FOCUS DISCUSSION GROUP E

•

•Effect of Increasing Plate Thickness on the Metering Accuracy of an 8 Inch

Orifice Meter Run.

U Karnik, Novacor Research & Technology Corporation

•

•

•EFFECT OF INCREASING PLATE THICKNESS ON THE METERING

ACCURACY OF AN 8 INCH ORIFICE METER.

U.KarnikNOVA Research & Technology Corporation2928-16th Street N.E.Calgary. Alberta, CanadaT2E 7K7

• ABSTRACT

The present study investigates the use of a thicker (0.25in.) orifice plate in a standard

8inch orifice meter run. In particular, the effect of retrofitting existing meter runs (old fitting)

designed for 0.125in. thick plates as well as the performance of newly designed meter runs (new

fitting) for 0.25in. plates is studied. For the 0.25in. thick plate, the effect of bevel angle and

orifice edge thickness on metering accuracy is also evaluated.

Results obtained indicate that:

a) 0.25in. thick orifice plates perform identically when used in old fittings or in new fittings. The

use of 0.25in. thick orifice plates would not only increase the capacity of the meter run but

also result in repeatable metering independent of [)-ratio. For 0.3:5 13 :50.6, (i.e. for mass

• flow> 3kg/s), the average error is around 0.31% with a 2a of ±O.13%.

b) in old fittings, for [)-ratios <: 0.4, there is no difference between the 0.25in. and the 0.125in.

plates. For [)-ratios < 0.4, the 0.25in. thick plates would over-register by around 0.3%.

c) bevel angle (30° :5a :545°)and orifice bore edge thickness (O.06in. :5e :5 0.16in.), do not

have a measurable effect on the metering accuracy in old or new fittings.

•

d) plate thickness (0.125in. vIs 0.25in.) does not affect the performance of non-beveled plates

in the old fitting.

e) data indicates that for the 0.125in. thick plate, the difference between the beveled and non-

beveled plates depends on the [)-ratio and does not indicate a specific trend. In the case of

the 0.25in. thick plates, the non-beveled plates tend to under-register in comparison to the

beveled plates at [)-ratios < 0.4.

1

1.D INTRODUCTION

It has been recognized in the past that the capacity of a orifice meter run could be

increased by increasing the differential pressure to values beyond the allowed 50kPa (200in.).

The issues to be recognized in adopting such a procedure are that

1. the plate should not deform permanently and

2. the plate should not undergo elastic deformation since this could lead to metering error.

The first issue can be easily addressed using available stress analysis expressions [1[.

The orifice fitting is treated as a simply supported circular plate with a central hole. Results of

this analysis are shown in Figure 1. The allowable differential pressure before permanent

deformation is govemed by the I3-ratio and the ratio of the support diameter to the plate

thickness (aspect ratio). If differential pressures of at least 200kPa are to be used for I3-ratios ~

0.2, then the aspect ratio of the plate should be around 40.

The second issue is also a function of the above two parameters [2[, [31, 141, (5[. (6[. Apart

from this work, originating from the British Gas Engineering Research Station, other work on this

subject matter has also been published (7[, (8[, (9[. The effect of plate deformation on metering

error [2] is presented in Figures 2a,b,c & d. It is evident that if the metering errors due to plate

deformation are to be minimized «0.05%), then the aspect ratio should be less than 40.

Experimental work performed on natural gas (10[ at the NOVA gas dynamic Test Facility,

confirmed the theoretical estimates shown in Figure 2 for aspect ratios of 32 (4in. meter run) and

80 (10in. meter run). Details of the theoretical estimates are also presented by (10[.

On examining metering standards and existing Sin. meter runs in the NOVA system, it

would, not be possible to increase the capacity of these meter runs under the current

specifications l.e, recommended plate thickness of 0.125in (an aspect ratio of 67,5), One

approach to increase the capacity would be to use a thicker plate e.g. O.25in. (an aspect ratio of

33,75). This results in the following issues that need to be addressed.

1. In the case of retrofitting existing meter runs with 0.25in. plates, one would be altering the

location of the pressure taps. Standards stipulate that the flange taps should be located

1inch from the upstream and downstream surface of the plate respectively .. On using the

thicker plate, the location of the taps from the respective surfaces would now be 0.9375in.

The question to be answered is whether, this small shift in tap locations causes a significant

2

•

•

•

•

•

•

•

•

error in metering? Further, can the R-G equation be modified to accommodate the new tap

locations as part of the tap term ?

2. A corollary of the above issue is that if the upstream edge is the governing parameter for the

flow structure, then the upstream edge will now be located 1.1S7Sin. from the downstream

tap instead of the 1.12Sin. as in the case of the O.12Sin.thick plate. This change in the flow

structure could alter the measured pressure drop and hence the measured flow rate.

3. If we buy into point #2, then consider the new fitting i.e. one manufactured for the O.2Sin.

thick plate with the appropriate location for the flange taps. In this case all specifications of

the A.G.A. standard are met i.e. the plate thickness does not exceed the maximum value

and the flange taps are appropriately placed. If the issue raised in point # 2 is now applied,

the distance from the upstream edge to the downstream tap is now 1.2Sin. Would this

further affect.metering error adversely?

