Influence of Flow Disturbances on Measurement Uncertainty ... · Typical flow disturbances were selected and agreed Two metering skids were designed, one for the 2” meters and the

Influence of Flow Disturbances on Measurement Uncertainty of Industry-Standard LNG Flow Meters- Highlights

Asaad Kenbar (NEL) & Menne Schakel (VSL)

LNGIII Workshop, 28 May 2020

Main objectives of this work

28/5/2020 LNGIII workshop 2

▪ Meet a target measurement uncertainty of 0.5% (k = 2) for small and mid-

scale LNG applications. Can we achieve this in practice?

▪ Investigate the effect of typical upstream flow disturbances on LNG flow

measurement uncertainty

▪ Investigate the effect of meter insulation on flow measurement uncertainty

▪ Assess transferability of meter calibrations with water at ambient conditions

to cryogenic conditions. Can we meet the 0.5% uncertainty above?

Experimental Programme


▪ A careful experimental programme was designed to achieve these objectives

▪ Six LNG flow meters were secured from 5 manufacturers (Endress + Hauser, Emerson,

Krohne, Yokogawa and Panametrics a Baker Hughes company) :

− 2 Coriolis (2” in size)− 2 Coriolis (4” in size) − 2 Ultrasonic (4” in size)

▪ Typical flow disturbances were selected and agreed

▪ Two metering skids were designed, one for the 2” meters and the other for the 4” meters

▪ These skids were tested with water at NEL first, then

▪ Tested with LIN in the VSL’s Cryogenic Research and Calibration Facility

LIN: Liquid nitrogen

Typical flow disturbances


Moderate From Kenbar & Schakel, submitted to FM&I, 2020

Severe

Metering configurations

LNGIII training 5

From Kenbar & Schakel, draft journal publication, 2020

Water calibrations at NEL

28/5/2020 6

Not displayed

Double bends

NEL’s Water Flow Measurement Facility


Water calibrations- setups and test matrix


▪ I. Ideal setup: no flow disturbance (open plate), no insulation

▪ II. As setup I, but with partially blocking plate

▪ III. As setup I, but with double-bend disturbance


Test matrix, 4” line

Cryogenic calibrations at VSL

8

VSL Cryogenic Research and Calibration Facility

2” metering line showing the double bend disturbance (1)

and part of its rockwool insulation (2)

From Schakel, 2019, VSL website


Cryogenic calibrations- setups and test matrix

28/5/2020 9

▪ I. Ideal setup: no flow disturbance and all meters are insulated

▪ II. As setup I, but with partially blocking plate

▪ III. As setup I, but with double-bend disturbance

▪ IV. As setup I, but with flow meter insulation removed


Calibration results

10

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0

Err

or

[%]

Mass Flow Rate [kg/s]

Water Flow TestsMeter 1

Ideal, 20 deg. C

Ideal, 36 deg. C

Partial blockage, 20 deg. C

Double-bend, 20 deg. C

-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0

Err

or

[%]


Liquid Nitrogen Flow TestsMeter 1

Ideal

Ideal-repeat

Partial blockage

Double-bend

Un-insulated

Nominal

Rate

(kg/s) Ideal, 20 °C

1 -0.018

2 -0.010

3 0.048

4 0.044

5 0.017

6 0.004

7 -0.013

8.5 -0.018

Nominal

Rate

(kg/s) Ideal

2 -0.19

4 -0.10

6 -0.07

8 0.00

10 -0.04

From: Schakel (2019) & Kenbar & Schakel, draft journal publication, 2020From: Kenbar & Schakel, draft journal publication, 2020

The tables show the

averaged error for

each test point for

the ideal case

Calibration results

11

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0

Err

or

[%]


Water Flow TestsMeter 6

Ideal, 20 deg. C

Ideal, 36 deg. C



Nominal

Rate


1 -0.026

2 -0.020

3 0.036

4 0.036

5 0.016

6 0.007

7 -0.006

8.5 -0.010

Nominal

Rate

(kg/s) Ideal

2 0.26

4 0.44

6 0.52

8 0.60

10 0.59


Calibration results

10-6-2020 12

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.0 5.0 10.0 15.0 20.0 25.0 30.0

