Professional Documents
Culture Documents
Calculation of Flow Rate From Differential Pressure Devices - Orifice Plates
Calculation of Flow Rate From Differential Pressure Devices - Orifice Plates
Peter Lau
EMATEM - Sommerschule
Kloster Seeon – August 2-4, 2008
Contents
Introduction - Why using dp-devices when other principles
provide higher accuracy? Situation of very high temperatures.
How is flow and pressure linked?
Why do we need a discharge coefficient?
What must be known to calculate flow rate correctly?
What help does standards like ISO 5167 provide?
Differences between different version (1991 to 2003).
Which other information's are essential in a given application?
What influences the flow measurement result and how important
is it?
What uncertainties must we expect?
1
Introduction 1 – Flow is always measured indirectly
Flow rate – one of the most complex and difficult quantity
to measure and to calibrate.
No principle for direct measurement exists.
Many different physical measurement principles – all
have advantages and drawbacks.
Roundabout way via speed, rotation speed, frequency,
phase difference, temperature difference, force, voltage
etc. (thousands of patents exist!)
But: All metering devices need a calibration in one form or
another.
Question: How should we calibrate meters for use in hot
water (> 300 °C, 180 bar) or steam?
2
Introduction 3 – The Situation for Power Plants
In the late 60-ties and 70-ties no robust flow measurement
technique was available for the high temperature/pressure
applications.
The nuclear power industry is conservative and focused on
safety.
The thermal effect is limited to a licensed value.
A flow measurement accuracy of 2 % kept the output at 98
% of rated effect.
dp-devices showed no stable behavior due to fouling.
Over the years a lot of measuring technique has been
exchanged.
3
Introduction 5 - Difficulties in Meter Calibration
Gravimetric Method - (short traceability to kg, s, °C)
• The water must be kept at constant pressure and temperature
• Large amount of water – big vessels are needed
• Diverter and weighing tank in closed system to avoid steam
evaporation
A1 qm
A2
v1 v2
4
Flow in pipe with change in cross area
ps p2 static pressure
u1 Due to friction
us within the medium
u2
and with the wall
pressure falls along
V1 the pipe.
m1 Vs us > u1
ms V2 Ö
m2 ps < p1
Mass m1 and volume V1 of a ”flow element” Average speed increases,
does not change if area changes static pressure falls where
the area is smaller.
Measurement principle
5
Bernoullis equation on energy conservation
2 2 p = p0 where u = 0
The sum of static and dynamic pressure is the same everywhere in pipe.
⇓
(
ρ 2
)
1
Δp = p1 − p 2 = u2 − u12 ρ u 2 : ”dynamic pressure”
2 2
Replace the speed terms through geometric measures and insert.
qm = A ⋅ ρ ⋅ u
ρ 4 2 qm ⎛⎜ 1 1 ⎞ ρ 4 2 qm ⎛⎜ D2 D ⎞
2 2 4 4
4 2 ⋅ qm
2
Δp = ⋅ 2 2 ⎜ 4 − 4 ⎟⎟ Δp ⋅ D2 = ⋅ 2 2 ⎜ 4 − 24 ⎟⎟
4
u1 =
2
2 π ⋅ ρ ⎝ D2 D1 ⎠ 2 π ⋅ ρ ⎝ D2 D1 ⎠
D1 ⋅ π 2 ⋅ ρ2
4
4 2 ⋅ qm 2⎛ D ⎞
2 4
u2 = 2 ⋅ Δp ⋅ ρ ⋅ π2 ⋅ D2 = 4 2 qm ⎜⎜1 − 24 ⎟⎟
2 4
D2 ⋅ π ⋅ ρ
4 2 2
⎝ D1 ⎠
1 π2 Exchange unknown D2
qm = ⋅ D2 ⋅ 2 ⋅ Δp ⋅ ρ
2 4
⎛ D2 ⎞ 4
4 2
with known d
⎜1 − 4 ⎟
⎜ D ⎟
⎝ 1 ⎠
Introduce a discharge coefficient C relating
the unknown diameter D2 to the known
d
β = orifice diameter d!
D Introduce an expansion coefficient ε
for compressible media (gas, steam).