4. The issues are further complicated by the fact that the orifice database on which the

equation was generated apparantly does not include any data from an Sin. meter run.

From NOVA's perspective, the need to increase meter capacity exists. This can be done

quite easily for the 4in. and 6in. meter runs without any adverse effects on the measured flow.

There exists a need to increase the capacity of Sin. meter runs. If NOVA is to request vendors

to supply meter runs appropriately designed for the O.2Sin. thick plate, it is. important that we

ensure that the performance of a retrofitted meter compares well with that of the new meter

design.

Hence, the objectives of the project are follows:

a) To evaluate the performance of old Sin. meter runs designed for O.12Sin.thick plates.

b) To examine the effect of retrofitting old Sin. meter runs with O.2Sin.thick plates.

c) To study the performance of the new Sin. meter run designed for O.2Sin.thick plates.

d) To study the effect of bevel angle and bore edge thickness on metering accuracy.

3

2.0 EXPERIMENTAL FACILITY

Experiments were performed in natural gas at NOVA's Gas Dynamic Test Facility at a

line pressure which varied between 5000-6000kPa. A sketch of the test loop is shown in Figure

3. Following a straight section of around 390, the flow was conditioned by means of a NOVA

SOE flow conditioner. The orifice plate was located 300 downstream of the flow conditioner,

ensuring a fully developed. symmetric velocity profile at the orifice plate. Past measurements

with the NOVA 50E 111J. (12( indicate that the profile is fully developed at 80 downstream of the

flow conditioner with a 50 upstream mixing length. The orifice meter run was a standard Daniel

orifice fitting. A thermowell was located 5.30 downstream of the orifice plate for the insertion of a

RTO to measure the flow temperature.

The static pressure was measured by means of a Rosemount Smart static pressure

transmitter and a Rosemount Smart temperature transmitter was used to power the RTD. The

differential pressure across the orifice plate was measured by a Rosemount Smart differential

pressure transmitter with a range from 0-200kPa. The static pressure transmitter was calibrated

with an Ametek dead weight tester and its performance was monitored. at zero flow conditions.

with another calibrated Smart transmitter located elsewhere in the loop. If the difference

between the two transmitters exceeded by more than 0.1% (stated accuracy of the transmitter).

the transmitters were re-calibrated.

The temperature RTD was calibrated with an Automatic Systems Laboratory (8125) dry

block calibrator using a NIST traceable ERTCO-HART 850 high precision platinum RTD as a

reference. Several checks performed during the course of the tests and the calibration was

stable to within 0.05deg. C. The differential transmitter was calibrated at ambient conditions with

an Ametek dead weight tester and was adjusted for the zero offset at line conditions. Calibration

checks indicated that the drift in the slope was not more than 0.05%. The calibration curve was

corrected for non-linear effects. Some typical calibration checks are presented in Table 1.

The set of orifice plates used in the present study were obtained from Daniel Industries

and were manufactured to NOVA specifications. namely, a bevel angle of 37.5 degrees and a

bore edge thickness of 0.06in among others. Other variations of orifice plates that were tested

are tabulated in Table 2.

The flow rate from the orifice meter was compared to that from the sonic nozzle bank.

Details of the uncertainty analysis and the traceability of the nozzle bank to the gravimetric

4

•

•

•

•

•facility [13). [14) indicate that the uncertainty in mass flow is around 0.2%. The maximum flow rate

capacity of the facility is currently around 14kg/s.

3.0 EXPERIMENTAL RESULTS

3.1 Beveled Plates

Results of testing beveled plates (0.12Sin. and 0.2Sin. thick), manufactured to A.G.A. and

NOVA specifications (bevel angle =37.Sdeg. and bore edge thickness = 0.06in.), in an old fitting

are shown in Figures 4a,b,c,d,e,& f. The data was taken such that, for any given J3-ratio,the

• 0.12Sin. and the 0.2Sin. plate were tested on the same day thus improving the system

repeatability. The first observation from these figures is that measurements taken at two tap

locations 180 deg. apart show identical trends. This indicates that the velocity profile at the

orifice plate is symmetric. Further, for J3 ~ 0.6, the metering error is within the ±O.SS%band of

the R-G equation.

The statistics presented in Table 3. (for ~P>20kPa) may imply that the 0.2Sin. plate over-

registers in comparison to the 0.12Sin. plate. However, on examining the figures and the 20'

levels of the data it can be concluded that, for J3 2: 0.4, there is no difference between the

performance of the two plates. For 13=0.2 and 0.3, the thicker plate clearly seems to over-

register in comparison to the thinner plate by about 0.3-0.4%. Using the R-G equation with the

appropriate tap distances (0.937Sin.) does not eliminate this difference.• In the past, the effect of plate thickness on orifice meter performance has been studied

[15) in detail for a 6in. meter run. Although in their plate thickness tests the location of the taps

with respect to the plate were constant (1inch) for all plates, some comparisons are made with

the present results wherein the tap locations changed with plate thickness. The findings in the

present study are similar to earlier findings [15). For example, for plate thickness of 0.2Sin., they

find that a 13=0.3over registers a 0.12Sin. thick plate by around 0.3%. Further, for the larger J3-

ratios (O.Sand 0.7) no significant differences were measured as found in the present study.

