Differential Pressure Producing Flow Elements

Differential Pressure Producing Flow Elements Standards Certification Education & Training Publishing Conferences & Exhibits Fort McMurray, Alberta Differential Pressure Devices for Liquid and Gas Flow Measurement Venturi Meters Orifice Plates Flow Nozzles Cone Meters Wedge Meters Meter Runs Performance Test Sections Averaging Pitot Tubes 1

Potential & Kinetic Energies of Flowing Fluids, Ideal Very Large V = H At the top of the tank, the potential energy of a reference mass of fluid is calculated: PE = mgh. At the discharge pipe on the datum line at the bottom of the tank, the kinetic energy of a reference mass is calculated: Gravitational Acceleration g = 3.174 ft/s Incompressible mv KE Fluid 1 v 1 To determine the kinetic energy content of the discharge jet, v 1 must be determined. With no frictional or other losses in the system, the discharge velocity is equal to the free fall velocity: v 1 gt g H g gh Potential & Kinetic Energies of Flowing Fluids, Ideal Very Large V = mv1 Since KE and v 1 gh, Incompressible Fluid Gravitational Acceleration g = 3.174 ft/s H v 1 then KE m(gh) mgh which is equal to the potential energy of the liquid at the top of the tank: PE = KE In an ideal, lossless system, ALL the potential energy is converted to kinetic energy.

Potential & Kinetic Energies of Flowing Fluids, Real V Hydraulic Gradeline Incompressible Fluid DP DH HGL Slope Due to Frictional Losses Venturi Meter The pressure produced by the line fluid at a given cross section of pipe is an indirect indication of the potential energy present at that cross section. The differential pressure produced by the venturi meter is caused by the conversion of potential energy at the inlet tap cross section to kinetic energy at the throat tap cross section. Theory of Operation, Orifice Plates High Pressure Pipe Tap High Pressure D & D/ Tap Low Pressure D & D/ Tap Low Pressure Pipe Tap High Pressure Vena Contracta Tap High Pressure Flange Tap Low Pressure Flange Tap Low Pressure Vena Contracta Tap Local Pressure High Pressure Corner Tap Low Pressure Corner Tap Plane of Vena Contracta Specific Headloss The Basics of Differential Pressure Measurement as Applied to the Orifice Plate 3

Theory of Operation, Cone Meters High Pressure Tap Low Pressure Tap Plane of Low Pressure Sensation Local Pressure DP Specific Headloss The Basics of Differential Pressure Measurement as Applied to the Cone Meter Theory of Operation, Segmental Wedge Meters High Pressure Tap Low Pressure Tap Local Pressure DP Specific Headloss The Basics of Differential Pressure Measurement as Applied to a Segmental Wedge Meter 4

Theory of Operation, Flow Tubes High Pressure Tap Low Pressure Tap Local Pressure DP Specific Headloss The Basics of Differential Pressure Measurement as Applied to the Lo-Loss Flow Tube Theory of Operation, Venturi Meters High Pressure Tap Low Pressure Tap Annular Chambers Local Pressure DP Specific Headloss The Basics of Differential Pressure Measurement as Applied to the Classical Venturi Tube 5

Why Specific Headloss? While permanent pressure loss appears to be simply a static pressure drop, it is really a dynamic value. Headloss is typically expressed in terms of PSI, inches of water column, kilopascals, etc., but the unit description in is: JOULES PER KILOGRAM OF FLOWING LINE FLUID Consequently, headloss represents an ongoing energy expense, the cost of doing business. Our duty is to minimize that cost. The Cost of Doing Business: A Comparison, Steam lb m /hr Steam Flow, P = 99.696 PSIA, T = 44 F r 1 =.61 34 lb m /ft 3 = 1%, Energy($) = 7 /kwh Operating 4 h/d, 365 d/yr 11.938 x 8.481 Orifice Plate DP = wc, DH = 95.9 wc AnnualCost.17 H Q $.17 (95.9) ( ) (.7) (1%) (.6134) $ 37,17 per year lb m /hr Steam Flow, P = 99.696 PSIA, T = 44 F r 1 =.61 34 lb m /ft 3 = 1%, Energy($) = 7 /kwh Operating 4 h/d, 365 d/yr 11.938 x 6.91 Venturi Meter DP = wc, DH = 11.5 wc AnnualCost.17 H Q $.17 (11.5) ( ) (.7) (1%) (.6134) $ 4,458 per year Using the Venturi instead of the Orifice Plate saves $ 3,714 annually. 6