C π
qm = ⋅ε ⋅ ⋅ d 2 ⋅ 2Δp ⋅ ρ1
1− β 4 4
6
Why a discharge coefficient?
7
What does the standards treatise on?
8
Isa - och Venturi nozzles (part 3 of ISO 5167)
D: 50 – 500 mm
β: 0,3 – 0,8 D: 65 – 500 mm
ReD: 7x104 – 2x107 för β < 0,44 β: 0,316 – 0,775
ReD: 2x104 – 2x107 för β > 0,44 ReD: 1,5x105 – 2x106
All specified measures have to follow certain tolerances (not at least the sur-
face roughness) and must be verified by measurement at several places.
a) D: 100 – 800 mm
Production by
β: 0,3 – 0,75 ReD: 2x105 – 2x106
a) cast
b) D: 50 – 250 mm
b) machined
β: 0,4 – 0,75 ReD: 2x105 – 1x106
c) welded
c) D: 200 – 1200 mm
β: 0,4 – 0,7 ReD: 2x105 – 2x106
Cylindrical Pressure tappings
Entrance section
Flow-
direction
Conical Conical
Cylindrical
converging diverging
throat section section
section
9
The standard gives advice by installation examples
1: conical diameter increase
2: totally open valve
3: place for orifice
4: 90° elbow
10
Measurement of pressure and differential pressure
• In flow computers or intelligent transmitters the density for the
known liquid or gas is calculated from equations of state if
pressure, temperature and gas composition is known.
( )
+ 0,043 + 0,080 ⋅ e −10L1 − 0,123 ⋅ e −7L1 ⋅ (1 − 0,11⋅ A ) ⋅
β4
1− β4
Data on upstream
pressure tappings
(
− 0,031⋅ M2 − 0,8 ⋅ M2
1,1
)⋅ β
1,3
Data on downstream
pressure tappings
d
β= relation between orifice and pipe diameter
D
0 ,8
⎛ 19000 ⋅ β ⎞
A function of β and ReD A = ⎜⎜ ⎟⎟ Reader-Harris
⎝ ReD ⎠ Gallagher
2 ⋅ L2
M2 function of β and L2 M2 =
1− β
Equation
11
Standards are updated from time to time
ISO 5167 (1991) & ASME MFC-3M (2004) ISO 5167 (1991)
Discharge coefficient
0, 7 0,75
⎛ 10 6 ⋅ β ⎞ ⎛ 106 ⎞
C = 0,5961 + 0,0261 ⋅ β 2 − 0,216 ⋅ β 8 + 0,000521 ⋅ ⎜⎜ ⎟⎟ C = 0,5959+ 0,0312⋅ β2,1 − 0,1840⋅ β8 + 0,0029⋅ β2,5 ⋅ ⎜⎜ ⎟⎟
⎝ Re D ⎠ ⎝ ReD ⎠
( )
0 ,3
⎛ 10 6 ⎞ −1
+ (0,0188 + 0,0063 ⋅ A ) ⋅ β 3,5 ⋅ ⎜⎜ ⎟⎟ + 0,0900⋅ L1 ⋅ β4 ⋅ 1− β4 − 0,0337⋅ L'2 ⋅ β3
⎝ Re D ⎠
Stolz equation
( )
+ 0,043 + 0,080 ⋅ e −10 L1 − 0,123 ⋅ e −7L1 ⋅ (1 − 0,11 ⋅ A ) ⋅
β4
1 − β4
(
− 0,031 ⋅ M2 − 0,8 ⋅ M2
1,1
)⋅ β 1,3
⎡ ⎤
( ) Δp
1
⎛ p ⎞κ
( )
ε = 1 − 0,351 + 0,256 ⋅ β 4 + 0,93 ⋅ β 8 * ⎢1 − ⎜⎜ 2 ⎟⎟ ⎥
⎢ ⎝ p1 ⎠ ⎥
ε = 1 − 0,41 + 0,35 ⋅ β 4 ⋅
κ ⋅ p1
⎢⎣ ⎦⎥
1991 2003 Diff [%] 1991 2003 Diff [%] 1991 2003 Diff [%]
887,69 893,51 0,651 0,9908 0,9909 0,008 0,60027 0,60416 0,643
Air 0 ºC 1261,25 1268,26 0,553 0,9827 0,9828 0,010 0,59940 0,60267 0,543
1783,36 1791,27 0,442 0,9673 0,9673 -0,002 0,59872 0,60139 0,443
2207,29 2215,05 0,350 0,9527 0,9524 -0,034 0,59838 0,60069 0,384
12
Changes in flow rate calculation according to ISO 5167
between 1991 and 2003
With the same conditions the latest version of ISO 5167 translates
the differential pressure Δp to 0,2 - 0,3 % higher flow rate qm
Difference in mass flow calculation qm
0,35
Increase in qm from 1991 to 2003 [%]
0,30
0,25
0,20
0,15
Orifice plate
0,10 d: 23,49 mm
D: 42,7 mm
β: 0,55
0,05 low flow rate 60 kg/min
high flow rate 179 kg/min
0,00
0 50 100 150 200 250 300 350 400
Water temperature [ C]
13
How are a dp-devices often used?
