.Report No.9
Association Copyright
1998 1998 Amer Americ ican an ll ig ts
pera perati ting ng eric erican an 1515 1515 li
as ilso ilso
to
as Asso Associ ciat atio io
ecti ection on ssoc ssocia iati tion on oule ouleva vard rd
ir inia inia U.S.A.
talo talo
o.
98
Noth Nothin in cont contai aine ne in an Amer Americ ican an Ga Asso Associ ciat atio io (A.G (A.G.A .A.)pub .)publi lica cati tion on is to be cons constr true ue as gran granti ting ng ny ri ht im lica licati tion on or ot er ise, ise, fo th anuf anufac actu ture re al us in co nect nectio io it ny et od appa appara ratu tu or prod produc uc cove covere re by lett letter er pate patent nt no as insu insuri ring ng anyo anyone ne agai agains ns liab liabil ilit it fo infr infrin inge geme ment nt of lett letter er pate patent nt ir
ff
to
.G epre epre enta entati ti n. ar anty anty or ua ante ante in co nect nectio io with with A.G. A.G.A. A. ubli ublica cati ti ns an he eb expr expres essl sl isc/ isc/ im an liab liabil ilit it or re pons ponsib ibil ilit it fo lo or dama damage ge es ltin ltin from from thei thei use;f use;for or an viol violat atio io .G. shou should ld be view viewed ed as an endo endors rsem emen en by A.G.A. A.G.A. of an part partic icul ular ar manu manufa fact ctur urer er's 's prod produc ucts ts
Copyright
1998 Amer Americ ican an Ga Asso Associa ciati tion on Al Righ Rights ts Rese Reserv rved ed
Noth Nothin in cont contai aine ne in an Amer Americ ican an Ga Asso Associ ciat atio io (A.G (A.G.A .A.)pub .)publi lica cati tion on is to be cons constr true ue as gran granti ting ng ny ri ht im lica licati tion on or ot er ise, ise, fo th anuf anufac actu ture re al us in co nect nectio io it ny et od appa appara ratu tu or prod produc uc cove covere re by lett letter er pate patent nt no as insu insuri ring ng anyo anyone ne agai agains ns liab liabil ilit it fo infr infrin inge geme ment nt of lett letter er pate patent nt ir
ff
to
.G epre epre enta entati ti n. ar anty anty or ua ante ante in co nect nectio io with with A.G. A.G.A. A. ubli ublica cati ti ns an he eb expr expres essl sl isc/ isc/ im an liab liabil ilit it or re pons ponsib ibil ilit it fo lo or dama damage ge es ltin ltin from from thei thei use;f use;for or an viol violat atio io .G. shou should ld be view viewed ed as an endo endors rsem emen en by A.G.A. A.G.A. of an part partic icul ular ar manu manufa fact ctur urer er's 's prod produc ucts ts
Copyright
1998 Amer Americ ican an Ga Asso Associa ciati tion on Al Righ Rights ts Rese Reserv rved ed
FOREWORD This This repo report rt is blis blishe he as re omme omme de ract ractic ic an is ot issu issued ed as st ndar ndar ha been been writ writte te in th or of pe form form ncence-ba ba ed spec specif ific icat atio io Mult Multip ipat at ultr ultras ason onic ic mete meters rs ho ld meet meet ex ee th applicable applicable installatio installatio recommend recommendation ations. s. Applications, writ writte te in Marc Marc 1996 1996 Ther Theref efor ore, e, in case case of an conf confli lict ct etwe etween en th info inform rm ti th tech techni nica ca note note (App (Appen endi di C) th cont conten en in th main main repo report rt prev prevai ails ls
in th main main re or an
Rese Resear arch ch test test resu result lt an flow flow-m -met eter er cali calibr brat atio io data data have have indi indica cate te that that mult multip ipat at ultr ultras ason onic ic flow flow mete meters rs ca accu accura rate tely ly me sure sure ga flow flow at when when inst inst ll with with upst upst ea ipin ipin le gt uffi uffici cien en to prod produc uc full full deve develo lope pe turb turbul ulen en flow flow-v -vel eloc ocit it prof profil iles es Vari Variou ou comb combin inat atio ions ns of upst upstre ream am fitt fittin ings gs valv valves es an meas measur urem emen en erro errors rs Th amou amount nt of mete mete erro erro will will depe depend nd on th magn magnit itud ud of th inle inle velo veloci city ty prof profil il di tort tortio io ro uced uced th up trea trea ipin ipin co figu figura rati ti an th mete mete 's abil abilit it to comp compen en at fo this this th incl includ ud leve levels ls of puls pulsat atio ion, n, rang rang of oper operat atin in pres pressu sure re an ambi ambien en temp temper erat atur ur cond condit itio ions ns th repo report rt Flow Flow-c -cal alib ibra rati tion on guid guidel elin ines es ar prov provid ided ed fo occa occasi sion on when when
if flow flow cali calibr brat atio io is requ reques este te by th
unce uncert rtai aint nt (see (see Appe Append ndix ix A) micr microp opro roce cess ssor or syst system em Ther Theref efor ore, e, this this repo report rt incl includ udes es by refe refere renc nce, e, stan standa dard rdiz ized ed se of inte intern rnat atio iona na te ting ting ecif ecific icat atio ions ns ap li able able to elec electr tr ni as mete meters rs Thes Thes te ts summ summar ariz ized ed in Appe Append ndix ix are se to emon emonst st at th ac epta epta le perf perf rman rman of th mult multip ipat at ultr ultras ason onic ic eter eter's 's elec electr tr ni yste yste desi design gn unde unde diff differ eren en infl influe uenc nces es an dist distur urba banc nces es This This re
rt
ff rs en ra crit crit ri fo th
ea urem uremen en of as
ulti ulti at ultr ultras as ni mete mete s.
incl incl de fo that that purp purpos os at th en of this this ep rt
iii
" -
-
ACKNOWLEDGMENTS Repo Report rt No.9 No.9
Facilitator
Members th
Associates tr
ic
Jeffrey L. Savi Savidg dg
In
Ga Rese Resear arch ch In titu titute te
Kevin L.
Staf Staf Exec Execut utiv iv ic
io
re
rn
if Tr
iv
le
fo
E.
Me
Tushar
In
is
an Bosi
nd Reidar Sakari ssen
St toillK-Lab
tr
Gert Sl et
A.G. .'s Transmission
collectively to prepar
easure en
asunie Research
om itte
em er represen
industry guidelines recommendation
vi
broa
an reports.
base
ex erienc in
,
G
vii
4.5.3 Inspection an Auditing Function
•...........•.........................................................................................•...•
4.5.4 Alarm ........•.••..............•............................................................................................................................... 4.5.6 Engineering Units ......•.................................•...•..............................................•........................................... 4.6
5.2
5.3
re
re
tu
itio
In lu
12
g R 6.1
eakage Test
6.2 Dimensiona 6.3
-Flo
13
Measurements ific ti
6.4 Flow-C libr tion
7.1 Environmental
7.2 Pipin
13 13
ro
Test
14
16
Considerations
16
Configuratio
viii
ltraso ic sure
fo
S O
tu
licati
. R
ix
Introduction 1.1 Scope
independen pair of measurin transducer (acousti paths) Typica applications includ measurin th la ti distribution system an larg end-us customer mete sets 1.2 Principl
of Measuremen
Multipat ultras ic eter ar in er ntia eter that derive th ga lo rate by meas ring th tran it time of high-frequency soun pulses Transi time ar measured fo soun pulses travelin diagonally acro th ip down trea with th as flow an pstrea agai st th as low. Th differ nc in thes transi time is relate to th av ra as lo elocit lo th ac usti at s. Nu eric alculati it line conditions throug th meter. he ccurac of
ltra onic as mete de en
pr cise eo etry of th mete body an io
everal fa tors such as ltra onic transducer lo atio
iq
uniformity it Th accuracy of th transit-time measuremen depend on co iste pr pe
detectio om ensati
ound ul
wa
fo si na delays
eferen
positi ns
electr ni components an tran ducers
Fo th purposes of this report th followin definition apply: auditor designer inspector
Representative of th operator or othe interested part that audits operatio of multipat ultrasoni meter. ultrasoni meters assurance purposes
manufacturer
Compan that designs, manufactures sell an delivers multipat ultrasonic meters
operator
Compan that operates multipat ultrasonic meters an performs normal maintenance.
SPU UM
th electronic microprocessor system
th
Multipat ultrasonic mete fo measurin ga flow rates.
3.1 Ga
Qualit
comp sition mixtures pecified in A.G.A. Report No.8 (pur methane) an 0.87 Th manu acture houl be consulte if an ca id lf an elemental sulfur compounds.
