gpa 2140 - liquefied petroleum gas specifications and test methods

8/10/2019 GPA 2140 - Liquefied Petroleum Gas Specifications and Test Methods

1/56


2/56


3/56


4/56


5/56


6/56


7/56


8/56


9/56


10/56


11/56


12/56


13/56


14/56


15/56


16/56


17/56


18/56


19/56


20/56


21/56


22/56


23/56


24/56


25/56


26/56


27/56


28/56


29/56


30/56

TABLE 2 Instrument ConditionsFlowRate, Carrier G

mLlmin

Substrate, Temperature,olumn

Diameter.mm. D Mass, % OC

ColumnLength, m

I

Silicon200/500 4 6.4 27 90 60 to 70 heliumBenzyl cyanide-silver nitr ate 9 6.4 36 40 45 to 55 heliumHexamethylphosphoramide 9 6.4 17 30 60 to 70 heliumDimethylsuifolanepius benzyl cyanide and 7 6.4 36 35 60 to 70 helium

Dimethylsuifolane 15 6.4 30 25 30 heliumHexamethyl phosphoramide 6 3.2 25 28 12 heliumDi-n-butyl maleate 4 6.4 25 28 60 heliumTricresyl phosphate plus silicone,550 9 6.4 30 35 70 heliumMethoxy ethoxy ethyl ether 9 6.4 30 30 60 helium

silver nitrate

TABLE 3 Piecirio n Datafor LPG Containing Less Than50 X PropeneConcentration Rangeof

C o m m n t s . mol % Repeatability Reproducibility

O to 70Above 70 0.2 1 % o amount present

use repeatabilitycurve in Fig. 3 use reproducibility curve in Fig.3

ventilation. Cylinders must by supported at all times. Hydrocarbonvapors that may be vented must be controlled to assure compliance withapplicable safety and environmental regulations. Vapor reduces oxygenavailable for breathing. Liquid causes cold bums.

8.2 Preparation an d Introductionof Sample-Attach thecylinder containing the gas mixture t o the sampling valve ofthe chromatograph so that a liquid phase sample is with-drawn. Adjust the flow rate from the sample cylinderso thatcomplete vaporizationof the liquid occurs at the cylindervalve. (An alternative technique is to trap a sample of onlyliquid phase in a short section of tubing, and then permit th eentire sample to vaporize into an evacuated container).Adjust the ratio of the two volumesso tha t a gage pressu re of69 to 138 kPa (10 to 20 psi) is obtained in the final container.

Then use this sample for the analysis. Fiush the sample loopfor 1 to 2 min at a flow rate of5 to 10 mL/min beforeintroducing the sample in to the &er g a s stream.

8.2.1 On propene concentrates, the sample maybe intro-duced as a liquid by means of a liquid sample valve or byvaporization of the liquidas above. On propene concentrateshaving a propene content of less than80 %, only thealternative technique of trapping a sample of liquid andvaporizing the entire sample into an evacuated containershall be used.

8.2.2 Sampling at th e sample source an d a t th echromatograph must alwaysbe done in a manner thatensures that a representative sample is being analyzed. Lack

of precision and accuracy in using this method can most

TABLE 4 PiecisionData for Propene ConcentratesCOmpOUnd Concentration,moi %

Ethane 0.0 o 0.10.2

Propene 70 to 7793 to 95

Propane 5 to 722 to 29

Butanes 0.0 to 0.10.50.61

Butenes 0.2

Repeatability

0.020.050.380.340.331 .o0.040.040.10.10.07

Reproducibility

0.040.061.51 o1 o1.70.080.20.30.50.2

often be attributed to improper sampling procedures. (See

Test Method D 3700.)8.3 Preparatio n ofthe Chromatogram-Obtain duplicatechromatograms of the sample. Adjust the attenu ator at eachpeak for maximum peak height within the recorder chartrange. Peak heights of like com ponen ts shall agree within1mm or 1 % (whichever is larger). If a reference standardmixture is used for calibration, obtain duplicate chro-matograms of the proper reference standard in a similarmanner. Use the same sample size for all runs.

