion of the measured line volume to base volume relies on the equation
of state for the particular gas.
( )
PV = nRT ∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙∙ (1)
In this equation
P is the absolute pressure
V is the volume
n is the number of moles of the gas
R is the universal gas constant and equals 8.31451 J/mol K.
T is the thermodynamic (or absolute) temperature
This equation is valid for n moles of gas and describes the relation between
the volume V, the (absolute) pressure P and the (absolute) temperature T.
9
10/14/2008
10
10/14/2008
Flow volume register
Change gears Index Assembly
Lubrication fitting
Main rotor
Encoder/sensor
Nose cone
Top plate
Meter body
Conditioning fins
11
10/14/2008
Flow volume register
Index Assembly
Coupling
Conditioning plate
Main rotor
Main shaft
Meter body
Bearing block
12
10/14/2008
Turbine meters operating at atmospheric pressure show a
very non‐linear performance curve
50 psig
175 psig
720 psig
1440 psig
Turbine meters operating in a high pressure line displays a
much more linear and predictable characteristic
13
10/14/2008
The Law of Conservation of Energy
Kinetic Energy = Dynamic Energy of Mass in Motion
In an turbine meter, a portion of the linear kinetic energy of the moving gas molecules is
converted into rotational energy of the rotor
14
10/14/2008
——————— (1)
——————— (2)
Q ∝ ωi ——————— (3)
15
10/14/2008
Index Assembly
Change gears
The index assembly typically houses a
readout register of flow volume and
Signal terminals one or more sets of encoder disc and
Magnet reed
sensor sensor for generating flow output
pulses for electronic measurement
systems.
16
16
10/14/2008
The primary rotor of a dual‐rotor turbine meter
is basically the same as that of a single‐rotor
design. A second rotor is added for checking
and/or improving the measurement integrity of
the primary rotor under various flow conditions.
• Adjusted Volume at Initial
Calibration
• Basic Adjustment Principle
• Operating Changes in
Retarding Torque
Retarding Torque
Cut‐out view of an Auto‐Adjust meter
• Self‐Checking Feature
17
10/14/2008
18
10/14/2008
Main rotor The main rotor is calibrated to register 110%
Sensing rotor of the actual flow passing through the meter.
The sensing rotor is calibrated to register
10% of the actual flow. By design of the two
rotors and their placement in the meter
body, the flow error experienced by the
sensing rotor matches that of the main rotor
sensing rotor matches that of the main rotor.
The “Adjusted Volume” therefore provides a
very accurate account of the true flow.
19
10/14/2008
The Auto‐Adjust Turbine Meter Equations:
⎡ V sensing ⎤
ΔΑ = ⎢ × 100 ⎥ − A (2)
⎣ V main − V sensing ⎦
Where:
Vmain = volume by main rotor
Vsensing = volume by sensing rotor
Vadjusted = adjusted volume
Ā = average value of the factory sensing rotor % adjustment
ΔA = % deviation in field operation from factory calibration
20
10/14/2008
The Auto‐Adjust self‐checking Indicator:
⎡ V sensing ⎤
ΔΑ = ⎢ × 100 ⎥ − A
⎣ V main − V sensing ⎦
The parameter ΔA (delta A) is a self‐checking indicator of the performance
j j
of an auto‐adjust turbine meter. It shows the amount of adjustment the
meter is making, thereby warning the user of meter or flow conditioning
problems.
21
10/14/2008
1.0
0.5
R (%)
0
ERROR
‐0.5
‐1.0
0 25 50 75 100 125
CAPACITY (%Qmax)
An ideal turbine meter has a flat error curve extending from Qmin to Qmax
22
10/14/2008
1.0 Causes for “non‐ideal” turbine
meter behaviours:
0.5
%)
‐ Dirty gas
yg
Error (%
0
‐ Mechanical friction
‐ Pertubations
-0.5
‐ Density effect
-1.0 ‐ Reynolds effect
0 25 50 75 100 125
Capacity (%Qmax)
Typical performance curve of a turbine meter
23
10/14/2008
1.0
0.5
%)
-0.5
-1.0
0 25 50 75 100 125
Capacity (%Qmax)
Performance curve of a “real” gas turbine meter
24
10/14/2008
25
10/14/2008
Dirt accumulated on the rotor blades has a tendency to speed
up a turbine meter, thus resulting in overestimated flow
volume.