This includes elativ de sities betwee
th foll wi io th
ar ex ecte ti PM
.554
1) ac usti wave attenuatin th
Deposits du to orma ga pipeline co diti ns (e g. conden ates or traces of oi mixe with mill-scale dirt or and) ma af ec th meter' accuracy by reducing th eter's cros -secti na area Deposits ma transducer 3.2
an in some de igns reflecte
th internal wall of th meter.
Pressures
Ultras ni transducer used in VM expected minimu
operatin pressure as well as th maximu
operatin pressure
3.3
The VM designer shal specif th expected operatin ga temperatur range.
C). The C).
electronics, ultrasonic transducers, cabling, etc. Th manufacturer shal stat th flowin ga an ambien ai temperatur specifications fo th multipat ultrasonic eter if they differ from th above. 3.4
Ga
Flow Considerations
te wi ga Th de igne ho ld determin th expected ga flow rate an ve if that thes values ar within th < I r n a x specifie by th manufacturer (see Sectio 5. fo definitions) Th accuracy requirements fo operatio within qminqtand qmaxar stated in Sections 5.2, 5.2. an 5.2. of this report Th designer is cautione to examin carefull th maximu velocity fo nois an piping safety (erosion thermowell vibrations, etc.) concerns. UM ha
th inhe en capability of meas ring flow in either directio with eq al accuracy i.e. they ar
m an u a c u re r c a p ro pe r
c on f g u
he SP
p a a me te r
3.5
tr meter' abilit to correctl in ig
---_ ..
...
io
ct
lo it
compen at fo such conditio s. Re earc wo ie
_ -
on mete in tallatio effect te to
Furthe recommendation ar provided in Sectio 7.2. of this report
io
Th mete body an al othe parts, includin th pressure-containin structures an external electronic components shal be designed an constructe of material suitable fo th servic conditions fo whic installation as specifie by th designer Unle ot erwi specifie by th esigne th eter hall uitabl or operatio in fa ilit su ject to th U.S. epartmen rans ortati 's DO regulati ns in C.F. Part 92 ransportatio of
4.2
Mete Body
4.2.1 Maximum Operatin Pressure
transduce assemblies. juri icti
in whic th
eter will be operated an fo th
ecifie
nvir
ment
te peratu
an e.
installation site an an othe requirements specific to th operator 4.2.
Corrosio
Al ette fluids.
arts
Resistance
th
eter
al
manufact re
at rial compatible with natu al as nd elat
specifie by th designer 4. .3
eterBod
en th
nd Bore
length to matc existing piping requirements Th UM bo
an th adjace
up tr am ip al
with flan es
ul
av th
am in id diam te to
"upstream 4.2.
Ultrasonic Transducer Port
Becaus natura ga ma contai some impurities (e.g., ligh oils or condensates) transducer port should be designed in wa that reduce th possibilit ofliquid or solids accumulating in th transducer ports.
al an nece ry additional evic unte on he tran ucer rt in rder to ak it ssible to epla th ul ra ic tr sd cers with ut pres ri in th eter ru that ase, le valv ay eq ired addi io to th is lation al to ns that es ur exists behi tran ucer before releasing the extractio mechanism 4.2.5
Pressure Tap
le ta ol
ne pres re ta al be ro ided ul et ee 1/8"
th inte na wall th Fo mete dy with Fe al pipe hrea
et bo wall thic
ho ld
meas ri
th
tati
re
re
th
eter
ac
re su e-
ho ld re rr an wi ed es an have mi im un ing. es le than 5/16", th hole should be 118"nomina in diameter
provided at each re ur ta fo
112" NP
1/4"
isolatio valve.
maintenanc access an proper drainage of gaug line condensate back into th mete body 4.2.6
Miscellaneous
UM is temporaril
se on th ground during installation or maintenanc work
installation Hoisting eyes or clearanc fo liftin straps should be provided 4.2.7
Mete Body Markings
nameplat containing th followin informatio should be affixe to th mete body th
anufactu er
axim axim
odel
er ti an
pres re an te
inim
ptio al pu chas arki ar ta pe h ou l b e u se d
ber, eria nu
actu
rder th
eter
onth an
ea
an fact re
erat re rang
at fl wi ber,
er nd
co diti rd
nu
y, low- tr ss ta
volu etri fl er nd/o
er ta
that ro ce
rate er
ur
mb ou de
tt
im ressio
4.3 4.3.1
Ultrasonic Transducer Specifications
Th manufacturer should stat th genera specifications of thei ultrasonic transducers, such as critical dimensions maximu allowabl operatin pressure operatin pressure range, operatin temperatur rang an ga compositio limitations. an fa tu er ho ld peci th inim er ting re re ased th ultrasonic ra duce model, UM size an expected operatin conditions This minimu pressure should be marked or tagged on th UM to lert th perato 's ield pers nnel that th eter ay regi te fl at re ce ip line pressures. 4.3.2
Rate of Pressure Change
expand inside th transducer Ifnecessary, clea instructions should be provided by th manufacturer fo maintenance and operation. 4.3.3
Exchange
It shal be possible to replac or relocate transducer withou
significan change in mete performance.
directed by th manufacturer th resultin shif in th meter' performanc shal no be outsid th limits electrical or othe measurements an adjustment have to be made 4.3.4
Transduce Tests
Ea
transd ce
ai
tr ns ucer
ho ld be test
th
an fa tu er an th
es lt docu ente
ermanent eria um er an be pr vi ed with th ge eral tran ucer at li te in Sectio .3.1 th SP requires specific transducer characterization parameters each transducer or transducer pair should metho used and characterizatio parameter(s) 4.4 4.4.1
Electronics General Requirements
Th UM's electronic system includin powe supplies microcomputer, signal processing components ex to th mete an is efer ed to as
Signal Proces in Unit SPU)
Optionally emot unit ontainin th ower up lies an th er to inte fa nonhazardous area an connecte to th SP by multi-conducto cable.
coul be in talled in
Th SP should operat over it entire specifie environmenta conditions within th mete performanc eq irements ecifie in Sectio 5. 5.2. nd .2.2 "Significant change is explaine in Sectio 4.3.3.
ve
ro ra
ault
nominal12V DC or 24 4.4.
Th
lo k- p. DC powe supplylbattery systems, as specifie by th designer
Output Signal Specifications
SP
ul be quippe with at leas rial at interface; .g., S-23 requency epresentin fl
et
ay al
Fl w-rate igna
be equi ould
low-fl ut ff fl rate is belo
ne
RS 85
rate
li
ed with an anal caleable up to 20
ncti inim
th
ollowi
ts
eq iv le
conditions (4-2 mA DC th
et r'
ould be provided that sets th valu no pplicabl to eria
wo eparat fl w-rate ut ut an
ut
ut
fo lo
aximum fl
at at line conditions ate,
'lmax.
low- at output to ze at ut t)
irectional st te output or eria
at valu
he th indi ated
sh ul
provided fo
computer(s an directiona stat output signal electronic design testin requirements of Appendix B. 4.4.
Electrica Safety Design Requirements
labo ator an then ac et ould la eled appr ve fo operatio in ationa El ctri de lass ivisio roup D, Ha ar us Area inim m. nt in ically safe esigns an expl ionroof enclos re esigns ar generall erti ie nd labele fo Di isio location Th esigne ay design. able jacket and grease 4. 4.5.
ub er
la ti an
ther ex os
part
ould
re ista
to ltra iole li ht fl me
il
ompute Progra ir ware
mp te odes es on ible fo th contro an peration of th eter ho ld emory. Al fl w-calcul tion co tant nd th erator-ent re pa ameter nonvolatile memory.
tore in ul al
no volatile to ed in
Fo auditing purposes it should be possible to verify al flow-calculation constant an parameters whil th mete is in operation. Th manufacturer should maintain
record of al firmware revisions, includin revision serial number
firmware. he fi mwar re isio um er evisio at eria um er and/ checks ould availabl to th audito by visual inspection of th firmware chip displa or digita communications port
eter dd additional feat re Th manufact re affect th accuracy of flow-calibrate meter.
hall otif th
pe ator if th
ir ware revi io will
4.5.2 Configuration an Maintenance Softwar
it la th followin meas re ents fl at at li conditio s, ea velocity averag pe of ound pe ound al ng each ac usti path an ltra ni ac stic ig al uality ecei ed by each tran ucer As software. 4.5.