9. Calculation9.1 Peak Height Method-Measure the peak heightof

each component and adjust this value to the attenuation ofthe same component in the reference standard mixture.Calculate the percentage by moleor liquid volumeof eachcomponent as follows:

Concentration, liquid volume or moi percent = PJP,) Swhere:P, = peak height of com ponen t in the sample,Po = peak height of component in reference standard

S = percentage of moleor liquid volume of componen t in

9.2 Area Method-Measure the area of each com pon entby multiplying the heightof the peak by the width at halfheight. The width shouldbe measured with the aid of amagnifyingglass (Note 3). Adjust the area to the attenu ationof the same component in the reference standard mixture.

NOTE -The use of planimeters or integrators is permissible pro-vided their repeatability has been established and the resulting repeat-ability does not adversely afiect the repeatability and reproducibilitylimits of the method given in Section 10.

9.2.1 Calcu late the percentage by moleor liquid volumeof each componentas follows:

where:A, = area of component in sample,A, = area of compon ent in reference standard m ixture, and

mixture, and

reference standard mixture.

Concentration, liquid volume or mol percent = (AJA, ) x S

26right Gas Processors Associationded by IHS under license with GPA Licensee=UK LOCATION/5940240005

Not for Resale, 08/21/2007 09:17:10 MDTproduction or networking permitted without license from IHS

--` ` ` `

, ,` ` `

, , ,` ` `

,`

, , , , ,`

, ,

` ` ` ` -` -`

, ,`

, ,`

,`

, ,` ---


31/56

S T D * G PA S T D 2 1 ' 4 O - E N G L 1 9 9 7 3 8 2 L i b 9 9 0 0 1 8 5 2 5 L(L(T=

Precision, 95 Percent Confidence Limit

FIG. 3 Precision Curves

S = percentage by moleor liquid volumeof component in

9.2.2 If pure components are usedfor calibration, calcu-reference standard mixture.

late the composition as follows:

where:A , = area of component in sample, mm 2, andA , = area sensitivity of compon ent, m m2 per percent.

9.2.3 Total the results and normalize to 100%.9.3 Normalization-Normalize the moleor liquid vol-

ume percent values obtained in 9.1or 9.2 by multiplyingeach value by 100 and div iding by the su mof the originalvalues. The sum of the original values should not differ from100.0 % by more than 2.0%.

Concentration, mol percent= AJAf l

judging the acceptabilityof results (959 confidence) onsamples containing less than50 % propene. The data inTable 4 shall be used for judging the acceptab ility of resultson samples containing m ore than50 % propene.

1O.1.1 Repeatabifity-The difference between successivetest results, obtained by the same operator with the sameapparatus under constant operating conditions on identicaltest material, would in the long run, in the normal andcorrect operationof the test method exceed the values shownin Table 3 or Fig. 3 and Table4 in only one case in twenty.

1O. 1.2 Reproducibility-The difference between twosingle and independent results, obtained by different opera-tors working in different laboratories on identical test m ate-rial, in the norm al and correct operationof the test method,exceed the values shown in Table 3or Fig. 3 and Table 4 nonly one case in twenty.

10.2 Bias-Since there is no accepted reference mate rialsuitablefor determining the biasfor the procedure in this testmethod, no statem ent on bias is being made.

11. Keywords

10. Precision and Bias510.1 The data in Table 3 and Fig. 3 shall be usedfor

5 The data from which thisprecision statementis based are not available. 1 1.1 analysis; liquified petroleum gas



- -

` ` ` `

, ,

` ` `

, , ,

` ` `

, `

, , , , ,

` , ,

` ` ` `

- `

- `

, ,

` , ,

` ,

` , ,

` - - -


32/56

APPENDIX

(Nonmand atory Information)

X1. PARTITION COLUMNS

X1.l The following four partition columns have beencooperatively tested and fo und suitable for use with materialsgiven in the scope of this test metho d.

