A REVIEW OF THE REVISIONS TO API 14.3/AGA REPORT...

2001 PROCEEDINGSAMERICAN SCHOOL OF GAS MEASUREMENT TECHNOLOGY

A REVIEW OF THE REVISIONS TOAPI 14.3/AGA REPORT NO. 3 - PART 2

Paul J. LaNasaCPL & Associates

P.O. Box 801304, Houston, TX 77280-1304

ABSTRACT

Periodically natural gas measurement standards arecreated or revised. In the period 1993 through 1999 Part2 of ANSI 2530/API MPMS 14.3/AGA Report No 3underwent revision. It is the intent of this paper to discussthe highlights of this revision.

BACKGROUND

The dictionary defines standard as “somethingestablished by authority, custom, or general consent asa model or example”. In the natural gas industry,standards are established to provide uniformity of actionwith the hope this will improve efficiency and avoiddifferences. Natural gas measurement is one area ofapplication for standards in the natural gas industry.Periodically, natural gas measurement standards arecreated or revised. In the period of 1993 through 1999Part 2 of ANSI 2530/API MPMS 14.3/AGA Report No 3has been revised.

ORIFICE PLATE REVISIONS

Two major orifice plate areas of concern are addressedin the 2000 revision of Part 2:

• Maximum allowable differential pressure acrossthe orifice plate,

• Recommended 8-inch orifice plate thickness.

The maximum allowed differential pressure across theorifice plate is redefined in the new standard. The 1992edition established an explicit maximum differentialpressure of 200 inches of water column (iwc) for all orificeplates sizes except 8-inch. The 8-inch diameter by 1/8thick orifice plate was limited to 150 iwc. The reviseddifferential pressures limits will vary and are now afunction of:

• Plate thickness• Plate material mechanical strength• Acceptable expansion factor,Y, uncertainty.• Maximum acceptable permanent pressure loss.

This could result in differential pressures as high as 1000iwc. Except for the additional expansion factoruncertainty, the influence of high differential pressures isarchived in the Reader-Harris Gallagher coefficient ofdischarge calculation. The flange tapped coefficient of

discharge database experienced differential pressuresas high as 1300 iwc.

The recommended orifice plate thickness for 8-inchorifice plates in the 1992 revision is 1/8 of an inch. The8-inch diameter by 1/8 thick orifice plate was limited to150 iwc in order to maintain the flatness tolerances. The2000 revision has changed the recommended orificeplate thickness for 8-inch orifice plates to 1/4 of an inch.The change from 1/8 inch orifice plates to 1/4 inch orificeplates can be achieved in some orifice plate holdingdevices without any modifications to the device providedthe differential pressure tap hole tolerances and limitsas specified are satisfied. Otherwise the orifice plateholding device has to be redrilled

METER TUBE REVISIONS

Meter Tube Surface RoughnessProper meter tube surface roughness is required toassure accurate measurement. New research dataresulted in a revision in the surface roughnessrequirements. The new requirements establishes aminimum pipe wall roughness for all pipe sizes and allowspipes larger than 12-inch nominal pipe diameter to havea greater roughness than previous revisions.

The surface roughness requirement for 12- inch nominalpipe diameter and smaller is:

• 300 micro inches Ra when βr ≤ 0.6• 250 micro inches Ra when βr > 0.6• Minimum Ra > 0.34 micro inches

The surface roughness requirement for greater than 12-inch nominal pipe diameter is:

• 600 micro inches Ra when βr ≤ 0.6• 500 micro inches Ra when βr > 0.6• Minimum Ra > 0.34 micro inches

Irregularities such as grooves, scoring, or ridges resultingfrom seams, welding distortion, offsets, and the like thataffect the tolerance in excess of that allowed by theTolerances and Restrictions detailed in section 2.5.1.3of the revision are not permitted. The existence of pits inthe surface of the meter tube, although undesirable, areallowed provided their individual measurements do notexceed the surface roughness and/or diameter tolerancerequirements of the meter tube and do not compromisethe meter tube’s pressure integrity. When these

2001 PROCEEDINGS PAGE 185AMERICAN SCHOOL OF GAS MEASUREMENT TECHNOLOGY

tolerances are exceeded, the irregularities must becorrected. The increase in the meter tube surface wallroughness requirement should not be interpreted asreducing the need to maintain clean meter tubes. The2000 revision states: “Due care shall be exercised to keepthe meter tube interior clean and free from accumulationof dirt, ice, grit, grease, oil, free liquid and otherextraneous materials, to the extent feasible. Damageand/or accumulation of extraneous materials in the metertube may result in a greater uncertainty for the orificeplate coefficient of discharge, Cd(FT).”