It is also clear from Figures 4a. through 4f., that a 0.25inch thick plate performs

identically when used in an old fitting and a new fitting.

•S

•3.2 Non-Beveled Plates

Results of testing the non-beveled plates in an old fitting, are presented in Figures

5a,b,c,d,e &f and tabulated in Table 4. For all I3-ratios, there does not appear to be a significant

or systematic difference between the performance of the 0.125in. and the 0.25in,. thick plates.

This is evident from the Figures as well as the statistics shown in Table 4. In the case of the

non-beveled plates, previous findings (15) also indicate that there is no difference in the

performance of a 0.25in. plate in comparison to the 0.125in. plate, for all I3-ratios. Husain &

Teyssandier (15) further state that for thickness upto 0.3in., within their system repeatability, there

is no difference between the beveled and unbeveled plates. This is true in the present

experiments for the 0.125in. plate as seen in Table 5a. where no definite trend seems to exist •

and the differences are comparable to the 2cr confidence levels. However, from Table 5b, it

appears that the non-beveled plate under-registers in comparison to the beveled plates for a

plate thickness of 0.25in. particularly at the lower I3-ratios.

3.3 Effect of Bevel Angle and Bore Edge Thickness

This study was conducted for I3-ratios of 0.2192 and 0.5952, so that two extreme, yet

stable cases could be examined. Further, these tests have only been conducted with a 0.25in.

thick plate because the effects are expected to become severe with increasing thickness.

Specific details of the plates are provided in Table 2. Results of these tests are shown in

Figures 6 & 7 and Table 6. Firstly, consider the data taken for the plates with ex= 37.5deg. and

e=0.06in. Results of two plates are shown for each I3-ratio. The two plates were ordered and

tested almost 1 month apart. The data is very repeatable with deviations not greater than 0.1%. •

From Figures 6 & 7 and Table 6, it could be concluded that, for both orifice fittings, there is no

effect of bevel angle and bore edge thickness, within the specifications provided by the

standards. This is confirmed by the experiments of Husain & Teyssandier (15), except for 13=0.3

where they measured a difference of around 0.4% between ex=30deg.and 45deg.

4.0 ANALYSIS AND DISCUSSION

Various theories can be postulated to explain the performance of an orifice plate.

A.lthough more information is needed, the following attempts to propose some theories

explaining some of the results that have been obtained in the present study. The performance of

•6

•

•

•

•

the orifice plate, as reflected by the pressure drop measured across the plate, is influenced by

the following parameters :

(a) tap location with respect to the leading edge of the orifice plate

(b) flow structure generated by the orifice plate

(c) I3-ratio

Consider the performance of the O.12Sin.and O.2Sin.beveled plates in the old fitting. It

is possible that the smaller I3-ratios result in a well defined jet and hence a region of pressure

(tap location) sensitivity. If the location of the vena-contracta is determined by the leading edge,

then for the smaller I3-ratios, if a thicker plate is used, then the vena-contracta occurs closer to

the plate. If the tap location is fixed, then the thicker plate sees a higher differential pressure.

Thus, at smaller I3-ratiosthe thicker plate over-registers in comparison to the thinner plate for a

fixed tap location as indicated by the measurements. On the other hand a larger I3-ratio could

result in a rather large wake causing an insensitivity to tap location.

Why then does the O.2Sin.plate not show tap sensitivity at lower I3-ratios (i.e. in the old

fitting and the new fitting) ? The answer possibly lies in the fact that the flow structure is similar

i.e. "wake like" so that the pressure downstream is not sensitive to small changes in tap location.

The same can be said with the O.2Sin. thick plates used in the test of bevel angle and orifice

edge bore edge thickness tests.

Consider the un-bevelled plates, the results indicate that there is no sensitivity to tap

location i.e. a O.12Sin.plate and a O.2Sin.plate behave identically in the old fitting for all I3-ratios

as seen in Table 4. In the absence of a bevel, the knife-edge of the orifice plate is absent and

this results in a "bluff body" like wake which camouflages any pressure sensitivity that may be

associated with varying I3-ratios.

There are some exceptions that do not conform to the above theories. For example, the

O.2Sin.thick plate with the edge thickness of O.16in. Should its perfonnance be similar to that of

the O.12Sin.and O.2Sin.unbevelled plate? This is not the case in the present experiments.

The fact remains that more than just the tap location is responsible for the performance

exhibited by an orifice plate. Although, some ideas have been proposed, there are exceptions

to the hypotheses. Further, studies involving the study of the structure of the flow just

downstream of an orifice plate for various geometries using flow visualization or computational

7

d) plate thickness (0.125in. vIs 0.2Sin.) does not affect the performance of non-beveled plates

in the existing fitting.•

•techniques are warranted. These studies need to be supported with laboratory experiments

where the relevant parameters involved can be artificially manipulated to isolate various effects.