Inspection an Auditing Function
configuratio parameters used by th SPU; e.g. calibratio constants, mete dimensions time averagin period an sampling rate Pr vi io
houl be ma
to pr ve
an accide ta
detect bl alterati
thos
ar mete
that
permanen programmable read-onl memory chip or password in th SPU. (Optional) should be os ible fo th au itor to ve if that al al orit ms co tant an config rati pa am ters bein used in an specific mete ar producin th am or better er orma ce as when th th verification becaus of th proprietar nature of some UM algorithms 4. .4
iarm
in voltage-free solid-stat switches isolated from ground li optional tr uble when an si ni ican erio time ti
everal
ia
4.5.5 Diagnostic Measurements
lo interface; e.g. RS-232 RS-485 or equivalent
onitored ar mete lt
fall ut id of or al operatio
or
Engineering Units
4.5.6
Th foll wi
nits houl be se fo th
ario
values associated with th VM
Parameter
SI Ilnit~
Il.S Ilnil~
kg pip diameter rP
pressure temperature
mls
e d
Pa-s
am /h
4.6
Documentation
lt as ni transducer an ze o- lo erificatio Th an facturer houl ls ro id al nece ar ata, ce ti ic te an docu entation fo correc configuratio set-up an se th particular mete if if io availabl fo th inspecto or th designer upon request. tu should be dated.
ta
escription of th meter, givi
c.
om nclature
arts with
an asse bl dr wi
th te hnical characte istics an th de cripti
h. 1.
J.
at rial of uc
it op ration
arts omenclatur
ic drawin of metrologically importan components
a w in g o f h e d a drawin
of co stitue
rinciple
with identification of th compon nt arts li te in th
wi dimensione
ld
ld
a.
g.
li
p la t o r a c p la t a n o f h e a r a ng em e n
o r n sc r p t o n
an auxiliar device
instructions for installation operation periodic maintenance and trouble-shootin
k. maintenanc documentation, includin third-part drawings fo an field-repairable components de cripti
of th electr ni SP
de cripti
of th availa le utpu
an it arra gement an
genera
escripti
ig al an an adju tmen mechanisms
it op ration
characteristics valu an operatin instructions regulations docu entati that th Requirements" r.
te 's perf rmance
cu entati that th Design Testing"
et r' de ig
eets th requirements
cces full
additional flow-rat measuremen erro of more than measuremen erro of more than
Se ti
"Per rmance
as ed th te ts in Ap en ix
"Elect onic
0.3%
0.3%
Th manufacturer should furnis specific mete outlin drawings includin overal flange face-to-face im si in id ia et r, mainte ance pace cleara es nd it on ection oi ts nd timate weight. Th manufacturer should provid
recommende
list of spar parts.
isolatin component; e.g. optica isolator relay, operationa amplifier, etc. This will allo to properly design th interfacin electronic circuits .6.2
ef re
th designer
ip
Prio to shipment of th meter, th manufacturer should make th followin availabl fo th inspector' measurements as required in Sectio 6.2.
This sectio specifie se of minimu measuremen performanc requirements that VMs us eet. If et is lo -calibrate he an fa ture al ro id uffi ie te at co fi in that each mete al ee thes er rmance re ir ents Th esig er al peci th eter lo measuremen performanc requirements before th applicatio of an calibration-factor adjustment Th
meas reme
ia erro
esig er is refe re to Ap endi
Sectio
.4.1
an expl atio of
th method an benefits of flow-calibrating mete an fo calibration-factor adjustment Th designer should also foll ca efully th installation recommendation of Sectio 7, as an in tallatio effect will ad to th overal measuremen uncertainty. efined in Sectio tbu 5.
ll te .1 Each UM whet er flow-calibrate or not, shal pe fo within th more ac urat to ~x it th te th ' I i n , as define in Sections 5.2, 5.2. an 5.2.2.
Definition
mete
nder test an th actual olum flow at
comp es ibilit
easure
etween th tw meters Th de iati
th ough th re erence mete Ty ically thre or mo nomina lo rate he test ru ca th repeatabilit as define below. Error [(Test Mete Readin -; (Reference Mete Reading)
te
used to etermine
Referenc Mete Reading) 100. Fi
an 5.2.2.
qltl2x
is
io li th flow-rat rang betwee qtand ' l m a x .
ll
The maximum
UM
5.2. an 5.2. fo larg an smal meters respectively transition
erro limi is ap lica le an wher an Sectio 5.2. nd 5. .2 ln
minimum
Figure unde
: s; ; O . l q r n a x .
UM
an Sectio
5. .1 an 5.2.
specific se oftest conditions
10
Se Figure
Referenc
et
measurernentaccuracy.
Repeatability
ti
te
conditions appr aching fr tr
th sa li
irecti n, fo
ull-
assumption of normal distribution Se Sectio 5.2. he malles step by whic th chan of th flow el city is indicate th mete Se Sectio 5.2.
Resolution
Velocity Sampling Interval
paths. Typically, betwee 0.05 an 0. seconds, dependin mete size Se Sectio 5.2. Zero-Flo
5.2
he maximu allowabl flow-velocit readin when th is i. lo components ar essentiall zero Se Sectio 5.2.
Readin
General
Th genera flow-measurement performanc making any calibration-factor adjustment. Repeatability:
of
shal meet th followin requirements prio to
±0 2% or q, ±0.4
Resolution:
qj
qmax
fo qmin qj< qt
0.00 ft/s (0.001 m/s) second qrnax
Zero-Flo
5.2.1
Reading:
<0.040 ft/s (1 mm/s fo each acoustic path
Large Meter Accuracy
te iz requirements prio to making an calibration-factor adjustment Maximu
rror
Qmax
40/ c.
11
VM
mall eter av been relaxe lightl ecause th di ficult in me suri turbulen ga flow when th path length ar shorter. aximum
acou ti tran it ti es in
±l.O% for qt:S;qj:S;qrna
rror
Fi
qj
ql
Izero-f1OW 1.6
1.4
Repeatabilit
1.2
±0.4
qj
qt
qt)
1.0 0.8
.... 0.6 ....
Q)
0.4 0.2
t t i
Maxi
-0.2 o, -0.4 -0.6
eak-to-pea
erro
_7
(q
l)
r ~-
-0.8
-1.0 (qj~qt)
-1.6
...,_-+
q t: : O . lq m a x '- ._
...-
qt
..
._
Performanc
Te
'_
'_ '_..
'_..
pecification
ummary
Co
Th UM shal meet th abov flow-measurement accuracy requirements over th full operatin pressure temperatur an ga compositio ranges withou th need fo manual adjustment unless otherwis stated by th manufacturer Ifth UM requires an al inpu to characterize th fl wing ga ondition (e.g., ty operator ca determin th need to change thes parameters as operatin conditions change
12
esigne
6.
th oper tor.
Di ension
easure ents
pairs.
(20
inspecto 6.
if present. ific io
13
--
know
an
documented
r' fin
Characterization
Method."
transducer pairs.
low-Ca ib atio
6.
es
VM rn
qmin'
0.70 < ! m a x , and q r n a x . '
andlor 0.
'In,ax
coul be usef l)
fo VM Xa duri
flow-calibrati
test 7.2.3, "Protrusions".
me
me
me
w-
wi
14
_.-._ -
manufact re to demo st at er rmance an will acilitat esearc eval atio urrent UM technology Th Ga Research Institut currentl ha progra to compil UM calibratio data 6.4.1 Calibratio
Factor Adjustment
If mete is flow-calibrated, th calibratio factor should normally be applie to eliminat an indicate mete bias error. Some suggeste method of applying calibratio factor are: a) Usin
low- ei ht
ea
rror FWME over th
eter's expe te
lo
ange (t
alculati
Usin or op isticate er corr ctio chem e. ., lti- oi or polyno ia al orit m, piecewis linear interpolatio method etc. over th meter' rang of flow rates. Fo bi-directional flow calibrations result of th
lo
second se of calibratio factor ma be used fo revers flow
calibratio to ptimiz th
eter's
er ll accura
performance.
he
an fa tu er
inte tional ia in rder to im rove ac urac at < I m i n . 6.4.
Test
eports
designer or th operator by th manufacturer Fo each meter, th report shal include, at minimum:
f.
th na
nd titl
th
erso (s wh
on ucte th te ts
h. th upstream an downstream piping configurations 1.
J.
k.
diagnostic report of th software configuratio parameters al test data includin flow rates, pressures, temperatures ga compositio an th measuremen uncertaint of th test facility descriptio of an variations or deviations from th required test conditions
retained in th an fa tu er's file Th to th perato on request, fo peri 6.
al ty
an acture ea afte
al ensure that th co pl te ep hi en of ny et r.
is availabl
ranc
Th manufacturer shal establis an follow writte comprehensiv quality-assuranc progra fo th as em ly nd te ting th mete an it el ctro ic te e.g. IS 000, AP Spec fication QI etc.). This quality-assuranc progra should be availabl to th inspector.