Flow ConditionersThe 2000 revision provides two changes to the 1992 flowconditioning section:

• It defines a new 19 tube uniform concentric tubebundle flow straightener.

• It provides recognition and acceptance of theuse of other flow conditioners.

In an effort to eliminate or reduce the potential for flowmeasurement bias in existing installations and to provideguidance for improved measurement accuracy in newinstallations, the 2000 revision provides constructionrequirements and installation recommendation for the19 tube uniform concentric tube bundle flowstraighteners cited in the installation effects research.Due to the significant (outside the designated uncertaintyband) coefficient of discharge differences experiencedfrom variations in the 1992 revision’s straightening vanetube bundle construction, only those tube bundle flowstraighteners meeting the 2000 criteria are specified toproduce “no additional uncertainty” when installed asrecommended. All other tube bundles should beconsidered as other flow conditioners.

It is not the intent of the 2000 revision to recommendany particular type of flow conditioner. However, the 2000revision does recognize and accept the use of other typesof flow conditioners. It requires that their use be basedon technical performance data obtained fromperformance test(s). The 2000 revision provides a uniformcriterion for evaluation of installation and/or flowconditioner performance (perturbation) test or tests. Thistest(s) is required by the 2000 revision to confirm theperformance level that can be achieved by an orificemeter installation using a flow conditioner. Details of theperformance test are given in Appendix 2-C & D of therevision. The performance test(s) will confirm the orificemeter diameter ratio, b, meter tube length, and flowconditioner location for which acceptable performanceis obtainable.

Required Meter Tube LengthsThe most significant change in the 2000 revision hasbeen to the required orifice meter tube lengths. Therequired orifice meter tube upstream lengths, both withflow conditioner and without flow conditioner (bare), have

undergone significant revision. The required orifice metertube downstream lengths remain unchanged. Theupstream meter tube length revisions are the result ofextensive research, both national and international, thathas been underway since the mid 1980’s. The 1992revision indicated that the research was in progress andthat revisions would occur when the research wasconcluded. All of the known data developed byresearchers that met the acceptance criteria was utilized.

The research data compiled was carefully reviewed byan international group of experts from Canada, Japan,Western Europe and the United States. Only after thisrigorous study of the data were the new meter tubelengths developed.

The required orifice meter tube lengths are divided intotwo categories:

• Meter tubes lengths without 19 tube concentrictube bundles (Bare meter tubes).

• Meter tubes lengths with 19 tube concentric tubebundles.

Meter Tubes Without the Uniform Concentric 19 TubeBundle Flow Straightener (Bare meter tubes)The bare meter tube consists of only straight pipebetween the last pipe fitting and the orifice plate. Theroughness of the pipe is the only agent that aids in theproper flow profile development. Any configuration notexplicitly identified in the designated tables (See attachedTable for more detail), the required lengths and the 1998Uniform Concentric 19-Tube Bundle Straightenerlocations of the “any other configuration” classificationshould be followed.

Meter Tubes With Uniform Concentric 19 Tube BundleFlow StraightenerThe meter tube lengths utilizing the uniform concentric19 tube bundle flow straightener have been revised. Theresearch data segmented the meter tube lengths intothe following two categories:

• 17Di ≤ Upstream Length < 29Di• Upstream Length ≥ 29 Di

Where:Di - published internal pipe diameter

The recommended uniform concentric 19 tube bundleflow straightener location for 17Di ≤ Upstream Length <29Di is designated as 12-13 Di form the outlet of the flowstraightener to the upstream face of the orifice plate.The recommended uniform concentric 19 tube bundleflow straightener location for Upstream Length ≥ 29 Di isdesignated as 12-13 Di form the downstream end of theflow straightener to the upstream face of the orifice plate.(See attached table for more details)


ORIFICE PLATE ECCENTRICITY

The 2000 revision removes the minimum limit on the highb ratio small diameter meter tube eccentricity tolerance.The 1992 revision had placed the minimum eccentricitytolerance for 2, 3, and 4-inch high β ratio orifice platesat 0.020 inches.

PULSATION ENVIRONMENT

The 2000 revision establishes for the first time a minimumpulsation threshold below which pulsation effects onorifice meter measurement is negligible. Accurate flowmeasurement with an orifice meter can be ensured onlywhen the root mean square (rms) of the fluctuatingdifferential pressure divided by the average differentialpressure is less than or equal to 10%.