In any event, comparing the performance of the plates (0.125in. and 0.2Sin.), using

Figures 8a and 8b, one could recommend the retrofitting of existing meter runs with O.2Sin.thick

plates or using them in new fittings. In fact, this would not only increase the capacity but also

improve the repeatability of the meter.

5.0 CONCLUSIONS

Based on the results obtained with older fittings designed for the 0.12Sin. plates the •

following can be concluded :

a} O.25in. thick orifice plates perform identically when used in existing fittings or in new fittings.

The use of 0.2Sin. thick orifice plates would not only increase the capacity of the meter run

but also result in repeatable metering independent of J3-ratio. For 0.3,;; 13 ,;; 0.6, (i.e. for

mass flow> 3kg/s), the average error is around 0.32% with a 20 of ±0.16%.

b) in existing fittings, for I3-ratios ~ 0.4, there is no difference between the 0.2Sin. and the

0.125in. plates. For I3-ratios < 0.4, the 0.2Sin. thick plates would over-register by around

0.3%.

c) bevel angle (300,;; exs4So) and orifice bore edge thickness (O.6in.,;; e s 0.16in.), do not have

a measurable effect on the metering accuracy in new or existing fittings.

e} data indicates that for the 0.12Sin. thick plate, the difference between the beveled and non-

beveled plates depends on the J3-ratioand does not indicate a specific trend. In the case of

the 0.2Sin. thick plates, the non-beveled plates tend to under-register in comparison to the

beveled plates. The larger under-registration occurring at I3-ratios< 0.4.

•8

Marks Standard handbook for Mechanical Engineers, Mechanics of Materials, 9th edition,McGraw Hill. pg. 5-52.Jepson, P. and Chipchase, R. 1975 "Effect of plate buckling on orifice meter accuracy",Journal Mechanical Engineering Science.Norman, R., 1982, "A further investigation into the effect of orifice plate deformation onmetering error". Report No. ERS R2554, British Gas Engineering Research Station.

Norman, R. , Rawat, M.S. and Jepson, P, 1983 "Buckling and eccentricity effects onorifice metering accuracy", International Gas Research Conference.Norman, R. and Rawat, M.S 1983 "An experimental investigation into the flowmeasurement errors caused by elastic deformation of orifice plates", Re[port No. ERSR2713, British Engineering Research Station.

6. Norman, R. , Rawat, M.S. and Jepson, P, 1984 "An experimental investigation into theeffects of plate eccentricity and elastic deformation on orifice meter accuracy". Presented atthe International Conference on metering of natural gas and liquefied hydrocarbon gases.

7. Gorter, J. 1977 "Buckling of orifice plates and resulting errors in flow measurements bymeans of orifice meter runs", Flow Con 77, Proceedings of the Symposium "The Applicationof Measuring Techniques", Brighton.

8. Gorter, J. 1978 "Deformation of orifice plates, Theory and Practice, Flow Measurement ofFluids, Gromingen.

9. Mason, D., Wilson, M.P., and Birkhead, W.G. 1975 "Measurement error due to the bendingof orifice plates". Presented at the ASME Annual Meeting, Houston.

10. Kamik, U. and Williamson, I 1992 "High Differential Orifice Metering" Reports to NGTLMay, Sept. and Dec. 1992.

11. Kamik, U. et al. 1995 "Meter Prover Traceability", December report to NGTL.12. Morrow, T. 1995 "Orifice meter installation effects in the GRI MRF". 3rd International

Symposium on Fluid Flow Measurement, San Antonio, USA.13.Williamson, 1.0., Sawchuk, B.D. and Kamik, U. ,1995, "The NOVA Meter Prover", 3rd

Intemational Symposium on Fluid Flow Measurement, San Antonio, USA.

14. Kamik, U. 1995 "A Compact Orifice Meter/Flow Conditioner Package" 3rd InternationalSymposium on Fluid Flow Measurement, San Antonio, USA.

15. Husain, Z. and Teyssandier, R. 1986 "The effects of plate thickness and bevel angle in a150mm line size orifice meter" International Conference on Flow Measurement in the Mid-

• 80's, NEL.

•

7.0

1.

• 2.

3.

4.

5.

•

6.0 ACKNOWLEDGMENT

The author would like to acknowledge the contributions of Shane Fry, John Geerligs,

Russ Given and Dr. Wojciech Studzinski of NRTC; Blaine Sawchuk of NGTL; and the support of

Fred Knoll of Daniel Industries.

REFERENCES

9

•Calibration of Differential Transmitter at Orifice Meter

First Calibrated on 27th September 1995.

Date SloDe 0/0 difference

27 Sept 1995 0.123225 countslPa 0.000%

27 Oct 1995 0.12317 counts/Pa ·0.045%

29 Nov. 1995 0.123164 counts/Pa ·0.050%

Calibration Checks of Static Pressure Transmitter at Orifice Meter at zero flow. •Calibrated on 27th September 1995.

Table 1. Typical performance of pressure and temperature transmitters. •

Date Pnoz Porta max. diff.

kPa kPa %

27 Seot. 1995 6257 6259 0.03%

17 Oct. 1995 5790 5787 0.05%

31 Oct. 1995 5255 5253 0.04%

06 Nov. 1995 6166 6165 0.02%

21 Nov. 1995 6031 6029 0.03%

Calibration Checks of Temperature Transmitter at Orifice Meter with dry block. •

Calibrated on 27th September 1995.