15
---------------------------------------------------------------------------------------------
In tallatio
quirem nt
ti is te to an in piping configuratio
ta that allows th UM to meet th expected performanc requirements
7.1 Environmental Considerations 7.1.1 Temperature
Th manufacturer shal provid ambien temperatur specifications fo th VM. Consideratio should be give to providin shade, heatin and/or coolin to reduce th ambien temperatur extremes 7.1.2 Vibratio
VMs should no be installe wher vibratio levels or frequencie migh excite th natura frequencie of
SP tr regardin th natura frequencie of th VM components. 7.1.
Electrical Nois
manufacture shall provid instrument specifications regardin electrical nois influences
7.
Piping Co fig ration
7.2.
Flow Directio
Fo bi-directional applications both ends of th mete should be considered "upstream. 7.2.2 Pipin Installations
io
ti
le to il is is ti th de igne ho ld consul with th ma uf ct re to re ie
ef ects is ng in eces ar fo th th mete ru
lo
esigne to alte th original pipi
co figuration
ir includ
th wo ta th latest te re ults
flow co ditioner as ar
To nsur that th UM when in talled in th operator's piping ystem, will pe fo within th pecified flow-rat easureme accuracy limits as hown in Sectio 5. 5.2. an .2 2, th ma uf ct re shal do on of th followin as desire by th designer/operator: 1.
ecom en lo ti
pstrea
supporte by test data
an
ownstrea
ipin co figu atio in io tall ' l i n and < l m a x .
16
inimum length
on with ut io lo li
2. Specif th maximu llowable flow disturbanc (e.g., th limits on swir angle, velo it profil asymme ry urbulenc intensity, tc at th me er's upstream flange or at some specifie axia distan upstream of th mete that il no re te dditiona flow-rat me surement erro of ha he ns fi rr ho ga flow rate betwee 'kin an qmaxTh recommendation should be supporte by test data flow-calibrate th UM in situ or in flow-calibra io made identica to th planne installation
facili
here th test piping configur tion an be
downstre from he po nt of initiation Sw rlin velo it prof le ma persis fo 20 pipe diam ters or re fl ro rl rt gh pipe requir to liminate th effect of an upstre flow disturbanc UM ma be bl compensate fo some leve of flow-profile disturbance. Research is stil bein conducte to quantify th sensitivit of different UM design to various flow-profile disturbances 7.2.3
Protrusions
loca disturbances to th velo it profiles Th UM bore flange an djacen upstream pipe should ll he r, flow di ur especially at th upstream flange Th upstream flange's internal weld should be ground smooth rt th 4,
he
re
an 12 o' lock position
ge
pr ud or ng keep th gasket entere hile tightening th nuts
by
be
1%
Thermowells, locate as specifie in Sectio 7.2.5, ar excluded from th abov protrusion limits 7.2.4
Internal Surface
kn -s on of deposits accumulate inside th UM th cross-sectiona area will be reduced, causin increase in ga velocity an positive measuremen error. Ex mples: iven 6.000" internal diam te M, wi ause 0.53 flow-measurement rror Fo 6% ro ri paper.) 7.2.5
ra correspondin
deposi ayer of only 0.008" roun th ns de surfac 20 000" me er th same 0.008" oating ould caus ne -p
Thermowells
Fo uni-directiona flow th designer should have th thermowell installe downstream of th meter. Th dire tional flow inst ll tions, th thermowell should be ocat
at le st 3D from ithe UM fl ng fa e.
fe ng ne manufacturer fo recommendation base on th most curren test data Th thermowell orientatio with respec to acoustic path should also be recommende by th manufacturer Catastrophic meta fatigu failur of th thermowell coul eventually result from th vibration.
17
7.2.
Accousti
Nois Interference
Some pressure-reducin al es ca inte fe
contro valves designed to reduce audibl noise, ma produc very high levels
it th operatio
ea by ltraso ic
et r. Re ea ch work
ltraso ic
is
pressure-reducing contro valve. 7.2.
Flow Conditioners
Fl co diti er mayo ay no ne es ar ependi th anufacturer' mete esig nd th severity of an upstream flow-profile disturbance. Th designer should consul with th manufacturer to determin th benefits if any, of installing variou type of flow conditioners give th upstream piping configuration. 7. .8
rientation of
eter
he esigne ho ld co sult with th manu acture to et rmin ther is pref rred et fo give upstream piping configuratio that is know to produc flow-profile distortions. 7.2.
orie tati
Filtratio
accumulati
of ep sits
to
ixture
di t,
il scale, conden ates an /o lu ri atin
il
ho ld
exist. 7.
Associated Flow Computer
Th UM's output is ty ically an unco re te volume (act al compressibilit
lu
li
co
itio ), either er
it
(t obtain standard cubi feet fo example) an to provid th necessar data retentio
manufacturer. "met
un
in
in le fl
comp te or with tw separate fl
co
ut rs
Fo othe applicable flow-compute requirements th designer should refe to AP MPMS Chapte 21.1 Flow Measuremen Usin Electronic Metering Systems. tha document. 7.3.1 Flow-Computer Calculation
Th eces ar ca culation re imilar to th equation escrib in A. .A of Ga by Turbin Meters an ar summarized in th followin expressions:
Qj(P/P
port
o.7, Measurement
a.n; (ZblZj)
IQbdt
Where: Qb
flow rate at base conditions flow rate at flowin conditions
18
-----.---------------------------------------------------------------------------
P, P,
1j Z;
cumulate integr te dt
integr ti
volu ov incr
as
nditio
ti ents of ti e, ty ically
second.
No.7. 7.4 Maintenance
ie dVerificationT st
electrical tests.
An extended analysis (beyon
19
------ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Ultrasonic Flow Measuremen fo American Ga Association, 1515 Wilson Boulevard, Arlington, VA 2220
Natura Ga Applications
American Ga Association, 1515 Wilson Boulevard, Arlington, VA 2220
Othe Relate Hydrocarbo
ompres ibilit acto of at ra Ga an Gases, American Ga Association, 1515 Wilson Boulevard, Arlington, VA 2220
National Electrical Code 1996 Edition, National Fire Protection Association, Batterymarch Park Quincy MA 0226 Electroni Metering Systems, American Petroleu
Institute, 1220
Flow Measuremen Usin Street NW Washington DC 2000
Standard Test etho fo Leak Usin Ultrasonics, American Societ fo Testin an Materials, 10 Barr Harbor Drive, West Conshohocken PA 19428-2959 U.S.A. Transportation of Natura Gas ff Washington DC 2040
Presen Status an Future Research on Multi-path Ultrasonic Ga Flow Meters, Transmission an Storage, Groupe Europeen De Recherches Gazieres asuremen of asfl Standardization, Case Postal 56 CH-I21
in lo ed co duit urbi Geneve 20 Switzerlan
et rs Internationa Organization for
Measuremen offlui flow in closed conduits ethods usin transi time ultrasonic flowmeters, Internationa Organization fo Standardization, Case Postal 56 CH-I21 Geneve 20 Switzerlan L R 6 Genera provisions fo ga volume meters 198 (E), Internationa Recommendation Organization Internationale de Metrologie Legale Bureau Internationa de Metrologie Legale 11 ru Turgot 7500 Pari France
Genera requirements fo electronic measurin instruments, 1994 (E), Internationa Document Organization Internationale de Metrologie Legale Bureau Internationa de Metrologie Legale 11 ru Turgot 7500 Pari France
20
-_
..
-
ltip th
ltraso ic
et
Fl
-C li ration Issues
Example: qt
qmax' qma.
qt qmin
UM
Error* (%
+0.953
qmln
0.10 qUI..
8,750
6,890
0.25
. .. .
21,875
21,980
0.40
q",u:
35,000
37,801
0.70
q",ax
61,250
60,415
q",..
87,500
A-1
90
UM diameter,
over
under
mete error.
1.6 1.4 1.2
aximum erro li it
1.0 0.8 0.6 0.4
.....
. ~(
0.2 Q) Q)
0...
-0.0 -0.2 -0.4
"' -,>:)--------------:}-~-
.\
',
.'-.--
----- ...... ------------,
,-
-0.6 Peak-to-peak erro
-0.8 -1.0
xi
-1.2
0.057% (qi~qt)
er or li it
-1.4 -1.6
70
30
100
q / q m a x (Percent)
re
ominal test rates.