∆Prms/ ∆Pave ≤ 0.10

This limit applies to:

• Single frequency flow pulsation with or withoutharmonics such as those generated byreciprocating compressors or closed relief orblow down valves,

• Broad band flow pulsation or noise such asgenerated by throttling valves.

Pulsation environments greater than the 10% limit givenin the formula are not addressed.

Additionally, in order to avoid any resonance in the gaugeline, the length of the gauge line should be as short aspossible or should have lengths, l, specified accordingto the highest frequency, ƒ, of concern from one of thefollowing formulas:

0≤l1≤0.25a/(2πƒ)l2 = 2.5a/(2πƒ)l3 = 5.5a/(2πƒ)l4 = 8.5a/(2πƒ)l5 = 11.5a/(2πƒ)

Where :a - speed of sound in the flowing fluid at

operating conditionsƒ – frequency of pulsation levelsπ - mathematical constant = 3.14159

The length of the gauge line determined from any of theseformulas will ensure that no resonance and/oramplification of pressure pulsation exist in the gauge line.Both gauge lines should be of equal length and have nosudden changes of the internal diameter, especially forlow pressure applications. In some cases, direct mountmanifolds may reduce the effects of pulsation.

THERMOMETER WELLS

The thermometer wells must be located to accuratelysense the flowing temperature of the fluid beingmeasured. The wells may be placed on the downstreamside of the orifice at the outlet of the downstream metertube or no farther than four time the length of thedownstream meter tube. If a flow conditioner is utilized,the thermometer well may be located upstream side ofthe orifice but no closer than 36 inches upstream of theflow conditioner. Thermometer wells exposed to theinfluences of the ambient environment may result inbiased measurement.

Care should be taken to ensure that the temperaturesensor indicates the flowing gas temperature and is notthermally coupled to the meter run pipe. Insulation ofthe meter tube can be required if the ambient temperatureeffects the accuracy of the flowing temperaturemeasurement.

CONCLUSIONS

The 2000 revision of Part 2 of ANSI 2530/API MPMS14.3/AGA Report No. 3 represent the mostcomprehensive revision to the orifice meter’sspecification and installation requirements since 1955.Adherence to the 2000 revision’s design specificationand installation requirements will certainly improvemeasurement accuracy.


TA

BLE

- O

RIF

ICE

ME

TE

R IN

STA

LLA

TIO

N R

EQ

UIR

EM

EN

TS

WIT

HO

UT

A F

LOW

CO

ND

ITIO

NE

R

Min

imum

str

aigh

t un

obst

ruct

ed m

eter

tub

e le

ngth

from

the

up

stre

am a

nd d

owns

trea

m s

ide

of t

he o

rific

e p

late

(in

mul

tiple

s of

pub

lishe

d in

tern

al p

ipe

dia

met

er, D

i)

Dia

met

era.

Sin

gle

90°

Two

90°

elb

ows

inTw

o 90

° el

bow

s in

Two

90°

elb

ows

Two

90°

elb

ows

Sin

gle

90°

a.S

ingl

e 45

°G

ate

valv

eC

once

ntric

Any

oth

erD

owns

trea

mra

tioel

bow

.th

e sa

me

pla

ne “

S”

the

sam

e p

lane

, “S

”in

per

pen

dic

ular

in p

erp

end

icul

arTe

e us

edel

bow

.at

leas

tre

duc

erco

nfig

urat

ion

met

er t

ube

ßb

.Tw

o 90

°co

nfig

urat

ion

conf

igur

atio

np

lane

s,p

lane

s,as

b.

Two

45°

50%

op

en(c

atch

all

leng

thel

bow

s in

the

spac

er10

Di <

S ≤

30D

iS

< 5

Di*

5Di ≤

S ≤

15D

ian

elb

ow b

utel

bow

s in

the

cate

gory

)*sa

me

pla

neS

≤ 1

0Di

not

as a

sam

e p

lane

“S

”w

ith S

> 3

0Di

head

erco

nfig

urat

ion

c.Tw

o 90

°el

emen

tS

≥ 2

2Di

elb

ows

inp

erp

end

icul

arp

lane

s w

ithS

> 1

5Di

UL

UL

UL

UL

UL

UL

UL

UL

UL

UL

DL

≤0.2

06

1010

5019

930

176

702.

8

0.30

1110

1250

329

3019

610

83.

0

0.40

1610

1350

449

3021

614

53.

2

0.50

3030

1895

4419

3025

714

53.

5

0.60

4444

3095

4429

3030

914

53.

9

0.67

4444

4495

4436

4435

1114

54.

2

0.75

4444

4495

4444

4444

1314

54.