SAGD Application: Flow Nozzle vs. Venturi Meter 3.86 x.53 Flow Nozzle DP = 5 wc, DH = 9.1 wc.17 H Q $ Annual Cost.17(9.1)(591.88)(.7) (1%)(.115684) $,773 per unit per year 3.86 x.54 Venturi Meter DP = 1 wc, DH = 1.5 wc.17 H Q $ Annual Cost.17 (1.5) (591.88) (.7) (1%) (.115684) $ 316 per unit per year Savings: $,457 per unit per year x 13 units = $319,41 per year! (Similarly sized vortex shedders and cone meters have even greater losses than the flow nozzle) Theory of Operation FLOW EQUATION FOR DIFFERENTIAL-PRODUCING FLOW METERS The flow equation for differential-producing flow meters is as follows: where: Q (kg/hr).16 446 856 d Fa CY 1β 4 Δ P ρ L g / g Q is the flow rate expressed in kilograms per hour; d is the diameter of the meter s throat or bore (millimeters); C is the meter discharge coefficient (dimensionless); Y is the expansibility factor (dimensionless); F a is the thermal expansion correction factor (dimensionless); DP is the observed differential pressure expressed in kilopascals; r L is the density of the line fluid at line conditions (kg/m 3 ); g is the local acceleration due to gravity (m/s ); g is the standard acceleration due to gravity (9.86 65 m/s ) Note that for most applications, g/g = 1; b is the ratio of the throat diameter (or bore) to the inlet (or pipe) diameter. 7

Theory of Operation (continued) Once the flow equation is understood, the underlying concepts reveal that the discharge coefficient, C, is actually a ratio: C Actual Rate of Flow Theoretical Rate of Flow (C 1, Y 1) The discharge coefficient simply relates an idealized flow rate to the real flow rate. When buying a flow meter, therefore, the client is essentially purchasing the manufacturer s knowledge regarding the value and the uncertainty of C (the discharge coefficient) and, if the line fluid is compressible, Y (the adiabatic expansion factor). Installation Effects, Nonimpact Venturis Straight Pipe Diameters 6 5 4 3 1 Effect of Concentric Reducer % Additional Uncertainty.1% Additional Uncertainty.% Additional Uncertainty.3% Additional Uncertainty.5% Additional Uncertainty.75% Additional Uncertainty 1.% Additional Uncertainty..3.4.5.6.7.8.9 1. Beta Ratio Straight Pipe Diameters Effect of Concentric Pipe Increaser 7 6 5 4 3 1 % Additional Uncertainty.1% Additional Uncertainty.% Additional Uncertainty.3% Additional Uncertainty.5% Additional Uncertainty 1.% Additional Uncertainty.% Additional Uncertainty..3.4.5.6.7.8.9 1. Beta Ratio Straight Pipe Diameters 18 16 14 1 1 8 6 4 Effect of Short Radius 9 Elbow..3.4.5.6.7.8.9 1. Beta Ratio No Additional Uncertainty.1% Additional Uncertainty.% Additional Uncertainty.5% Additional Uncertainty.75% Additional Uncertainty 1.% Additional Uncertainty 1.5% Additional Uncertainty In order to address concerns regarding a given metering design s uncertainty once installed, sensitivity tests must be run to determine the errors caused by common pipe fittings. Only flow test data can answer this question, the opinion of the seller does not matter. 8

Concentric Reducer Installation Effects, Nonimpact Venturis, Detail Effect of Concentric Reducer Straight Pipe Diameters 6 5 4 3 1 % Additional Uncertainty.1% Additional Uncertainty.% Additional Uncertainty.3% Additional Uncertainty.5% Additional Uncertainty.75% Additional Uncertainty 1.% Additional Uncertainty..3.4.5.6.7.8.9 1. Beta Ratio Concentric Reducer Installation Effects, Impact-Type Venturis, Detail Effect of Concentric Reducer Straight Pipe Diameters 6 5 4 3 1 No Additional Uncertainty.1% Additional Uncertainty.% Additional Uncertainty.3% Additional Uncertainty.5% Additional Uncertainty.75% Additional Uncertainty 1.5% Additional Uncertainty BVT-IL..3.4.5.6.7.8.9 1. Beta Ratio 9