A-2
--
--------------------------------------------------------------------------
ab ve exam le de on tr te th cert inty Th tota low- as re
te tial al nt erro of
mi imizin UM's co sist tw rt
ea
emen inaccuracy random (or precision)
systematic
meter' operation. Random errors normally follow
certai statistica distribution Th magnitud of th
accepted statistica principles Systematic errors normally caus repeated UM measuremen readings to UM ca help eliminat or at least, minimize th measuremen erro of th meter. Operationa experience as sh wn hat, in os ca s, he aj rt on of th tota lo -mea uremen rror an uncali rate UM is ue to sy temati er rs Du to machinin tolerances variations in componen manufacturin processes, variations in th mete minimize
particular UM's flow-measurement uncertainty, th manufacturer ca flow-calibrate
rror rrecti tech ique is availabl to th manu actu er th need of th operator
epen in on th
et
UM
applicatio
operationa range, more sophisticate erro correction techniques ca be applied. Fo instance
nd
higher
characterize th meter' output base on th availabl test data An exhaustive discussion of th variou mete erro correction techniques is beyond th scop of this discussion Th designer or operator should consul with th manufacturer regardin th availabl option fo particular UM
(F
nl
ingl calibratio -fac or correction
ppli
to th
et r' outp t. Ap lication
FWME fa to is
ly ne everal lter tive et ds adjustment to to minimize th flow-measurement uncertaint of th mete (see Sectio 6.4.1) Th exam le se in Sectio A. ab mete
ould experi nc
ri
will
be
ield er ice.
ed to em nstrat
ingl cali rati
this fact
to
M' cali ra io
to alculate th FWME
-facto (i e.
ne
ME orrectio
compared with th requirements specifie in Sectio 5.0.
A-3
.
---.-- -
FW
Al
s,
FWME Where: SU
SUM
Al
lc
qm x)
is th summatio of th individual term representing each of th test-flo m a is
weightin factor (wfJ fo each test lo
E, is th indicate
flow rate erro (i
points
oint an
erce t) at th te te
lo rate qi
(An alternativ method fo computin th FWME that decrease th contribution of th highes flow-rat in in 0.95 q m a r Th de ig er or th operator
middle or pp
ra ge of low. qmax
pr du es th re ults hown in able A.2. Note that column la eled i; th weightin acto that is ap lied to each value. ActualTestRateReferenceMeter (acflh
wf,
in
s,
q/qmax
wf,
(%
(%
+0.953
0.0397
3,475
87,500 acf/h)
+0.0378 +0.0296
6,890 0.2512
21,980 37,801 60,415
-0.2569
86,500
-0.3618 SUM
2.4807
SUM
in correction bein applie to th data). FWME
SUM (wi;
single calibratio factor F, measuremen error. Th valu of 1001 (100
lc
SUM (wj;)
-0.7672/2.4807
ll
to
-0.3093%
is calculated usin th followin equation
FWME)
Fo this exam le th FWME is 0. 09
each E, ha been adjusted to obtain
an th
ingl calibratio fact r, F, is calculated to be 1. 031.
calibration-factor-adjuste
value,
ie !,
usin th followin equation
A-4
------- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
--
s,
icl
(%
iC
(%
(%)
+0.953
+0.0503
+0.376
+0.0541 0.2512
-0.318 -0.315
-0.0025
-0.372 -0.0563
-0.366
SUM=
2.4807
SUM=
1.6r-~-~1---~1~-~1~~1--~~1--~~1--~~1--~~1--~~1--~~
1.4
f-
axim .8
::
Q)
0.6
r-
0.4
f-
0.2
f-
.
l<.,
......................
,>
er or li it
FWME-corrected
-0.0
- -
-!\ .---- -----------
-.. ----:'
Q)
-0.4 r-0,6
f-
-0.8
f-
-1.0
f-
-1.2
f-
Uncorrected aximum erro limi
-1.4 r-1.6
I
10
I
I
I
I
20
I
I
80
q/qmax (Percent)
A-5
I
Fi th te erro ha been virtuall eliminated by applying
th single FWME calibration-factor adjustment to al of th
data.
calibration-factor adjustment does no completely eliminat th measuremen erro becaus th UM ha ti it
meter's range.
Southwes Research Institute, Sa Antonio, Texas.
A-6
El ctroni
sign Te ting
he de ig of th UM's electronic should be te te to de on trat that th UM will continue to meet th performanc requirements of Sectio 5, whil operatin unde th influences an disturbances specifie eneral rovi io of Volu eters, Genera Requirements fo Electronic Measurin Instruments. OIML is th Organisation Internationale de Metrolocie Legale (i.e., th nter ationa Or anizatio of Lega Metr logy). OIML pu lishes thes oc ment fo th expres ed purp se armonizing national pe fo manc re uire ents an te ti procedures fo ga meters
climatic conditions th excl ding pola an eser en ironment ." Fo th echa ic co diti ns th requirements shal be fo clas 3/4M5, "locations with significan or high levels of vibratio an shock" it la or determinin th se erit leve fo each te t.
interconnectin wiring an customer wiring terminals, Th electronic shal be in operation, measurin electrostati discharg tests, th mete ma temporaril within 30 seconds. th in be expo ed to th sa
lt te te co dition as ther arts
Th foll wing ection pr vide
brie de cripti
stop functionin bu shal automaticall recove
ll ll th electronic ystem. th required test an se erit levels Note that th
publications. .1
Static
emperature
Heat
temperatur hall no exceed 1. Flmin (10 C/min) during heatin up nd coolin down th ai shal be uc that co densatio is voided at al time
he umidit
temperatur hall no exce 1. Flmin (10 C/min) during heatin up an coolin down Th of th ai shal be uc that co densatio is avoide at al times.
B-1
umidit
B.3
C) unit. B.4
He
in
B.5
Random Vibratio
Hz:
(0.048 m/s'
Duration: B.6
Sinusoidal Vibratio
B.7
Mechanical
Sh ck
edge. B.8
B.9
Powe
Volt ge Va iati
Main
voltage:
Main
frequency
lt
10% 2%
Sh
Reduction:
100% urin
10 ms (milliseco ds
8-2
B.10 Bursts (Transients)
po re to ur
erio
rsts
ltag
re etitio ti
pi es avin
bl ex nent al
inte al sh ll
Th
ea
av form al
ac
ik
al
av
is
sh ll be .5 V.
B.11 Electrostati Discharg
Ex
ur to 10 lect static di harges
it
time inte va betw en ac di harg of
econ
electrod is approachin th electronic an th discharg occurs by spark, th test voltag shal be B.12 Electromagneti
th kV
Susceptibility
Ex os to adiate el ct omag etic fiel strength of 10 Voltslmete (VIm).
Th
requ nc ra
al
.1 to 50 MHz, with
fi ld
B-3
------ - - - - - . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Engineerin
Technica Note M-96-2-3
This technica note contains referenc informatio
fo measurin high-pressur natura ga usin large-
measuremen performance, erro analysis calibratio an referenc literature.
Ultras ic
re eterin
as
Gr
1996 American Ga Associatio Al Rights Reserved
C-1
" " -
ACKNOWLEDGMENTS ic
Ul
Drenthen Ro McCarthy an Huib Dane ofInstrome Ultrasonic Technologies Kevi Warner of Daniel Flow Pr duct provided an ther significan ortion al ng with John Stua of Paci ic Ga an Electric timely editin were provided by Terr Grimle of Southwes Research Institute.
Pi (GERG) an it Ultraflo
ti project.
genera suppor an provisio of th speed-of-sound data
formerly of NO
Ga Transmission Ltd. wh laid th groundwork fo th Technica Note
Phil Barg Chai Ultrasoni Task Grou Marc 1996
C-2
ie
2
E
c F
CM
RR
DA
C-3
Intr
ctio
Am ia referenc informatio fo measurin natura ga with ultrasonic flow meters Ultrasonic meters measur flow by measurin
ib it
en ur ro er un er tandin result ca be achieved Single path an multipat
hi
turndown rati an incu
igh-frequenc
mall pres ur lo s. he unce tainty of
of th characte istics an li it ti ns of thes mete s,
meters ar addres ed
especially when dealin with non-unifor se as
ul es
to
instance ultrasonic meters av
fo
el city in th ga stream usin
lo
he
ther is
that th expected
refere ce to th number
path
th
ga velocity profiles an othe disturbances such as swirl.
io in refere ce in contract
ts
ic ti
a) Develo an A.G.A. Engine ring echnical Note to addres th following:
experience of th European Community. Develo an un er tandin Identify technica is ue c)
1.
hi
th potentia ap licati ns an relate busi es pa ameter
limitation an
evie curren in ustr tandards with operatio of ultrasonic meters neer ng
chnica
el te re earc need ie
to de el ping an A.G.A. report or th in tallatio
te Scop
echnical Note is limite to ultrasonic mete
in igh-pres ur
atural as transmis ion. Alth ug
recommendations relat specifically to high-pressur natura ga applications.
ri cipl of 2.