5

Rec

omm

end

ed44

4444

9544

4444

4413

145

4.5

len

gth

for

max

imum

ran

ge ß

≤ 0

.75

UL

-M

inim

um m

eter

tub

e le

ngth

up

stre

am o

f th

e or

ifice

pla

te,

UL,

in p

ublis

hed

inte

rnal

pip

e d

iam

eter

, D

i Str

aigh

t le

ngth

sha

ll b

e m

easu

red

fro

m t

hed

owns

trea

m e

nd o

f th

e cu

rved

por

tion

of t

he n

eare

st (

or o

nly)

elb

ow o

r of

the

tee

or

the

dow

nstr

eam

end

of

the

coni

cal p

ortio

n of

red

ucer

or

exp

and

er.

DL

-M

inim

um d

owns

trea

m m

eter

tub

e le

ngth

, DL,

in p

ublis

hed

inte

rnal

pip

e d

iam

eter

s, D

iS

-S

epar

atio

n d

ista

nce

bet

wee

n p

ipin

g el

emen

ts in

pub

lishe

d in

tern

al p

ipe

dia

met

er, D

i, m

easu

red

from

the

dow

nstr

eam

end

of t

he c

urve

d p

ortio

n of

the

upst

ream

elb

ow t

o th

e up

stre

am e

nd o

f the

cur

ved

por

tion

of t

he d

owns

trea

m e

lbow

.N

ote:

The

tol

eran

ce o

n sp

ecifi

ed le

ngth

s fo

r U

L an

d D

L is

±0.

25D

i.*

-Th

is in

stal

latio

n ex

hib

its s

tron

g ef

fect

of

Rey

nold

s nu

mb

er a

nd p

ipe

roug

hnes

s on

the

rec

omm

end

ed le

ngth

due

to

rate

of

dec

ay o

f sw

irl.

The

pre

sent

rec

omm

end

atio

ns h

ave

bee

n d

evel

oped

for

high

Rey

nold

s nu

mb

ers

and

sm

ooth

pip

es t

o ca

ptu

re t

he w

orst

cas

e.


TA

BLE

A -

OR

IFIC

E M

ET

ER

INS

TALL

AT

ION

RE

QU

IRE

ME

NT

S W

ITH

199

8 U

NIF

OR

M C

ON

CE

NT

RIC

19-

TU

BE

BU

ND

LEFL

OW

ST

RA

IGH

TE

NE

R F

OR

ME

TE

R T

UB

E U

PS

TR

EA

M L

EN

GT

H O

F 17

Di ≤

UL

< 2

9Di.

Dia

met

er R

atio

,S

ingl

e 90

° el

bow

Two

90°

elb

ows

out

ofS

ingl

e 90

° Te

eP

artia

lly c

lose

d v

alve

sH

igh

swirl

com

bin

edA

ny fi

ttin

gD

owns

trea

m m

eter

ßR

/Di =

1.5

pla

ne S

≤ 2

Di

used

as

an(a

t le

ast

50%

op

en)

with

sin

gle

90°

Tee

(cat

ch a

ll ca

tego

ry)

tub

e le

ngth

R/D

i = 1

.5el

bow

but

not

as

a he

ader

ele

men

t

UL2

UL2

UL2

UL2

UL2

UL2

DL

0.10

5-14

.55-

14.5

5-14

.55-

115-

135-

11.5

2.8

0.20

5-14

.55-

14.5

5-14

.55-

115-

135-

11.5

2.8

0.30

5-14

.55-

14.5

5-14

.55-

115-

135-

11.5

3.0

0.40

5-14

.55-

14.5

5-14

.55-

115-

135-

11.5

3.2

0.50

11.5

-14.

59.

5-14

.511

-13

**11

-13

***

3.5

0.60

12-1

313

.5-1

4.5

*N

ot a

llow

ed*

Not

allo

wed

3.9

0.67

1313

-14.

5N

ot a

llow

edN

ot a

llow

edN

ot a

llow

edN

ot a

llow

ed4.

2

0.75

14N

ot a

llow

edN

ot a

llow

edN

ot a

llow

edN

ot a

llow

edN

ot a

llow

ed4.

5

Rec

om

men

ded

1313

.5-1

4.5

139.

513

9.5

4.5

tub

e b

und

leß≤

0.67

ß≤0.

67ß≤

0.54

ß≤0.

47ß≤

0.54

ß≤0.