Date Tre, TortS

dea.C deg.C

07 Nov. 1995 14.69 14.69

9.18 9.19

29 Nov. 1995 14.42 14.37

8.42 8.42

10

11

,,

•

•

J3-Ratio bevel angle Plate thickness Bore edge thicknessadea. inches inches

0.2193 37.5 0.125 0.06

37.5 0.250 0.06

37.5 0.250 0.16

30.0 0.250 0.06

30.0 0.250 0.16

45.0 0.250 0.06

45.0 0.250 0.16

0.3132 37.5 0.125 0.06

37.5 0.250 0.06

0.4072 37.5 0.125 0.06

37.5 0.250 0.06

0.5012 37.5 0.125 0.06

37.5 0.250 0.06

0.5952 37.5 0.125 0.06

37.5 0.250 0.06

37.5 0.250 0.16

30.0 0.250 0.06 -

30.0 0.250 0.16

45.0 0.250 0.06

45.0 0.250 0.16

0.7205 37.5 0.125 0.06

37.5 0.250 0.06•

Table 2. Details of orifice plates used In the present study.

•

•Mean Error '%1 with 20' Difference New Fitting

B-Ratlo O.12Sinch O.2Slnch O.2Sinch

0.2193 -0.29 ±O.18% O.08±O.11% -0.37% 0.11+0.13%

0.3132 0.09±O.07% 0.36±O.09% -0.27% .0.36±O.13%

0.4072 0.13±O.1S% O.25±O.08% -0.12% 0.28±0.11%

0.5012 0.17±O.08% 0.25±0.06% -0.08% 0.23±0.16%

0.5952 0.34±O.1S% 0.35±O.07% -0.01% 0.36±0.08%

0.7205 0.69±O.28% 0.79±O.32% -0.1% 0.60±O.02% •Table 3. Summary of results for beveled plates.

.

Mean Error [%] with 2(1 Difference

B-Ratio O.1251nch O.25inch

0.2193 -0. 1O±O.17% -0.18±O.15% 0.08%

0.3132 -0.045+0.09% 0.00tO.08% -0.045%

0.4072 0.20±O.11% 0.09±O.01% 0.11%

0.5012 0.14±D.08% 0.15±O.05% -0.01%

0.5952 0.11±0.21% 0.21±O.12% -0.10%

0.7205 0.55±0.O8% 0.57±O.05% -0.02%

•

Table 4. Summary of results for non-beveled plates in old fitting

•12

•

•

•

•

Mean Error f%] wHh 20" Difference

a-Ratio Beveled Non-Beveled Beveled minus Non Beveled

0.2193 -0.29±O.18% -O.10±0.17% -0.19%

0.3132 0.09±O.07% -0.045±O.09% 0.14%

0.4072 0.13±O.15% 0.20±0.11% -0.07%

0.5012 0.17±O.08% 0.14±O.08% 0.03%

0.5952 0.34±O.15% 0.11±O.21% 0.23%

0.7205 0.69±O.28% 0.55±O.08% 0.14%

Table 5a. Beveled vis non-beveled for 0.125 In. thick plates.

Mean Error [%) with 20" Difference

p-Ratio Beveled Non-Beveled Beveled minus Non Beveled

0.2193 0.08±0.11% -0.18±O.15% 0.26%

0.3132 0.36±0.09% 0.00±O.08% 0.36%

0.4072 0.25±O.08% 0.09±O.01% 0.16%

0.5012 0.25±0.06% 0.15±O.05% 0.10%

0.5952 0.35±0.07% 0.21±O.12% 0.14%

0.7205 0.79±O.32% 0.57±O.05% 0.22%

Table 5b. Beveled vis non-beveled for 0.25 in. thick plates.

13

Bevel Angle orifice Error±2cr Error±2a

(deg.) edge (new Fitting) (old Fitting)(In.)

45 0.06 0.19±O.D4 0.10±0.22

37.5 0.06 0.11±O.13 0.08±O.10

30.0 0.06 0.26±0.05 0.22±O.10

0.20% 0.13%

45 0.16 0.02±D.12 0.10±O.11

37.5 0.16 0.12±D.09 0.10±0.13

30.0 0.16 -0.01±O.06 -0.02±O.09

0.04% 0.06%

0.0 0.25 -0.18±O.15

•

•Bevel Angle orifice Error±2a Error±2a

(deg.) edge (new Fitting) (old Fitting)/In.)

45 0.06 0.35±O.08 0.41±O.17

37.5 0.06 0.36±O.O8 0.37±O.11

30.0 0.06 0.39±O.08 0.48±0.11

0.37"10 0.42%

45 0.16 0.24±O.D4 0.40±O.08

37.5 0.16 0.23±O.05 0.52±O.06

30.0 0.16 0.19±O.00 0.40±0.12

0.22% 0.44%

0.0 0.25 0.21±O.12

•Table Sa. Effect of bevel angle and orifice bore edge thickness (p-Ratio = 0.6) using a

0.25in. th ick orifice plate.