Installation Effects, Impact-Type Venturis Straight Pipe Diameters 6 5 4 3 1 Effect of Concentric Reducer No Additional Uncertainty.1% Additional Uncertainty.% Additional Uncertainty.3% Additional Uncertainty.5% Additional Uncertainty.75% Additional Uncertainty 1.5% Additional Uncertainty..3.4.5.6.7.8.9 1. Beta Ratio Straight Pipe Diameters 18 16 14 1 1 8 6 4 Effect of Concentric Increaser No Additional Uncertainty.1% Additional Uncertainty.% Additional Uncertainty.5% Additional Uncertainty 1.% Additional Uncertainty.% Additional Uncertainty 3.% Additional Uncertainty 4.% Additional Uncertainty..3.4.5.6.7.8.9 1. Beta Ratio Straight Pipe Diameters 15 1 5 Effect of Short Radius 9 Elbow..3.4.5.6.7.8.9 1. Beta Ratio No Additional Uncertainty.1% Additional Uncertainty.% Additional Uncertainty.5% Additional Uncertainty 1.% Additional Uncertainty 1.5% Additional Uncertainty 3.% Additional Uncertainty The differential pressure produced by BVTs is an indication of the difference in the kinetic energy content of the line fluid between the high and low pressure tap cross sections. Due to differing velocity profiles, a given flow rate can possess different kinetic energies, and thereby introduce errors in the indicated flow rate. This is the essence of the study of installation effects and installed accuracy. Installation Effects (continued) Effects of a Single Elbow on Single Path and Dual Path Ultrasonic Flow Meters % Dual Path, Parallel to Elbow Plane Dual Path, Perpendicular to Elbow Plane Flow Error -5% -1% -15% Single Path, Parallel to Elbow Plane Single Path, Perpendicular to Elbow Plane 5 1 15 5 Pipe Diameters after Elbow 1

Installation Effects (continued) Effect of Eccentricity on the Discharge Coefficient of Orifice Plates Orifice C L Orifice Bore Orifice Plate OD Flange Bore Pipe ID Flange C L Eccentricity Pipe CL Source: Hobbs and Humphreys, Flow Measurement & Instrumentation, Vol. 1, No., pp. 133-14, 199 DEVIATION FROM STANDARD b =.4179 5% b =.53 4% b =.67 b =.7313 3% % 1% TAP ECCENTRICITY EXPRESSED AS A PERCENTAGE OF PIPE INSIDE DIAMETER -7-6 -5-4 -3 - -1 +1 + +3 +4 +5 +6 +7 Effect of Edge Sharpness on the Discharge Coefficient of Orifice Plates 1.6 Percent Change in Disch harge Coefficient 1.4 1. 1..8.6.4. O 5167 Limit per IS Source: Hobbs and Humphreys, Flow Measurement & Instrumentation, Vol. 1, No., pp. 133-14, 199 Radius <.4d..4.6.8 1. 1. 1.4 1.6 1.8. Radius/Bore x 1-3 11

Effects of Density and Viscosity on Coriolis Meters +.5 Error r (%) -.5-1. Benzene Diesel Oil Water Dual U-Tube P = 45 PSIG T = 6 F - 1.5 4 6 8 1 Flow Rate (Percent of Full Scale) Flow Calibration Water The discharge coefficient for a given meter design can only be determined through flow calibration. DP Timer 13.4 SEC Venturi Meter FLOW Weigh Tank or Volumetric Tank C.16 446 856d Collected Mass/Time F a P L 4 1/(1- ) Actual Rate of Flow (kg/s) Theoretical Rate of Flow (C 1, Y 1) 1