X .1.1 Silicone200/500 Column-This colum n separatesethane, propane, n-butane, isobutane, n-pentane, andisopentane andis therefore suitab le for analyzing LP gasesfree from unsa turated hydrocarbons.

X 1.1.2 Benzyl Cyanide-Silver Nitrate Column-Thiscolumn separates isobutane, n-butane, the butenes,n-pentane and isopentane, and accordinglyis best suitedforuse with L P gas butane con taining unsaturatedC, hydrocar-bons.

X .1.3 Hexamethylphosphoramide(HMPA) Column-This column separates ethane, propane, propene, isobutane,

n-butane, the butenes, n-pentane, and isopentane, and ac-cordingly is suitable for use with all typesof LP gases.

X . 1.4 Dimethylsufolane (DMS)-Benzyl Cyanide-SilverNitrate Column-This colum n separatesail components incomm ercial LP gases.

NOTEX 1 . 1-There are commercial suppliers of gas chromatographyequipment and columns who can make (and guarantee) that thecolumns they provide will meet the specifications (see 6.7 Columns) ofthis test method.

NOTEX1.2: Warning-toxic. Precaution-See the product safetybulletins from the suppiier of the chemicals used in preparing thesecolumns or before Benzyl Cyanid asilve r Nitrate Column; XI. .3 Hexa-methylphosphoramide (HMPA) column , and X 1.1.4 Dimethylsufolane(DMS)Benzyl Cyanide-Silver Nitrate Column.

The American Society for Testing and M aterials takes no position respecting the validity of any patent rights ass me d in connectionwith any item mentioned in this standard. Users of this standa rd are expressiy advised that determination of the validity of any suchpatent rights , and the risk of infringement of such rights, are entirely their own responsibiiity.

This standard is sub ject to revision at any time by the resp onsible technic al comm ittee and must be reviewed every five years andif not revised, either reapproved or withdrawn. Your comments are nvited either for revision of this standard or for additional standardsand should be addressed t o ASTM Headquarters. Your comments w ill receive careful consideration at a meeting of the responsibletechnical comminee, which you may attend. If you feel that your comments have not received a fair hearing you shouki make yourviews known to the ASTM Committee on Standards, 1916 Race St., Philadelphia, PA 19103.



- -

` ` ` `

, ,

` ` `

, , ,

` ` `

, `

, , , , ,

` , ,

` ` ` `

- `

- `

, ,

` , ,

` ,

` , ,

` - - -


33/56

S T D - G P AS T D Z L q O - E N G L 1997 M 382Lib77 0038527 2L2 M

Designation:D 2713 - 91 An Amencan N a W Standard

Designation:395/92

Standard Test Method forDryness of Propane (Valve Freeze Method)This standard is issued under thefixed designation D 27 13; the num ber imm ediately following the designation indicates the year oforiginal adoptionor, in the case o f revision, the year of last revision.A num ber in parentheses indicates the yearof last reapproval.Asuperscript epsilon(6) indicates an editorial change since the last revisionor reapproval.

1. Scope1.1 This test method covers the measurement of the

dryness of propane-type products such as, but n ot lim ited to,commercial prop ane (see SpecificationD 1835).

NOTE -This test method is not applicable to propane-type prod-ucts containing antifreeze agents. However, the relative freeze times ofsuch materials tested by this procedure maybe an indicationof the

tendency of these productsto cause freezing in pressure reducingregulators.

1.2 This standard does not purport to address all of thesafety problems, if any, associated with its use. It is theresponsibility of the user ofthis standard to establish appro-priate s afety and health practices and determ ine the appiica-bility of regulatory limitations prior to use.