An
nd
er ti
Introd ctio
te
ic
-4
ic
Txl and Tx2, D. In
Tx2
Tx1
tD
til
(1)
where
denotes
!:... ( t U + t D ) tUtD
locity
space
t:
v(x,t).
-s
(2)
radius R,
ze
approximated by
im
il
th
ec
semi-empirical powe la
(r)
(3)
ax
where is equation applie (Schlichting, 1968
an pipe roughness. Fo smooth pipes, Prandtl'
.8
(4)
n, the
velocity profil v(r) ca be co puted, whic es entially is
Re=106 Cn=9.266)and Re=10
teady- tate de cription of th flow Fi ur mux
ab (n=11.109).
Re=10
1.0
£;
0.9
(I)
0.8 (I)
ro
0.7
0.6
0.5 ~--~--~~--~--~----~--~----~--~----~--~ 1.0
it
F ig ur e
S m oo t p ip e u rb u e n
C-6
e lo c
p ro f e s f o
,1
(n=7.455),
(x, t)
(x, t)
t)
(5)
where
2.2.2
Ultrasonic
Flow Measuremen
In (Lighthill
1972).
8P
where
denote
(6)
th pressure
th partia
-v
erivativ
Th thermo ynamic
ty
(x-
v(r),
(r)
(r
cos
constant
(r)
(7)
rp(r)
sound written
dx
r)
dt dr
(r
dt drp dt
, , -
C-7
dr
(8)
-
Re
(Ma),
Tx2
Tx1
Figure
Ex ggerat
ur atur
of cousti path
Equations (1)
(9)
and
(10)
(11)
C-B
averaged over th cross-sectio
(12)
dA
(13)
where k,
(14)
rv(r)dL
and
v(r)
Re, th
correction
Re.
(15)
0.011 1 0 g J O Re
Re will be different.
In
Rl2,
k, k,
C-9
th
verage flow calculation. lt
ofp. Since
(p)
(16)
2 ) ·R
wher 17
p)
p,
(p
Wi
(17)
V(Pi)
i=J
where
Pi
Wi
ir
A,
ic
VA
2.2.
(18)
eneratio of Ultrasonic Signal
th piezoelectri
element.
o n h a b ee n w e
d o u m n te d
S t p an i e n
1 97 1 a n H a
C-10
1 98 1
Whe
on
uo
an mi
is direction. Th acoustic frequenc an puls repetition rate ma vary betwee differen designs. 2. .4
ig
Proces in
In signal processing tw groups of method ca be distinguished: time-domai method an frequency-
time-of-flight versus th period of th ultrasonic pulse, or th path length versus th acoustic wavelength
method ar employed ea re en method an th correlatio ea -shi ethod. In carrie ut firs th ecei ed ul is etecte nd then it ti of arri al is stimated
0.5 Q) Q) __J
c:
0.0
.~ (/
-0.5
pred termined zero cr si gs in th ecei ed ul e. im le wa whic is widely used is to trig er on predetermine amplitud leve in th received puls an then detect th firs subsequent zero-crossing, ze o-cr ssings in th
ta le pa
of th pulse. In
is
indi id al transi time corr spondi to ea ze o-cr ssing. or adva ce method is to ak us th elativel fi ed am litu patt rn in th tr nsient ar of th ul e. some ha di ferent ap roac is to us correlatio techni ue i.e. to correlat th emitte an re eive puls an calculat th tran it ti as th time co re pondin to th ea of th orrelati fu ctio
C-11
._---
_ _ -
re re
3.1
C-14
through C-16).
93/0181.
C-12
_
-
CHARACTERISTIC
AN
COMPOSITIONS
GR Ekofis Coast
Air
Spee of Sound: ftlsec (m/s) 14.73, 60F HV Btulcf Mol Percent: Methane Nitrogen
Propane
78.03
0.4596
n-Butane i-Pentane
0.0473
n-Pentane
0.0324
0.0443
0.0480 0.0000
Speed-of-sound calculations were provided throug GRI' Basi Research Progra
C-13
on Flui Properties
2200 2100 2000
;:
1900
J)
1800
(J)
1700 1600
:::'I
(J
1500 1400 1300 1200 500
in 0.58 G,
700
650
600 "0 (J)
550
(J
"0
500
::::'I
(J
450
400
350 5.0
0.0
15.0
30.0
G,
C-14
J.
1550
1500
(J)
140f
1450 100F
1400 00
1350 ::::l
00
1300
I----r--
------r---___J
- - --- -
1250
......
1200 1000
1250
1500
(J) -...
8C Q)
00 c: ::::l
----:------
---_l_
.0
C-15
1450
1400 .5 C/)
lf
1350 A~ar A~aril illo lo
0.61 Q)
t" as
1300
(J)
"Eko "Ekofis fisk" k" Ga
1250 ::J (J)
1200
1150
1100
C/)
Q) Q)
J)
,0.61
:'A arill arillo" o"
10.6 10.6
Ekof Ekofis isk" k"
as
as
c: ::J J)
5.0
C-16
.0
7.0
9.0
10.0
3.2
Inst Instal alla lati tion on
Requ Requir irem emen ents ts
Mete Meters rs avai availa labl bl from from diff differ eren en manu manufa fact ctur urer er have have diff differ erin in reco recomm mmen ende de Gi en th if er nt esig esig an ropr roprie ieta tary ry fe tu es avai availa labl bl this this is ay re om en in tall tall tion tion ased ased .G A. Repo Repo o.7, o.7, th re ma or le trin trin en equi equi emen ements ts epen ependi ding ng te ti th Sing Single le-p -pat at mete meters rs will will requ requir ir cons consid ider erab ably ly more more stri string ngen en requ requir irem emen ents ts an th perf perfor orma manc nc of mult multip ipat at mete meters rs An im
rtan rtan co id rati rati
fo th
er is th
th bo
of te ting ting an re ar
prac practi tice ce fo inst instal alla lati tion on re onab onable le Ma fa tu er manu manu actu acture rers rs av one. one. ar stil stil unli unlike kely ly to matc matc is till till
it
mall mall rela relati tive ve
requ requir irem emen ents ts base base on thei thei test testin ing, g, bu th part partic icul ular ar inst instal alla lati tion on ma no prod produc uc th same same resu result lts. s. 3.2.1
Installat Installation ion Configur Configuration ation
Si ce th ultr ultras as ni et is lu d-ve d-velo lo it meas measur urin in evic evice, e, timu timu er or ance ance will will ac ie ed when when th pipi piping ng conf config igur urat atio io imme immedi diat atel el upst upstre ream am of th mete mete prov provid ides es well well-d -dev evel elop oped ed flow flow prof profil il ente enteri ring ng th mete mete Sinc Sinc mult multip ipat at ultr ultras as ni mete mete meas meas re th fl id el city city seve severa ra loca locati ti s, it infl influe uenc nc of mino mino flow flow dist distor orti tion on on mete mete perf perfor orma manc nce. e. To inim inimiz iz os ible ible lo -dis -disto to tion tion effe effe ts manu manu actu acture rers rs ma re om en lo cond condit itio ione ne in th ipin ipin pstr pstrea ea th eter eter Fl wirl wirl co diti diti ay be caus caused ed pi fitt fittin ings gs al es regu regula lato to eced ecedin in th et inle inle pi ing. ing. egar egardl dles es loca locati ti n, fl on itio itio er il no elim elimin in te th effe effect ct of stro strong ng jett jettin ing. g. Th inst inst llat llatio ions ns ec mmen mmen ed ip ia et rs up tr am th
anuf anufac actu tu er vary vary ut ener eneral ally ly re ir le et nd mi al pi ia eter eter wnst wnst ea
th th
to et r.