Flow Calibration (continued) Flow Calibration May lessen the effect of unseen manufacturing tolerances on flow measurement Establishes the data base for a given meter design May lessen the probability of litigation relating to the flow measurement A Meter Can Be Flow Calibrated Using an incompressible fluid (typically water) as the calibration medium Using a compressible fluid (typically air) as the calibration medium Directly against primary standards (gravimetric or volumetric) Indirectly against secondary standards (transfer masters) Flow Calibration (continued) To lower calibration uncertainty, one must limit the unknowns. If all critical metering dimensions time are known without error, then the only remaining error sources are those associated with C, the discharge coefficient Y, the expansibility factor. Flow calibration using a compressible fluid as the calibration medium allows for two significant sources of uncertainty, C and Y. Flow calibration using an incompressible fluid as the calibration medium eliminates the uncertainty associated with expansibility. 13

Uncertainty of C Calibrated Uncertainty reflects the uncertainty of the flow calibration Uncertainty of volume and/or mass determination Uncertainty of elapsed time determination Errors associated with installation Calibrated Uncertainty: Typically ±.% to ±.5% Uncalibrated Uncertainty is determined as follows: Geometrically similar meters are fabricated and flow calibrated The mean discharge coefficient is calculated The standard deviation is calculated The precision is determined Uncalibrated Uncertainty: Typically??? Only with test data can uncalibrated uncertainties be determined. Uncalibrated Uncertainty: Example 1 Flow Calibrated Meters N Flow Calibrated C Calculated Values 1.996 Mean Discharge Coefficient, C.99.9931 Standard Deviation, s ±.16% 3.9915 s (95% Confidence Level) ±.3% ( ) N 1 C 4.991 Precision (95% Confidence Level) ±.11% 5.9919 6.9935 UNCALIBRATED UNCERTAINTY 7.9894 (95% Confidence Level) ±.34% (Student's t) Precision N U 95 () (P) 8.99 9.997 1.9946 14

Data Analysis Discharge Coefficient in the Function of Pipe Reynolds Number 1.1 Disch harge Coefficient - C.99.97 C =.9891 Low R D tests defines C-shape, but forces extrapolation for higher R D s..95. 1.. 3. 4. 5. 6. 7. 8. Pipe Reynolds Number (x1^-5) Inlet Diameter (inches): 3.381 WYATT ENGINEERING, LLC Throat Diameter (inches): 1.377 4" LVM-B METER RUN Beta Ratio (dimensionless):.473 Serial Number: 494 For Pipe Reynolds Numbers >.74 x 1^5, Mean Discharge Coefficient:.9891 Bettis Atomic Research Laboratory March 5, Data Analysis (continued) Discharge Coefficient in the Function of Pipe Reynolds Number 1.1 Disch harge Coefficient - C.99.97 C =.998 High R D tests defines C-value, but provides no information for low R D s..95. 1.. 3. 4. 5. 6. 7. 8. Pipe Reynolds Number (x1^-5) Inlet Diameter (inches): 3.381 WYATT ENGINEERING, LLC Throat Diameter (inches): 1.377 4" LVM-B METER RUN Beta Ratio (dimensionless):.473 Serial Number: 494 For Pipe Reynolds Numbers > 1.6 x 1^5, Mean Discharge Coefficient:.998 Bettis Atomic Research Laboratory March 5, 15

Data Analysis (continued) Discharge Coefficient in the Function of Pipe Reynolds Number 1.1 Disch harge Coefficient - C.99.97 C =.99 To consolidate data mathematically violates knowledge of low R D C-shape and high R D C-value..95. 1.. 3. 4. 5. 6. 7. 8. Pipe Reynolds Number (x1^-5) Inlet Diameter (inches): 3.381 WYATT ENGINEERING, LLC Throat Diameter (inches): 1.377 4" LVM-B METER RUN Beta Ratio (dimensionless):.473 Serial Number: 494 For Pipe Reynolds Numbers > 1.49 x 1^5, Mean Discharge Coefficient:.99 Bettis Atomic Research Laboratory March 5, Data Analysis (continued) Discharge Coefficient in the Function of Pipe Reynolds Number 1.1 Disch harge Coefficient - C.99.97 C =.999 Treating data in a physically meaningful fashion respects knowledge of the low R D C-shape and the best estimate high R D C-value..95. 1.. 3. 4. 5. 6. 7. 8. Pipe Reynolds Number (x1^-5) Inlet Diameter (inches): 3.381 WYATT ENGINEERING, LLC Throat Diameter (inches): 1.377 4" LVM-B METER RUN Beta Ratio (dimensionless):.473 Serial Number: 494 For Pipe Reynolds Numbers >.74 x 1^5, Mean Discharge Coefficient:.999 Bettis Atomic Research Laboratory March 5, 16