Dirt on the
rotor blades
1%
Error
-1%
Flow rate Q
26
10/14/2008
Dirt accumulated in bearings slows down a turbine meter,
therefore results in underestimated flow volume.
1%
Good bearings
Error
-1%
Flow rate Q
Damaged bearings
27
10/14/2008
0
At a constant inlet
‐2 pressure, increase in
ERROR (%)
mechanical friction
mechanical friction
‐4
due to bearing wear
‐6 has more significant
effect on LOW FLOW
‐8 accuracy.
NEW IN OPERATION
‐10
0 25 50 75 100
CAPACITY (% Qmax)
Q )
Damaged bearings slow down a turbine meter considerably
28
10/14/2008
The spin time of a turbine meter is a very good indicator of its condition
29
10/14/2008
Effect of spin time on the proof of a T‐35 Mark‐II turbine meter
30
10/14/2008
31
10/14/2008
Turbo‐Meter Oil
Alemite Fitting
32
10/14/2008
A i l Kf
A single K‐factor is often used to express the calibration of a turbine meter. It is
i f d h lib i f bi I i
simple but does not represent the operating characteristics of the meter
throughout the entire flow range.
33
10/14/2008
Meter Factors
34
10/14/2008
35
10/14/2008
36
10/14/2008
37
10/14/2008
Calibration adjustment of the mechanical output of a turbine
meter is typically accomplished by choosing an appropriate set
of change gears.
38
10/14/2008
Linearization
Linearization of flow meter
, p
If the error of a flow meter is known, it can be corrected for. Some flow computers
have the ability to carry out this correction. First the correction data resulting from
calibration are fed into the instrument. Next, the appropriate correction factor at
the particular flow rate is determined and applied. The result will be perfectly
linear.
39
39
10/14/2008
40
10/14/2008
AGA-7
41
10/14/2008
AGA-7
42
10/14/2008
AGA-7
43
10/14/2008
44
10/14/2008
45
10/14/2008
46
10/14/2008
47
10/14/2008
48
10/14/2008
49
10/14/2008
A straight AGA‐7 compliant meter run produces an uniform flow
profile with the same flow velocity across the cross‐section of pipe
An elbow or “tee” introduces a low level perturbation to the flow
50
10/14/2008
An additional out‐of‐plane elbow adds swirl to the already uneven flow profile
51
10/14/2008
An up‐stream regulator and out‐of‐plane elbow cause a high level of swirl and
jetting at the meter run
52
10/14/2008
Expanding from a smaller diameter pipe into a larger one introduces jetting
Expanding from a smaller diameter pipe into a larger one introduces jetting
which cannot be removed by a tube‐bundle flow straightener
Addition of an out‐of‐plane elbow upstream compounds the problem by adding
a swirl component to the flow
53
10/14/2008
19‐tube bundle straightening vanes
Flow conditioning plate
54
10/14/2008
55
10/14/2008
Integral conditioning plate
typically allows a turbine meter
to be installed in a non‐ideal
b i ll d i id l
meter run (e.g. short meter
run, elbows….) and maintain its
accuracy
Example of a turbine meter with
integral conditioning plate
56
56
10/14/2008
The pressure loss of a turbine
meter is directly proportional to
the flow pressure and specific
gravity and to the square of the
flow rate:
Δ Pm ∝ Pm × G × Q 2
Where
Constant Pm and G ΔPm = pressure drop across meter
Pm = absolute flow pressure
G = specific gravity of gas
Q = flow rate
57
10/14/2008
The pressure loss across a turbine meter is
directly proportional to the line pressure
and specific gravity and to the square of the
flow rate:
ΔPm ∝ Pabs × G × Q 2
In which
ΔPm is the pressure loss across the meter
Pabs is the absolute line pressure
G is the specific gravity of the gas
Q is the flow rate
58
10/14/2008
Properly sized blow down valve prevent over‐spinning
of turbine meter during line purge operation
59
10/14/2008
Typical maximum rate of
pressure change rating for
pressure change rating for
Rate of pressure change = Δ P turbine meter:
Δt
60
10/14/2008
TTurbine Meters display different response characteristics
bi M t di l diff t h t i ti
while speeding up and slowing down.