mi al should
minimum
cond condit itio ions ns exis exis (e.g (e.g., ., lo
leve levels ls of swir swirl, l, smal smal asym asymme metr trie ie in flow flow prof profil ile) e)
in id di mete mete of th inle inle nd utle utle ipin ipin sh ul be th requ requir ired ed to esta establ blis is th effe effect ct of smal smal chan change ge in diam diamet eter er
am size size as th
et r. Fu th
te ting ting is
Ther Ther shou should ld be no pipe pipe conn connec ecti tion on with within in th upst upstre ream am or down downst stre ream am pipi piping ng othe othe than than pres pressu sure re taps taps temp temper erat atur ur well wells, s, dens densit itom omet eter er conn connec ecti tion on or flow flow cond condit itio ione ners rs if they they ar used used Th manu manufa fact ctur urer ers' s' reco recomm mmen enda dati tion on fo th upst upstre ream am leng length th fo sing single le-p -pat at mete meters rs ar gene genera rall ll long longer er Th ultr ultras ason onic ic mete meter' r' desi design gn make make it inhe inhere rent ntly ly bi-d bi-dir irec ecti tion onal al In bi-d bi-dir irec ecti tion onal al inst instal alla lati tion ons, s, upst upstre ream am pi in is eq ired ired on ot en th mete meter. r. manufacturer. 3.2.2
Pressu Pressure re and Temper Temperatu ature re Measur Measureme ements nts
th re is tw -w velo veloci city ty ro il lo ated ated betw betwee ee
fl down down tr am loca locati tion on is im ract ractic ic l, he ef ec ho ld inim inimiz iz d. Ty ical ically ly ther ther well well le th an 5D th eter eter is ec en ed
C-17
ther ther well well pene penetr trat atio io on he e-th e-thir ir of th ip ia et r,
3.2.3
Density Measurement
create an unmetere bypass When used densitometer should be installe downstream of th ultrasonic meter. References should be made to manual on th variou densitometer fo furthe information. 3.2.4
Contamination, Strainers and Filter
ontami atio
th transducer
includin de ri that sticks to th surfac (suc as grease
wi
io to
oi buil
th
should be made in
iv
th ip in
th in
tr
ll li tl
te
en
e.
iz
tric
tr
3.2.5
ti
te
is iz
mo it re to ensu
ig
th strainer
filter is in go
im
th
in
conditio so as to preven flow istortio
Over-Range Protectio
il lo in re ul of th operatio of th
ie ownstrea
in th tr piping sy tem. Th accuracy of th mete ma be affected once
expected over-rang conditions ti
lt
ic
ld
damage to th transducers. 3.2.6
Bypass
ce measurement.
ll
installati ns withou this ca abilit
ic
te wi
installati ns re uiring ro in in ti ip
ti
or in tallatio
flow interruption is no allowed. 3.2.7
Pressure-Reducin
Valves an Nois
gene ates lt asonic nois (suc as co trol valves). On pa ticula proble trim
ti
C-18
is that nois attenuatin trim li ic te
closed valve, in clos proximit to th mete (especiall upstream is no recommended. Generall is in th frequencie is ot as frequencie coul interfer with th meter. 3.2.
Vibratio
Pipi
ibrati
ould be ep to
inim
uc
ro le
ltho gh ig
ince th el ctro ic pparat
oi
le el
ou te
th
esonan
ip will
manufacturer. 3.2.9 Orientation
orientations should work just as well primar consideratio is whethe th transducer ar mounte in ec se th dy an heth th se will be su ce ti le to collecting ir nd li ui in th as tr am Th orientatio of th transducer in th mete body differ amon manufacturers. 3.2.1
Piggin
So eter esigns il allo ig to as withou inte fere ce whil ther ay re ir et acti th tran ducers er th tran ducers ar rece ed ovisio to lean th transd ce pocket ho ld be
3.2.11 Pulsation
Preliminar testin ha show that pulsatio ha little effect on ultrasonic mete performance. In theory th im ac is all, ince an errors ho ld averag out. 3.2.12 Ga Composition
it in th
as phase, om ositio
ould ot av
larg
ff ct
erfo ma ce
hi
co po itio th ga sh ul availabl ro th ee ound owever this ma an indication that th compositio is changing high percentage of carbon dioxid (C caus problems as CO tend to attenuat th signal
erce ta usef
only as
3.2.13 Stress an Strain Issues
change in mete factor resultin from high-pressur ai ed th re lt cali ration 3. 3.3.
operatio is either availabl from th manufacturer
Mete Construction Genera
Th ultrasonic mete consists of tw basi components
C-19
constructe of material suitable fo th servic conditions fo whic th mete is rated. connection standards. Th
od
houl
identified to ho
th following.
manufacturer's name
1.
3. maximu
allowabl op rating re sure psig
in 8. 3.3.3
inimum operatin pres ur Electronics
he electronic incl de th circ it an device informatio to
equi ed to emit an receiv ultrasonic pulses measur
flow computer
Th lo informatio ca be tran fe re to flow computer ia se ia comm nication ort, throug frequenc signal or by othe analog methods. Sinc th ultrasonic mete is bi-directional th electronic provid flow in or atio in positive egativ nits to identify flow dire ti n. An directio indicators should als be available.
3.
Perfor ance Characteristic Ef
Th ultrasonic mete is esigne at th mete inle ha igni ican
or co dition that appr ache axia fl at th mete inlet. If th fl id wi (mai ly tangential comp nent ), th indi idua path velocities at lo us be re uced to an in ignifica leve thro gh proper installation practice as de cribed revi usly
meter design la containe
it
ul on in
os
ow
me
id
z in g o t
pa
g em e
uc
ho
eliminate by us of weightin factor determined theoretically or empirically.
C-20
.1
3.4.
eter Error, Uncertaint an Accuracy
ause by very smal
ee sh ul
iscr pancie in th
co sulted fo th
easu emen
agnitu
th up tr am an down tr am tr nsit ti es at
th zero ff et
Fo increase accuracy meters ca be flow-calibrated. An timu mete acto nd ze ffse re then calc late that th eter er at an iv fl rate within th eratin ange th eter il fall within th manufacturer's specifie erro band fo flow-calibrate meter, whic is typically+ 0.5% om lete tate en of system.
ea rement erro
eeds to incl
th meas re en uncertai ty
th prover
3.4.3 Mete Performance Curv
pr duce te ting each eter agai pr ve te at everal if er nt fl rate then pl ttin th perc nt ge er or in fl re di vers th prover fl rate Th erro curv ca lotted gain ct al flow rate averag velocity or Reynolds number dependin on th intended purpos of th curves pressure an temperatures ar generall
family of distinctiv curves
When th mete erro is plotte agains th Reynolds number th mete performanc curv tend to show flow-profile effects. This method of plotting woul normally be used only fo specific experiment an is
function of th Reynolds number 3.4.
ress re
os
protrusions, obstructions or constriction to th flui flow 3.4.
xi
Flow Rate
Ultrasonic eter re enerally esig ed fo aximum el city ax maximu actual flow rate Q n I Q X . to receiv undistorte ultrasonic pulses whic ca be function of both transducer design an th signal temperatures within th stated maximu
mete operatin range.
Th maximu
unctio
flow rate Q m a . T is
impl
Qmax
th
axim
max
appropriate pressure temperature and compressibilit corrections.
C-21
el city gi en (19)
3.4.
Mini um Flow Ratean Rangeability
it whic th mete will operat within om specifie
th te ncertainty or rror li it Obviously, th mini um 1.0% Ge erally th
compressibility corrections 3.4.
In
Pulsatio
Effect
number of measuremen applicatio (e.g., co pres or tati ns), th flow te Fr ti ie
ay be ulsati io
instea
io
ti
coul occu if th pulsatin flow is synchronized with th transducer firing rate .4.8
of
th
in
ly
compensation method It ma also be influenced by th flow characterstics of th specific installation
3.5.
eneral
Th most comm nl
applie
ield chec
ar th
is al in pectio an speed- f- ound test Operationa
accurately known, this measured peed of soun ca
compared with
calculated peed of ou
ased
measuremen an on length measuremen of th spoo piec ar within acceptable limits 3.5.2 Visual Inspection
In visual inspections, th mete bore an transducer port should be inspecte fo accumulation of solids erosio or othe damage that woul affect mete performance. 3.5.
Maintenanc
T h u l a so n
an
Inspection Frequenc
m e te r h ou l a le r h e o pe ra to r w h e
p ro b e m o cc u
wi
h e o pe ra t o n o f h e m e te r
ti it is installations, periodic inspection ca be establishe prio to expected alar conditions to preven thei la th th ultrasonic et ca co ti ue to function with th lo al ut ne of th path meas reme ts
C-22
--
valu tion of
ea urem nt
rfor
nc
systems. It try
complex.
th
ti
im
code.
Ultrasonic
igna Diagnostic
C-23
fo an accurate transit-time measurement. Ifther is proble with th ultrasonic signal themselves th transit-time measuremen and, therefore, th flow measuremen will be incorrect. It is impossible to stat ta
detection).
sy te ma expe ienc etection problems impl exam le of detectio roblem is cy le ki pi g. olde analog meters th threshol detectio techniqu woul determin detectio poin by choosing amplitud sometime change du to pressure change an th threshol is crosse by im resultin in an appreciabl erro of severa percent. This particular detectio proble checking speed-of-sound measurements among different paths.
differen portio (lit) different
ca be identified by
An
effect ar discusse here fo reference. Thes effect do no necessaril caus in tr signals.