The primary flow device is bought with a belonging discharge
coefficient C – (one value).
¾ calculated to a standard for a specified application
(medium, nominal flow rate, pressure, temperature > ReD)
or
¾ calibrated in a similar reference condition > ReD
Change in Change in
Δp(act ) Error in
qm (act ) = ⋅ qm (nom) indicated Calculated simplified
Δp(nom) Δp flow qn calculation
+ 1% 0,5 % 0,002 %
Example for orifice plate
In water at 90 ºC
Δp = 0,1 bar (10000 Pa)
β = 0,7361 +10 % -4,86 % 0,023 %
Re= ∼540 000
14
Not registered process parameters changes lead
to measurement errors
qm calculated from Δp C π
if everything else is known qm = ⋅ε ⋅ ⋅ d 2 ⋅ 2Δp ⋅ ρ1
1− β 4 4
Pressure difference Δp
Density of medium ρ1
Orifice diameter d
Measure-
Diameter relation β =d/D
P ment com-
Expansibility factor ε
puter or
Discharge Coefficient C
intelligent
Except β all other factors tansmitter
change with pressure P
and/or temperature
15
How are the variables and parameters interrelated?
Not relevant for liquids
Orifice diameter d β
pipe diameter
D
Pipe diameter, coef-
ficient of expansion D(T) qm
C
Dynamic viscosity μ1 ReD
of medium qm
prel. flow estimation L1
L2 β
Geometry of pressure
tappings up/down stream d (T)
Orifice diameter, coefficient of expansion
measured Δp
Density of medium at process conditions ρ1(T,p)
1,2
Viscosity of water as a Density of water as a
1000
1,0
function of temperature function of temperature
Viscositet [mPA s]
Temperature [C]
900
0,8
800
0,6
700
0,4
600
0,2
0,0 500
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400
For hot water these relations are reasonably well known - IAPWS
16
Reference calculation for testing - simulating
saturated steam
Iterative calculation of flow rate Calculate C1 Calculate qm
Assume C2 =0,61 Calculate qm1
Calculate qm2 Calculate ReD
Calculate ReD1 Calculate C Calculated
Process parameters
C π 4 ⋅ qm
qm = ⋅ε ⋅ ⋅ d 2 ⋅ 2Δp ⋅ ρ1 C = f (ReD ; β;L ) Re =
1− β 4 4
π ⋅ μ1 ⋅ D
17
How is flow measurement affected?
Unrecognized process Viscosity change: + 1 % or + 2 %
parameter change: at 90 and 200 ºC
C π 4 ⋅ qm
qm = ⋅ε ⋅ ⋅ d 2 ⋅ 2Δp ⋅ ρ1 C = f (ReD ; β;L ) Re =
1− β 4 4
π ⋅ μ1 ⋅ D
Δp = is assumed constant
C π d 4 ⋅ qm
qm =
1− β 4
⋅ε ⋅
4
⋅ d 2 ⋅ 2Δp ⋅ ρ1 C = f (ReD ; β;L ) β= Re =
D π ⋅ μ1 ⋅ D
Δp = is assumed constant
ΔD = 0,5 mm @ 1,17 % of D
ΔD = 1 mm @ 2,34 % Flow changes at 90 ºC high low flow
ΔD = 0,5 mm @ Δβ -1,16 %
ΔD = 1 mm @ Δβ -2,29 %
ΔD=0,5 mm @ -0,25 to -0,26 %
ΔD = 1 mm @ -0,49 to -0,51 %
Orifice/pipe 200 ºC
Flow changes at
d: 23,491 mm
D: 42,7 mm ΔD=0,5 mm @ -0,249 to -0,254 %
β: 0,5501 ΔD = 1 mm @ -0,483 to -0,496 %
The flow rate and temperature influences marginal
18
How is flow measurement affected?