Underestimated
volume on rapidly
increasing flow
Overestimated
volume on rapidly
decreasing flow
Flow Rate (ACFH)
Actual flow
Flow registered by
turbine meter
Intermittent Flow Response of Turbine Meter
Material quoted in part from iMeter publication
61
10/14/2008
Due to the unsymmetrical transient response of turbine meters, they are
susceptible to overestimating the flow volume of pulsating devices such as
susceptible to overestimating the flow volume of pulsating devices such as
compressors and regulators.
Intermittent Flow Response of Turbine Meter
62
10/14/2008
Reynolds Number
ρνD
Reynolds Number =
η
ρ = fluid density
ν = flow velocity
D = pipe diameter
η = fluid viscosity
Recent research conducted at CEESI and SwRI on behalf of AGA
Recent research conducted at CEESI and SwRI on behalf of AGA
has demonstrated that commercially available gas turbine
meters have markedly different responses to given volumes of
natural gas at different Reynolds number.
63
63
10/14/2008
Effect Of Fluid And Non‐fluid Retarding Torques On Gas Turbine Meter Performance For Reynolds Number Below
100,000 (Source: Invensys Metering Systems)
64
10/14/2008
65
10/14/2008
Equation of State
The “State” of a gas
To calculate quantity in terms of base or standard volume one needs to
know the quantity of matter, e.g. the number of moles, that occupies
the actual volume measured under operating conditions.
This is done by using a suitable “Equation of State” for the type of gas
measured and by using measured pressure and temperature.
66
66
10/14/2008
Composition and compressibility
The composition of the gas influences the constants in the Equation of
State. This is mostly translated in the “Compressibility factor” or “Z”.
67
10/14/2008
Elevated Pressure Operation
1. Maximum Capacity in SCFH increases directly
as does the Boyle’s Law pressure multiplier
factor.
2. Minimum (Low Flow) Capabilities increases
directly with the square root of the Boyle’s Law
pressure multiplier factor.
68
10/14/2008
Calculating Rangeability
69
10/14/2008
Calculating Rangeability
Pressure Multiplier = (Line Pressure + Average Atmospheric) / Base Pressure * Compressibility Ratio
= (500psig + 14.48psi) / 14.73 * 1.0863
= 37.942
Maximum Flow Rate = Meter Rating * Pressure Multiplier
= 18,000acfh * 37.942
= 682,956 scfh = 683,000 scfh from table
Minimum Flow Rate = Meter Rating * Square Root of Pressure Multiplier
= 1200acfh * (37.942) 0.5
= 7391scfh = 7400 scfh from table
Range = Maximum / Minimum Flow Rater
= 683,000 / 7400 = 92:1
70
10/14/2008
Non‐hazardous Area
Hazardous Area
I t i i ll safe
Intrinsically f
NAMUR sensor Pulse
Flow Computer / RTU
or dry contact Amplifier
Power
Turbine Meter
Pulse amplifier converting NAMUR signal to a standard 24V digital signal
71
71
10/14/2008
NAMUR Signal
Inductive Sensor Capacitive Sensor
Supply Voltage = 8
8.2
2 VDC
Source impedance ~ 1 kΩ
Typical sensor current versus sensing distance
72
72
10/14/2008
Low flow
Low Flow High Flow
High flow
N
73
10/14/2008
Incorrect supply voltage or source impedance results in missed pulses
74
10/14/2008
4 1 271 208
1,271,208 8 898 458
8,898,458 44 492
44,492 4 10 990 320
10,990,320 76 932 238
76,932,238 384 661
384,661
75
10/14/2008
Questions ?
Oct. 8 2008
76
10/14/2008
Oct. 8 2008
77