Table 6b. Effect of bevel angle and orifice bore edge thickness (\3-Ratio = 0.2) using a

0.25in. thick orifice plate. •14

Figure 1. Allowable pressure difference across an orifice plate before yield .

·e

250 I- !\· ,!\· \~:. ! \, • 1•~i, ~ \, . ~, ..., '\· '\, '\· ,· ,· ,'\ '. \. ,

\ . ':.

\ \. \\ -, '\\ \ ., .. ., .. ....'" .. ~..•.........

.... .... ...., <; :~.~.:>-::"::::---

- - - :.:.:-~~o~~~~~~I~~~~~I~~==~~~,-~-~-~-~-~-~-~~-~~--~-.o 50 100

- {J=O.2---- {J=O.4._----- {J=O~6- {J=O.7200 -

e........ca..-'it.'-'"':a 150 I-

100.1-

50 I-

150

Dl/t

e

•

:

-

-

-

-

200

0.04 r-w-r-r-..-,.,..-r-.,........, ....,.-..,......,...-.--..-,.'""T"-r-,.........-:.,......,..-r-.-,-{l=O.2 . - - {l=O.4 D/t=35.25·•••••{l=O.6 -{l=O.7

-0.02 t- . •t-

-0.04- ...L --1 --1 I

0 50 100 150 200 250~P(kPa)

o .20 r-.-.,......,.....,......:..... ,......,......,....-,-r-".-,.....,.....,....,r-...,....~ ........,"""'-"'''''''--P=O.2 ._- .(1=O.4 D

1t/=51.5

"''''P=O.6 -P=O.7

0.02 r

0.00 I-

0.15 r

E<,E 0.10-'0

0.05 I-

•2(a) -

.

•2(b) .

, ., -

'.-". --

, -'.'

....'..'.,»:.......-....'" -'-,--.»:<-

-'."--..'..'..-

?# ..-

~

t>P(kPo) •

'1

!

e0•.30 • I I • •

-/1=0.2 .•• /1=0.4 D.lt=67.5000

······/1=0.6 -(l=O.?0.25 - -

2(c)0.20 ,, ·~ ,, ~

~ ,., .../"~ ,,,13 ,, -:0' 0.15 I-

, ..'13 , ,, -, ,-.:I ,,

. , ,,,..0.10 / "

I- , .' ·'.-,'.'"..e ','.

0.05 I- -

0.00 I • • I ,a 25 50 75 100 125 150

4P(kPa)

0.5 , , I I °

• {J=O.2 . - - .{l=O.4 D. t/=B5.5.•.•..{J=O.6 -(J=O.7

0.4--

2(d)

e. ~~ ..~ 0.3 I-

,, ·e ..<, / ,E /

. .0.0. .0'0 / .

/ . .'

0.2 I- / .. ...•...., . ..'

/ . .'.0'.0

o·

001.

I-

0.0 I I , I

a 25 50 75 100 125

• Figure 2. Total error in metering due to elastic deformation of an orifice plate.

FLOW

--<>

ORFICE t.ETER RUN

DETAIL 'A'

_!,~?-r;~R~__~ _ NGRAVIMETRIC

1FACILITY <,

~O IAUTO-ADJUST

----i -~ r-- \a"TIJRBINE METER SEE DETAIL'''' ULTRASONIC METERS

NOZZLE a"ORIFICE ~----------- ~

BANK.\

METER'-- ----------------- - -- NPS8 'I~

METERBUILDING FLOW

" , "i- ~

Figure 3. Schematic layout of the Test Facility

•

•

•

•

• Figure 4a&b Comparison of performance of beveled orifice plates (a) p=O.2193(b) p=O.3132.

~.,

•1.0 .' , , , ,

• O.125in. In old filling Beta c 0.219271

0.8 f- .., O.125in.1n old filling (_180 dog) -0 O.2SIn.1n old tillingV O.25In.1n old filling (laps at 18Oo:1og.) 4(a)0.6 0 O.25in. in new tilting -

0.4 -~~ 02

g~E~

~8~:=;. 0.0 (l~

E 0•" -0.2 •0• §. •-0.4 .., ..,-0.6 ••-0.8

-1.0 :f' I I I

0 20 40 60 80 100 120 140 160IlP (kPa)

1.0 , , , .

4(b) Beta = 0.3132440.8 f-

0.6 f- -!]l

0.4 t- s Il!I ~0 c-:::..) i

~ I;J -' ,• ~ 0.2 r~ 0 1.. 'J.f .., , • jEe 0.0 r.., 1t -02 rl 10 'J§. j

-0.4 r --0 O.2Sin. New Fitting

--0.6 r ~• 0.125in. in old fitting.., 0.125in. in old fitting (taps180 deg) c-0.8 f- 0 0.2Sin. in old fitting

'V 0.2Sin. in old fitting (taps at 18Odeg.)