1.3 The values stated in SI units are to be regardedas thestandard. The values in parentheses are for information only.

2. Referenced Documents2.1 ASTM Standards:D 1265 Practice for Sampling Liquefied Petroleum (LP)

D 1835 Specification for Liquefied PetroleumLP) Gases2Gases2

3. Summary of Test Method3.1 A liquid-phase sample of the product tobe tested is

flowed through the prop ane w ater test valve to cool the valvebody. After cooling, the test valve is partiaily closed to asmall preset flow rate and the time required for the valve tofreeze, and thus interrupt the normal flow, is recorded. Theaverage observed time for several successive observations isrecorded as the observed freeze time.

4. Significance and Use4.1 This test is a functional test in which the water

concentration in the product is related to product behaviorcharacteristics in a pressure-reducing system of special designto arrive at a measure of product acceptability in co mm onuse applications. Experience has demonst rated that excessivewater c onten t (dissolved water) will cause freeze-up difficul-ties in pressure reducing systems.

This test method is under the jurisdictionof ASTM CommitteeD2 onPetroleum Products and Lubricants andis the direct responsibilityof Subcom-mittee W 2 . H on Liquified PetroleumGas.

Current edition approved M arch 15, 1991. Published July 1991. Originallypublished as D 27 13- 68 T. Last previous editionD 2713 - 86.

2 Annual Book of ASTM Standards, Vol 05.01.

5. Apparatus5.1 Propane Water Test Valve3-A specially constructed

and calibrated valve manufactured solely forthis test (Note2). The valve has two open positions, a wide open positionfor flushing, and a small preset flow position for testing.

NOTE -The propane water test valveis a precision instrument andit shouldbe so treated. It should notbe dropped, strainedin any way, or

disassembled, except to clean the filter in acco rdancewth the manufac-turers instructions. Valves suspected of being defective shouldbereturned to the manufacturer for inspection, reconditioning orrecalibration.

5.2 Stop Watch.5.3 Sample Cylinder, having a minim um capacity of11.4

5.4 Cloth, dry, clean.L (3 gai).

6. Sampling6.1 The sensitivity of moisture test measurements to

uncontrollable sampling errorsis such as to warrant con-ducting all important tests at the propane supply sourcerather than o n samples taken from the bulk supply. Referee

tests shouldbe conducted on the bulk supply.6.2 If the test cannotbe run by connecting the apparatusdirectly to the bulk propane supply, a sam ple canbe takeninto a sample cylinder having a m inimum capacity of11.4 L.In such cases, the sam ple shallbe taken strictly in accordancewith direc tions given in PracticeD 1265.

7. Procedure7.1 Connect (Note3) the propane water test valveto the

liquid line of the bulk produc t source or to the liquid phaseconnection of the sample cylinder described in6.2, so thatthe body of the valveis horizontal and the outlet opening isaimed vertically upward. The valve shouldbe positionedsothat the internal surfacesof the outlet opening are clearlyvisible to the operator. Open the main valve on the samplesource (Note4) and set the valve on the test apparatu s in thepurge position. Purge the sample line and th e app aratus for15 s. Close the test valve for2 or 3 s, open it for2 or 3 s, closeit for 2 or 3 s, and continue this intermittent opening andclosing untila uniform frost cover has accumulated on thehousing around the outlet of the test valve. Snap the valveclosed to the test position and simultaneouslystart the stopwatch. Stop the watch at the instant the liquid propaneceases to flow through the valve (Note4).

A list of supplies of LPGasfreeze valves is available fromASTM.



-- ` ,, ` ,,, ` , ,,,,, ,, ` - - ,, ,, , ,, ---


34/56

S T D - G P AS T D 2140-ENGL 1997 3 8 2 4 b 9 9 0018528 157

4mD2713NOTE3-Use only clean, dry pipeor metallic tubing for makingthis record the average time for three consecu tive determ inations

as he freeze-time.onnection.Do not use mbber hose or plastic lined hose.