46

loca

tio

n fo

r

max

ran

ge

of

ß

NO

TES

:Le

ngth

s sh

own

und

er t

he U

L2 c

olum

n ar

e th

e d

imen

sion

s ex

pre

ssed

as

the

num

ber

of p

ublis

hed

inte

rnal

pip

e d

iam

eter

s (D

I)b

etw

een

the

dow

nstr

eam

end

of t

he 1

998

Uni

form

Con

cent

ric 1

9-Tu

be

Bun

dle

Flo

w S

trai

ghte

ner a

nd th

e up

stre

am s

urfa

ce o

fth

e or

ifice

pla

te.

* -

13D

i allo

wed

for

up t

o ß

= 0

.54

** -

9.5

Di a

llow

ed fo

r up

to

ß =

0.4

7**

* -

9.5D

i allo

wed

for

up t

o ß=

0.46

S -

Sep

arat

ion

dis

tanc

e b

etw

een

elb

ows

UL1

= U

L(To

tal u

pst

ream

leng

th) -

UL2

Not

e : T

he t

oler

ance

on

spec

ified

leng

ths

for

UL,

UL2

and

DL

is ±

0.25

Di.

Not

allo

wed

mea

ns th

at it

is n

ot p

ossi

ble

to fi

nd a

n ac

cep

tab

le lo

catio

n fo

r the

199

8 U

nifo

rm C

once

ntric

19-

Tub

e B

und

le F

low

Str

aigh

tene

r d

owns

trea

m o

f the

par

ticul

ar fi

ttin

g fo

r al

l val

ues

of U

L


TA

BLE

B -

OR

IFIC

E M

ET

ER

INS

TALL

AT

ION

RE

QU

IRE

ME

NT

S W

ITH

199

8 U

NIF

OR

M C

ON

CE

NT

RIC

19-

TU

BE

BU

ND

LEFL

OW

ST

RA

IGH

TE

NE

R F

OR

ME

TE

R T

UB

E U

PS

TR

EA

M L

EN

GT

H O

F U

L ≥

29D

i.

Dia

met

er R

atio

,S

ingl

e 90

° el

bow

Two

90°

elb

ows

out

ofS

ingl

e 90

° Te

eP

artia

lly c

lose

d v

alve

sH

igh

swirl

com

bin

edA

ny fi

ttin

gD

owns

trea

m m

eter

ßR

/Di =

1.5

pla

ne S

≤ 2

Di

used

as

an(a

t le

ast

50%

op

en)

with

sin

gle

90°

Tee

(cat

ch a

ll ca

tego

ry)

tub

e le

ngth

R/D

i = 1

.5el

bow

but

not

as

a he

ader

ele

men

t

UL2

UL2

UL2

UL2

UL2

UL2

DL

0.10

5-25

5-25

5-25

5-13

5-23

5-13

2.8

0.20

5-25

5-25

5-25

5-13

5-23

5-13

2.8

0.30

5-25

5-25

5-25

5-13

5-23

5-13

3.0

0.40

5-25

5-25

5-25

5-13

5-23

5-13

3.2

0.50

11.5

-25

9-25

9-23

7.5-

159-

19.5

11.5

-14.

53.

5

0.60

12-2

59-

2511

-16

10-1

711

-16

12-1

63.

9

0.67

13-1

6.5

10-1

611

-13

10-1

311

-13

134.

2

0.75

14-1

6.5

12-1

2.5

12-1

411

-12.

514

Not

allo

wed

4.5

Rec

om

men

ded

14-1

6.5

12-1

2.5

12-1

311

-12.

513

134.

5

tub

e b

und

leß≤

0.75

ß≤0.

75ß≤

0.75

ß≤0.

75ß≤

0.75

ß≤0.

67

loca

tio

n fo

r

max

ran

ge

of

ß

NO

TES

: Leng

ths

show

n un

der

the

UL2

col

umn

are

the

dim

ensi

ons

exp

ress

ed a

s th

e nu

mb

er o

f pub

lishe

d in

tern

al p

ipe

dia

met

ers

(DI)

bet

wee

n th

e d

owns

trea

m e

nd o

f the

199

8 U

nifo

rm C

once

ntric

19-

Tub

e B

und

le F

low

Str

aigh

tene

r and

the

upst

ream

surf

ace

of t

he o

rific

e p

late

.S

- S

epar

atio

n d

ista

nce

bet

wee

n el

bow

s.U

L1 =

UL

- U

L2N

ote

: The

tol

eran

ce o

n sp

ecifi

ed le

ngth

s fo

r U

L, U

L2 a

nd D

L is

±0.

25D

i.

A REVIEW OF THE REVISIONS TO API 14.3/AGA REPORT...

Documents

Transcript of A REVIEW OF THE REVISIONS TO API 14.3/AGA REPORT...