-1.0 .1 " , I ,I·

0 20 40 60 80 100 120 140 160 180 200t>P (kPa)

•1.0 • • 'T I 'T

4(C)Beta = 0.4072t7

0.8 :.. -

0.6 -

0.4 i,J ~

-~

~.gS 9

~"! 02 - • -E e • •- -'""i 0.0 :.. .;- -E•" -0.2 -0E •~

-0.4 ~ --0.6 0 025in.1n new titling -• 0.1251n.1n old titling

T 0.125111. In old titling (tapsl80 dog)

-0.8 - 0 025in.1n old filling -V 0251n.1n old filling (laps at 18Odeg.)

-1.0 I I

0 20 40 60 80 100 120 140 160 180llP (kPa)

1.0 , I I I •,

I-4(d) Beta = 0.50119

0.8

0.6 -

0.4 ~ 1

0 -t

~g

tlll 1 •e.. B .'f} B 0 j

0.2 ~ ~ tf -1e g B 1..§ ; J

'""'- 0.0 lE~. •"t -0.2 - -0 •.s -

-0.4 .,

-0.6 ~ 0 O.2Sin. in New Fitting

• O.125in. in old fittingT O.125in. in old fitting (taps180 deg)

-0.8 0 0.2Sin. in old frtting

\l 0.25in. in old fitling (taps at 180deg.)

-1.0 I, ,

0 10 20 30 40 50 60 70 80f>P (kPa)

Figure 4c&d Comparison of performance of beveled orifice plates (c) 13=0.4012

(d) 1l=0.5012. •

1.0 rr~~.,...,.,~"""",C"T"""~~,-,-,~~"""'~-"C"T"""'-'-rr-,,-,-,,,,,,4(e) Beta ~ 0.59516 _

- 0 025 in. : New Filling-0.6 - • O.125in. in okl fitting

" O.I25in. in old fitting (tapsl80 dog)-0.8 0 025in. in old fitting

\l O.25m. in old fitting (taps at 180deg.)-1.0 w...~.......L~~..J'L..-~~.L.....,~~.L. '~.....JW-~~L.o...'~..0....1

5 10 15 20 25 30 35

1.4 ·~~~~~~~~~~I~~~~~·~ ,~~~~

4(f) Beta = 0.7204611.2 F- -

•0.8 -

0.6 !-

0.4 ~

0.2

0.0," -02{•

1.0

0.8 F-

• ::0' 06~ ...E.-0.4 F--->.

e" 0.2IioS 0.0

-0.2 F-

-0.4

-0.6

-

SI'Y B••

-Q ~•• -

•-

-

40

~(kPa)

o()8 o •8

Q .8 •I5 I·oo ~. j

<1J

I.V.• V

o O.2Sin. in new fitting

• 0.125in. in old fitting... O.12Sin.in old fitting (taps180deg)o 0.2Sin. in old fitting

"V 0.2Sin. in old fitting (taps at 180deg.,-0.8 L.....~...L~..o..l.~~~~...L.~.....L~'--'-'Lo- ....:.

7

Figure 4e&f Comparison of performance of beveled orifice plates (e) P=O.5952• (f) P=O.7205.

8 9 10 11

~P (kPa)

12 13 14 15

•1.0 : I 'T

......-eIIed plates

0.8 I- Beta • 0.219271

0.4 I-

•

, I ' I

• O.1ZSIn. In old fillingo 0251n. In old lilting .

0.6

-

S(a) .

02 I- bevelled O.251n, 2..-0.11.~-----------------------0.0 o

Qeo-02 •• -

-O_~I- •beveUed 0.125n.; 2u=O.18. -

-OJ. I-

-4LS I-

-1_0 W-.. ...........-L...1~...L ...........----L~c......L~.......JI ...........~L.o..".~~'~o 20 40 60 SO 100 120 140 160

~p(kPa)

norH>eveUed plates0.8 I- aeta=O.313244 5(b)

• O.125in. in old fitting

o 0.25in. in old fitting -

0.6 I- -

o 4 l- beveUed O.2Sin.: 20=0.09. J.~----------------------l0.2 I- beveDed O.12Sin.; 20=0.07. l

o i,-'

•80.0 I- 8• ••

-e -0.2 f-E'-0.4 I-

-0,6 I-

-0,8 I-

40 60 80 100 120 140 160

",p (kPa)

Figure 5a&b Comparison of performance of non-beveled orifice plates in oldfitting (a) 13=0_2193 (b) 13=0.3132_ •

j

•

•

•

1.0 ~~~,~~~~~~,~.~ ,~~~~~~~,~~"O"-bevelled plates S(c) • 0.' 2S1n. in old lilting

0.8 I- Be.. a 0.407217 0 025in. in old fitting -

0.6 I-

u~ -bevelled 0.25in.; 2...0.08.