How Codes Can Mislead Tube type flow straightener Valved vent Upstream pressure taps Throat taps Compressed gasket thickness not to exceed 1.8 mm (1/16 in.) Flow D d Throat tap nozzle D D D 16D min. D min. 1D min. ASME PTC-6 TEST SECTION How Codes Can Mislead Reference Curve for PTC-6 Nozzle Calibration 1..995 C.99.985.98 1.E+5 1.E+6 1.E+7 1.E+8 Throat Reynolds Number 17

How Codes Can Mislead in Microinches eter in Inches Surface Finish Throat Diame 1 1 1 Throat-Tap Nozzle Required Surface Finish to Produce a Hydraulically Smooth Surface 1 1 1 Throat Reynolds Number x 1^-6 Data Analysis The meter was calibrated so it must be good... The following slides come from different manufacturers web sites and promotional literature. If they truly understood the data, they couldn t brag about the results. Judge the data for yourself. 18

3/15/13 What Were They Thinking? Some Manufacturers Do Not Realize How Poorly Their Devices Perform: DC = +.3% C =.9918 Per ASME Brand X Website Error in flow coefficient value Error in flow coefficient Re behavior Physically impossible performance:. Nozzle creates energy (C > 1.) What Were They Thinking? Some Manufacturers Do Not Know the Difference between Data Scatter and Accuracy Error in flow coefficient values Error in flow coefficient Re behavior Data incorrectly analyzed Physically impossible performance C =.99 per ASME For Some, the Irony is Totally Lost Impressive, but the Patent Office Has No Record of Such an Application Brand Y Promotional Literature Who knows what they were thinking? 19

3/15/13 What Were They Thinking? Some Feel Safer Providing No Data at All No data, just sales talk. but Do Not Hesitate Using the Seal of Approval of a Recognized Test Facility. Brand Z Website What Were They Thinking? HHR Flow Tube is manufactured in accordance with ASME codes and standards quality controlled manufacturing to consistently produce the HHR Flow Tube with an accuracy of +/-½%. Fluidic Techniques maintains a database with over 1, independent laboratory flow calibrations Data incorrectly/not analyzed Incorrect citing of ASME Std. Incorrect mean C-value Incorrect uncertainty analysis Claims cannot be verified Shown above are the results of over 7 calibration runs of the FTI HHR Flow Tube. The data produces an average coefficient of discharge of.987 with standard deviation of.5. The data illustrates that for Reynolds numbers greater than, the coefficient of discharge is independent of Reynolds number. The data shown represents a wide range of tests including various line sizes, beta ratios and Reynolds numbers. Brand Z Website

3/15/13 Pressure Vessel Venturi Meters Fabricated, Wide Design Choice Stainless, Hastalloy, Monel, Inconel, etc. Flanged or Butt-Weld Ends Custom Laying Lengths Eccentric Design Concentric Design Insert-Type Venturi Meters Fabricated, Constructed from Practically Any Metal 3-Series and 4-Series Stainless Steels Hastalloy B & C Monel, Inconel, etc. Fabricated, Constructed from Various Composites Vinyl ester or polyester resin with fiberglass reinforcement PTFE, CPVC, etc. 1

3/15/13 ASME, ISO, AGA, API Products Flow Nozzles In accordance with ASME or ISO Subcritical or Critical Test Sections Orifice Plates & Meter Runs Paddle Type Square Edge Concentric Square Edge Eccentric Segmented Quadrant Universal Type In accordance with ASME, AGA, or ISO Custom Engineering (Manufacturers Should Work to Solve Clients Problems) Eccentric Design (Multiphase Fluids) Diaphragm Seals (No Tap Plugging or Emissions) Wafer-Style Insert Meters (Lower Cost) Special Materials (Demanding Applications) Titanium Teflon Kynar Monel, Inconel, etc.