Err

or

[%]


Water Flow Tests

Meter 0

Ideal, 20 deg. C

Ideal, 36 deg. C



-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 5 10 15 20

Err

or

[%]



Ideal

Ideal-repeat

Partial blockage

Double-bend

Un-insulated

Nominal

Rate


7 -0.041

16 -0.017

25 -0.018

35 -0.012

Nominal

Rate

(kg/s) Ideal

4 -0.07

8 0.02

12

16 0.13

20 0.10


Calibration results

13

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.0 5.0 10.0 15.0 20.0 25.0 30.0

Err

or

[%]


Water Flow Tests

Meter 3

Ideal, 20 deg. C

Ideal, 36 deg. C



-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 5 10 15 20

Err

or

[%]



Ideal

Ideal-repeat

Partial blockage

Double-bend

Un-insulated

Nominal

Rate

(kg/s)

Ideal,

20 °C

7 -0.107

12 -0.115

17 -0.123

22 -0.145

28 -0.137

Nominal

Rate

(kg/s) Ideal

4 -0.21

8 -0.20

12 -0.12

16 -0.12

20


Calibration results

14

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.0 5.0 10.0 15.0 20.0 25.0 30.0

Err

or

[%]


Water Flow Tests

Meter 4

Ideal, 20 deg. C

Ideal, 36 deg. C



-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 5 10 15 20

Err

or

[%]



Ideal

Ideal-repeat

Partial blockage

Double-bend

Un-insulated

Nominal

Rate

(kg/s)

Ideal,

20 °C

7 -0.155

12 -0.148

17 -0.118

22 -0.111

28 -0.135

Nominal

Rate

(kg/s) Ideal

4 0.67

8 0.43

12 0.38

16 0.27

20


Calibration results

15

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.0 5.0 10.0 15.0 20.0 25.0 30.0

Err

or

[%]


Water Flow Tests

Meter 8

Ideal, 36 deg. C



-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 5 10 15 20

Err

or

[%]



Ideal

Partial blockage

Double-bend

Un-insulated

Nominal

Rate

(kg/s)

Ideal,

36 °C

7 0.204

16 0.021

25 -0.035

Nominal

Rate

(kg/s) Ideal

4 0.29

8 -0.25

12

16 -0.22

20 -0.22

From: Schakel (2019) & Kenbar & Schakel, draft journal publication, 2020

From: Kenbar & Schakel, draft journal publication, 2020

Conclusions


▪ For water testing, the combined measurement uncertainty (facility + meter) varied

between 0.10% and 0.14% over the test flow range for all test cases.

▪ Therefore, the measurement accuracy of 0.10% to 0.15% specified by meter

manufacturers are generally met since the measured error was within its accuracy

band.

▪ The corrections used to transfer the water calibration to LIN conditions resulted in meter

errors, for the ideal case, within ±0.50% for about 85% of the results. However, it is

important to note that the correction models are specific to the meter type and

manufacturer.

▪ Also, the target measurement uncertainty of 0.50% can be achieved if the expanded

standard error of the mean value measured by the meter is smaller than 0.40%.

Conclusions continued


▪ The influence of flow disturbances is limited (meter error below ±0.50%), however,

these disturbances were followed by straight piping equivalent to 20D. This conclusion

will be substantiated further when testing repeated with LNG.

▪ The influence of meter insulation can be more significant than the influence of the

tested disturbances (meter error > ±0.50%) .

▪ The results presented here cannot be extrapolated, with meaningfully low uncertainty,

to flow meters of different type, size, model and make (manufacturer) than those tested

▪ Many of the above conclusions will be substantiated further when testing is conducted

with LNG in the future

▪ Future work will include other effects such as 2-phase flow and meter orientation.

Acknowledgment


The participation and support received from the LNG flow meter

manufacturers is highly appreciated!

Thanks to:

▪Panametrics a Baker Hughes company,

▪Emerson,

▪Endress + Hauser,

▪KROHNE, and

▪Yokogawa

Documents

Influence of Flow Disturbances on Measurement Uncertainty ... · Typical flow disturbances were selected and agreed Two metering skids were designed, one for the 2” meters and the