02 r-.--....-- ..-------:---------:.- _eI1edO.'2SIn.;2""O.'~ 0 0

~~ -

-0.2

-0.4

-0.6 :-

-0.8 I-

-1.0 , ,0 20 40

-

-

-

, , ,60 80 100 120 140 160

6P (kPa)

1.0,......,~..,...~.,.,~..,...,..~,..,.,..,.....,...,...,......,~ -r-r,""""''''''''~'''''''non-bevelled plates 5(d) • 0.12Sin. in old fitting

0.8 r- Beta = 0.50119 0 0.2Sin. in old rilling -

0.6 I- -

0.4 I- -bevelled O.25in.; 2ct=O.06%.~-----------------------0.2 ~

• -.- - ~ bevelled 0.1"25in.: 2~=O.08c..{,. jl1-1J~J

0.0 I-,t -0.2 r-EO

-0.4

-0.6

-0.8 ..:;J

, . , ,-1.0

0 10 20 30 40 50 60 70 806P (kPa)

• Figure 5c&d Comparison of performance of non-beveled orifice plates in oldfitting (c) P=O.4072(d) p=O.5012.

•1.0 n....-:.lIed~ I ,

• 0.1251n.1nold titling

O.B Beta a 0.69616 5(e) 0 02510. In old fitting -

0.6 - -

0.4 ~b.... 1Ied O.25In~2....0.07... -~ bevelled 0.12~ 2...0.15,.. ee.... 0E'!

0.2 0

I • 0~ 0.0 -E

1i -0.2 f- • •g-0.4 f- • --0.6 f- .

-O.B f- -

-1.0 I I

a 5 10 15 20 25 30 35l>P (kPa)

1.4 T , Inon-bevelled plates

S(t) •O.12Sin. in old fitting

1.2 f- Beta = 0.720461 0 D.2Sin. in old fitting ..:

1.0 f- -

;f 0.6 F- bevelled O.125in.;2a=028%. o c: 0 j •~ •.. aE 0.4, 0E 0.2 f- gt ~0g 0.0

~-0.2 f-

-0.4 ,j,-iJ

-0.6 -'

-0.8 . I 14 6 8 10 12 14 16

6P (kPa)

Figure 5e&f Comparison of performance af non-beveled orifice plates in oldfitting (e) P=O.5952 (fl P=O.7205. •

O.25in. thick plates in old fitting

0.8 - B.ta = 0.595166(b)

•1.0 , , , , ,

0 aa37.5deg.: 0-0.0&. V a-45deg.: .. 0.16.

0.8 ~ aa37.5deg.: 0-0.06 • a-45deg.: .. 0.06. -(,__ another ""'''')

0 aa37.Sdeg.; e-O.16. • a-3Odeg.; e-O.06.

0.6 t:>. aa3Odeg.: .. 0.16. -- ""_",,,"':2-0.15%

0.4 -~ I Q;-2.- 02.. •E ,

~--:.. 0.0 -eE," -0.2 -a• .s

-0.4

-0.66(a)

-0.8 :- B.ta = 0.219271

0.25in. thick pia"" in old filling

-1.0 , , , ,20 40 60 80 100 120 140 160

toP (kPa)

0.6 -

0.4 -~• e; 0.2 -..§~ 0.0 ~E

" -0.2a.§.

-0.4

-0.6

-0.8 ~

-1.0 ~0

-

LJ

o 1 -3j,-tj

~j

..J

1'V a=45deg' e=O16 J- un bevelled plate: 20--0.12-"

o a=37.5deg.: e=O.06.

... a=37.5d.g.; ee 0.06(repeat with another plate)

o a=37.5deg.: e=O.16... ! , , , , . I, , t , ! , t. , !

,a.=3Odeg.; e=O.06.-:

c: a.=30deg.; e=O.16. i,,

• a=45deg.: e=D,D6. :

•! , . , .

5 10 15 20toP (kPa)

25 30 35 40

Figure 6. Effect of bevel angle and orfice bore edge thickness on meterperformance; 0.25in. thick plates; (a) [3=0.2193(b) [3=0.5952.•

•1.0 , , "T I I 1

o ..-37.5deg.; e-O.06 (standard) • ..-30deg.; .-0.06 .0.8 0 u-37.5deg.; -0.16. ~ ...3Odeg.; .-0.16 .•

V ..-45deg.; ... 0.16.

O.S • ..-45deg.; e-O.06. -

0.4 I- -~ •~ 02

~

• e ,-"!!g~E- 6.!fJ _l! 0.0 =sE

15 -0.2 - -E~ •-0.4 - -

-0.6 - -7(a)

-0.8 - Beta = 0.219271

-1.0O.25in. ~ plate~in new filling , I

20 40 SO 80 100 120 140 160

6P (kPa)

1.0 I I I . I

O.2Sin. thiek plates in new fitting 7(b)0.8 f- Beta = 0.59516

0.6 l- S0.4 I- ~ I .

8 fit~~ 0.2 I- !it tl ~ . •;; {'<,E::=:..~ 0.0 - -E~ -0.2 -0.s i-0.4

a~4Sd.g.: .=0.16. ~-0.6 0 a=37.5deg.: e=O.06.'V

0 a=37.5deg.; &=0.16. • a=4Sdeg.: I!o=O.06. ~

-0.8 • a=30deg.: &=0.06. ~c: a~3Odeg.: ~=O.1S'1

-1.0 , ,0 5 10 15 20 25 30 35 40

lloP(kPa)

Figure 7. Effect of bevel angle and orfice bore edge thickness on new meterperformance; 0.2Sin. thick plates; (a) P=0.2193 (b) P=0.5952. •