3/15/13 Custom Engineering North Sea Venturi Meter with Diaphragm Seals and Integral Electronics Module Subsea Water Injection to Maintain Reservoir Pressure Placed 4 meters below North Sea Surface 3-1/8 5 PSI API Pressure Connections to Accommodate Diaphragm Seals, Eliminating Tap Plugging Fieldbus or 4 ma DC Output Signal Custom Engineering Hot Tap Process Seal Option Allows Use of Venturis for Coke Fines, Slurries, and Viscous Fluids Prevents Plugging and Contamination of Secondary Instrumentation Minimizes / Eliminates Fugitive Emissions Allows for Removal & Calibration under Pressure 3

3/15/13 Custom Engineering Miscible Vapor / Water Venturi System Single Meter to Measure Both Fluids Improves Efficiency and Output of Existing Wells Miscible Vapor Water Max. Flow Rate: 7 Nm3/d Min. Flow Rate: 8 Nm3/d Turndown: 5 to 1 Pressure: 5 MPa Temperature: 5 C Max. Flow Rate: 15 BPD Min. Flow Rate: 5 BPD Turndown: 6 to 1 Pressure: 1 MPa Temperature: 7 C Solution: ONE 5mm, 75mm, or 1mm Venturi Metering System Differential Pressure: 7.5 kpad Pressure Loss: 17 kpa Differential Pressure: 15. kpad Pressure Loss: 8.7 kpa Custom Engineering Miscible Vapor Measurement Uncertainty Band Flow Measurement Error Band Flow Measurement Error Band (%) 15. 1. 5.. -5. Single Transmitter System -1. Dual Transmitter System -15. 1 3 4 5 6 7 8 9 1 Percent of Maximum Flow Rate 4

3/15/13 Custom Engineering Water Measurement Uncertainty Band Flow Measurement Error Band. Flow Measurement Error Band (%) 15. 1. 5.. -5. -1. Single Transmitter System -15. Dual Transmitter System -. 1 3 4 5 6 7 8 9 1 Percent of Maximum Flow Rate Custom Engineering Erosion of Standard Weld Overlay Notes: 1. Crazing of the weld overlay, which will lead to erosion. Total loss of overlay throughout; field repair necessary 3. Unacceptable pressure tap edge; error inevitable 4. Overlay chipped and cracked 5. Possibly suitable for pipe, but not for flow measurement 5

3/15/13 Custom Engineering Comparison of Overlays Hard Facing, SlurryShield Hard Facing, Typical High density Intermetallic matrix 1.6 mm Thick Rockwell C 68 Abrasion Resistance Factor: 18- Throat finish: 1.6 mm RA Inlet section finish: 6. mm RA Uniform Thickness Calibration not required for.5% uncertainty band Resists channeling and uneven wear Tungsten or Chromium Carbide Overlay 6 mm Thick Rockwell C 58, typical Abrasion Resistance Factor: 4-5 Beaded interior finish, not smooth Interior must be machined to achieve predictable performance Flow calibration required for accurate performance Subject to uneven wear and chipping Custom Engineering Eccentric Venturi Meters For Multiphase/Wet Gas Metering; and For Slurry and Hydrotransport Flow Measurement Minimizes / Eliminates Build-Up Low Permanent Pressure Loss Flow Calibrated Uncertainty: ±.5% 6

3/15/13 Third Party Certifications GET THE CERTS! Make sure the ISO 9-series certification is current, applies to quality management of the product(s) of concern, and valid for the address where the product(s) is fabricated. Installation and use of pressure vessels in the European Union that do not conform to the Pressure Equipment Directive (PED) can result in civil and criminal penalties. New Product Release Features: For Use with Liquids and Gases Multiphase Meter for Wet Gases and Slurries Rugged Design: Use in Nearly Any Process or Environment Minimal Straight Run Requirements Conductivity and Velocity Are Not Issues Extremely Wide Turndown >1:1, Depending on Pressure Loss and Uncertainty Requirements Low Permanent Pressure Loss Constructed from Almost Any Material ±.5% Uncertainty without Flow Calibration ±.5% Uncertainty with Flow Calibration It s a Venturi Meter! 7