4.1
measuremen problem,
Signal Amplitud
ig
in contro becaus th effect is so widespread Some transducer ar stronger than others especially specia
am litu th igna be au of prea in an si na catterin within th as Very larg mete av weaker ig al Partic late li uids in th ga ma catter th si na an weaken it Di tran uc face ma weaken th signal Temperatur also ha some effect on th ultrasonic signal strength 4.2
Puls
Shap
transducer in orde to produc pulses with well-defined fron edge that simplify detection. Much coul
ls
th
lo
th
selectivel attenuat certai frequencie within th ultrasonic signal and, therefore, change it shape.
C-24
can
4.3
Noise
outside
4.4
Batc
Diagnostic
C-25
32
ill
om inatio
ia
stic
with irre ul
sa
velo it
lock
measur ment
le
62.
STATA
64.
STATC
65.
STATD
66.
STYSTAT
67.
FAIL
68
BIFAIL
69.
CIFAIL
71.
A2FAIL
72.
B2FAIL
73.
C2FAIL
74.
D2FAIL
C-26
ma
indi at
flow-profile irre ularitie
Detailed explanatio of code in messag block: A..
ti
reject signal ifnois 1se
leve exceed limit.
reject signal if IN rati belo
limit.
reject signal if measuremen qualit chec failed ejec
ig al if peed of ound ut of an e.
reject signal if delt time chec failed reject signal if span chec failed reject signal if polarity chec failed reject path if flow-profile chec failed. reject signal if peak puls widt exceed limit. reject signal if signal qualit belo flo
limit.
change indicator
reject path if intra-chor qualit chec ou of range. reject path if speed-of-sound deviatio ou of range. path is manually se to
inacti e.
path failed 66 SYSSTAT, Genera system status
puls accumulato error. processo RA progra
error.
memory error.
EEPROM error. DS progra
memory failure.
DS
memory er or
DS
em ry er or
number or operatin path belo 67
4, A.. A...
FA
.. Number
minimum.
ailu es er atch
designates whic path in th multipat meter.
downstream).
C-27
rAn
is 10
IS
an eviation cr Th root-sum-of- quar
ta dard
erro
(20)
where
length
ar necessary.
presente
here
followin
expression
(21)
C·28
(22)
Ro
ga pressure Po (at
w,
therma
expansio ra
re
I,
Ro p,- Po
RJ
(23)
(X) 2xl0
Nzmrrr (29xl0 Iblinch
Nzmrrr',
R,
1.000458
Ra
(24)
=435.6775
B.!_
X,
!1T
(25)
(26)
),
14xl0-
°K
-6
1% variation.
C-29
" -
An increased
5.1.
Factors Relatin
to Calculatio Technique
Wi
.1.3 Errors in th
Ti
easure en
tu
and
tD
ro
(27)
Ve
ra
fl
5.2 Variance
rv times
re tu
and
tD
C-30
-
cloc chosen is sufficiently small, th effect of quantization nois ca be neglecte when operatin well ac stic at as th main ca of th ra om er rs in ultras ni ea re en tationar fl w. ariati th ee ou d, ecause it is low, ca egle te as om ar it turb le velo it variations Equation (11) show that th measured velocity actually is an averag alon th acoustic path us th longer th ac stic path le gt th mo th tu bule fl ct atio ar aver ge ut re ulti
vera in peri measured velocity
ne ec nd an
fficiently hi
repeti io rate th st ndar
eviati
O"
(28)
where cousti at This elatio expres es lowe ou th achi vabl ta dard eviati in ultrasonic velocity measurement. Dividing both side by ield ex re io or th tandar eviati th relative error
(29)
iven ix length measurin time at lo
6.
approaches ze o, th relative erro ec es arbitraril large, ea ri time Practi al calibratio roce re therefore, ho ld increa th velocities as compared with higher ones
Dry Calibration
Dry cali ration with ut luid fl w) involves th ac urat piece, whic includ th spoo diameter D, nd ea
ea rement th di en ions th po th at th imen ions and X. Errors in th dimensions directly affect th measuremen accuracy Th expected uncertaint of multipat mete em nstrat this oncl iv ly hi er accu ac is ee ed ndlo traceability to required flow calibratio is recommended.
atio al ta
ar is
specific se of electronic an transducers. ne et od is to un tw tr ns ucer in re ac uratel nown he cham er is ille with
ri ed test cell he epar tion th tran ucer ga us ally itrogen) fo whic th el city of
tu
C-31
and
tD
us nd
D.
6.1.1
.2
Velocity Distribution
Flow Calibratio
re
Any
facility. It
m in e
he
nd
ma
C-32
or
me
m en t
he ow
li
ilitie
capacity.
6.
ra sd ce
eplace en
6.4.
Mechanical Procedur
outside
6.4.
Computationa Procedur
C-33
in th
ignal-processing
lgorithm
6.4.3 Verification
%)
gas composition.
[t
(0
ecom endation 7.1
Industry
mistake.
7.2
Users
application.
rs
fi
C-34
tl
].
7.
an facturer
Manufacturer should continue to develo th technology an strive to provid increase reliabilit an technology will need to come in larg part from th manufacturers.
.4
es ar
er
Industry-funde research as well as research funded by individual users, need to augmen work done by th manufacturers. Research on variou piping configurations will increase th confidence of al partie in th technology
C-35
advancemen
of custody-clas
measurement.
C-36
SL. }';O.
TITLE
F lo w
ns ct ns
ut
as
en
nd
AlITHOR(S)
SPO},;SORS
DATE
L.C. Lynnworth
IEEE
1981
C o u mb i G a S y t e Servic Corp
1982
.D
un
Measurement
Z L
s a d G
AG, Switzerland M.E. Nola J.G. Q'Hair
Flows nt Ultrasoni he
ac
nd
al
A.M. Lynnworth
a th s
pm
on
of he
ga
ec
on
al
ea
P. Hojholt, Danfos
on
F lo w
e su l
11
or an
as Pa
1988
Co
g ra l
ou
uc
1988
us as ni
on
12
London Research
/S
. O
ga industry
10
1983
. E. N an . S C he un g
e te r
M.E. Nola
as ni
Panametrics
L.C. Lynnworth
Flowmeter ev
h G
ct
P ip e
or La
1989
n D
F lo w e te r
Delsing
na
ns
of
1989
1989
c tu r R.S. Flemons an
14
en va on
of
po
am
an
A.J. Algra,
Smid.
c co
el n,
1990
R.J.W. Peter H ig h P re s u r
1990
Gasunie
ys
15
17
er
U lt ra so ni c
1990
NEL
M et e P ip e i n
F ow
G a J ou rn a
1990 1991
e su l pb E.G. DIBello
18
e- at
as ni
F lo w
e te r
e lo c
nt
1991
at nd
19
D ev el op me nt s
i n U lt ra so n
f lo w M et er in g
es 1991
C-37
SL. 1'
TITLE
AUTHOR(S)
SPONSORS
DATE
Statoi!
1992
20 J.
P et er s
A. Lygre, T.Folkestadt R. Sakariassen, D.Alda1 Jouk
E. Halttune
sa
u nt t
. H
23
K. an
1992
e ll e
Daniel Industrie
1992 1993
25 T i e s i n H ig h- Ra ng ea bi l t y
P. Soni
as
lo
S on i
K.S Mylvaganam
M et er i
l tr as on i
as
lo
et rs
. D F JJ .
k S i j m an s
.W
28
n J W. Russwurm
. M
Noel Bignel k S
Sonic
1994 Flowmete
Calibrations 1994
. v 34
Ultrasonic
Mete
1995
Experience R. Wilsack
1995 1995 1995 l.W. Smalling
1984
Panametrics
L.D Braswell L.C Lynnworth D.R Wallac C on t o l
1995
9 . p . 9 2- 10 1
(Oct 1995 40
I.
C-38
K ol m k ov ,
.G
at
Metrologiva (publication
1987 USSR)
SL.
RESEARCH
SPONSORS
DATE
NGTL
1995
~O.
Contamination Instrornet pera iona
1995
Effects
esting
Offshore Cu tody
ultipath Ultras ni
eter
rans er Se vice
TCPL
1995
Statoil
1995
in
1994
d 1
k H - M
1994 1995
" M
1994
10
11 12
1990 2" Fluent
Cali ration
Fluent
@K-Lab
1991
@ K 14
24 Daniel
Cali ration
@ K
15
24 Daniel
Cali ration
@ K
18
@ K @ K @ K
21
Gasunie Lab
1993
Gasunie
1994
22
C-39
NO. 24
Soni
Swir
Gasunie
1993
Oklahoma Na Ga
1994
Britis Auckland
1994
C-40
"-
REFERENCES
moving media, incl ding
Flanges, DU
au tics
5675/0007/4.
En
in
Ex
En
in
Me
ro
USA.
USA.
Jr.,
C-41
._-_
...•
-