Unrealized error in orifice diameter d +0,05 mm or +0,1 mm at 90 ºC
at 200 ºC
C π d
qm =
1− β 4
⋅ε ⋅
4
⋅ d 2 ⋅ 2Δp ⋅ ρ1 C = f (ReD ; β;L ) β=
D
Δp = is assumed constant
Δd = 0,05 mm @ 0,21 % of d
Δd = 0,1 mm @ 0,43 % of d Flow changes at 90 ºC high low flow
Δd = 0,05 mm @ Δβ -0,21 %
Δd = 0,1 mm @ Δβ -0,43 % +0,05 mm @ -0,473 to -0,474 %
+0,1 mm @ -0,948 to -0,950 %
Orifice/pipe Flow changes at 200 ºC
d: 23,491 mm
D: 42,7 mm +0,5 mm @ -0,472 to -0,473 %
β: 0,5501 +1 mm @ -0,946 to -0,947 %
The flow rate and temperature influences marginal
19
Uncertainty in expansibility factor ε
Uncertainty depends on geometry and pressure conditions.
With β = 0,7361 (d=73,61 mm / D=100mm)
ΔP = 15000 Pa ( 0,15 bar); P1 = 115400 Pa (1,15 bar absolute)
κ = 1,399 (for air at 50 °C).
Venturi tubes (
U(ε) = 4 + 100 ⋅ β8 ⋅
Δp
p1
% )
1,64 %
(4 + 100
Venturi nozzles U(ε) =Not ⋅ β )⋅
relevant p1
% 8 Δp
1,64 %
when measuring water flow
Δp
ISA nozzles U(ε) = 2 ⋅ % 0,26 %
p1
Δp
Orifice plates U(ε) = 3,5 ⋅ κ ⋅ p 1
% 0,336 %
μ JT =
Ru ⋅ T 2 ∂Z
⋅ Δϖ =
( )
1 − β 4 ⋅ 1 − C2 − Cβ2
⋅ Δp
p ⋅ c m,p ∂T p 1− β 4
⋅ (1 − C ) + Cβ
2 2
20
What uncertainties then must we expect in
measurement?
U(model) 5167
U(C) Way and equipment Installation induced
U(ε) To perform the calculations errors
(iterative approach)
U(β) Flatness
U(d) Concentricity
Difference to Distance to
experimental disturbances
experience
U(qm)
U(Δp) U(κ) Tolerances
U(T) U(μ1)
Continuous
U(P1) U(ρ) control
Measurement of Model of calculating
process variables process parameters
Quality of
Reproducibility Δp-device
of measurement
Analysis of the measurement uncertainty according to GUM
for an idealized situation using GUM-Workbench
21
Variable declaration
22
The total uncertainty budget
23
Uncertainty contributions at idealized conditions
24
Total measurement uncertainty
If all process parameters were known without any error:
• Simple and cheap production • Less dynamic flow range for fixed
(especially for big diameters) diameter
• Cheap in installation • Large pressure loss
• No moving parts - robust • Non linear output, complex calcula-
tion - needs urgent calculation help
• Works as ”primary element” –
• Process parameters must be
no calibration !? known for flow calculation
• Good characterized behavior • Problems with different phases or
• Simple function control flow variations
• Suits most fluids and gases – • Can be sensitive for wear and
even extreme conditions disposition
• Available in many dimensions • Very sensitive for disturbance and
installation caused effects
and a large flow range
• To achieve the measurement
uncertainties indicated by ISO
5167 many conditions must be
fulfilled
25
Limitations for Orifice Plates – med ISO-5167
26