7.2 Sample pressure, at the inlet to the test valve, must notbe more than 100 psi (690kPa) above the vapo r pressureofthe product at the sample temperature. When sample sourcepressure is above this limit a liquid propane pressureregulator mustbe used to hold th e pressure, at the inlet to thetest valve, within this limit.

NOTE-The instant liquid propane stops flowing through the valve,frost will form on the internal surfaces of the valve outlet. The watchshould be stopped at this instant. The operator is forewarnedof thisinstant of stoppage if he will watch the frost line climb androll over thelip of the valve outlet.

7.3 Disregard the observed time for the initial freeze-offmn. uickly wipe the test valve outlet threads witha clean,dry cloth. Open the test valve to the purge position for ab out15 s to ensure the removal of ice from the preset opening.Repeat the operation sas described in the trial run until threesuccessive freeze-off times check each other to within+2 s.For freeze-off times of less than1 min this may requireasmany as seven or eight test runs.If the freeze-off time onthree consecutive test run s is3 min or greater, the test on th eproduct can be discontinue d and the freeze-off time recordedas greater than3 min. For freeze-off times shorter than3 min

NOTE5-Failure to purgethe apparatus with the valve opento thepurge position for about i 5 s between tests will give erroneous results.Purging assures that ice formed in the preset opening in the precedingtest wll be removed.

8. Report

productas pass.

9. Precisionand Bias9.1 In the case of pass-fail data or results from other

qualitative tests, no generally accepted method for deter-mining precision o r bias is curren tly available.

NOTE -Data froma series of tests conducted in 1967 indicate thefollowing: At moisture levels of 14 and26 ppm, all valves can beexpected to give freeze times over3 min; at a moisture level of 49 ppmit can be expected that none of the valveswill give freeze times overI8s; at a moisture level of93 ppm it can be expected that none of thevalveswiil give freeze times ov er5 s.

9.2 Bias-The procedure in Test MethodD 2713 formeasuring drynessof propan e has no bias because the valueof dryness is defined only in term s of this test m ethod.

10. Keywords

8.1 If the valve does not freeze off within60 s, report the

10.1 dryness; propan e.

The American Society for Testing and Ma terials takes no position respecting the va lidity of any patent rights asserted n connectionwith any item m entioned in this stand ard. Users of this standard are expressly advised that determ ination of the v alidity of any suchpatent rights, and the rlsk of infringement of such rights , are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years andit not revised, either reapproved or withd rawn . Your comments are invhed either for revision of this standard or for additional standardsand should be addressed to ASTM Headquarters . Your comnients w ill receive careful consideration at a meeting of the responsibtetachnical committee, which you m y attend. /i you feel that your comments have not received a fair bearing you should make yourviews known to the ASTM CommMee on Standards, 1976 Race St., hilade lphia, PA 19703.



--````,,```,,,```,`,,,,,,,````-`-`,,,,`,`,,`---


35/56

S T D * G PA S T D 2 1 4 0 - E N G L 1777 m 3 8 2 4 b 7 7 0038527 075 m

TbDesignation:D 2784 - 92 An American National Stand ardStandard Test Method forSulfur in Liq uefied Petroleum Gases (Oxy-Hydro gen B urnerorLamp)'

This standard is issued un der the fixed designation D 2784: the numb er immediately following the designation indicates the yearoforiginal adoption or. in the caseof revision. the yea r of last re\ ision.A numb er in parentheses indicates the yearof last reapproval.Asuperscnpt epsilon e ) indicates an editorial change since the last revision or reapproval.

1 . Scope1 . 1 This test method covers the determination of total

sulfur in liquified petroleum gases containing more than1pg/g. Specimens should not contain more than 100 pg/g ofhalogens.

1.2 To attain the quantitative detectability that themethod is capable of, stringent techniques must be employedand all possible sources of sulfur contamination must beeliminated.In particular. cleaning agents, such as commonhousehold detergents which contain sulfates, should beavoided.

1.3 The values given in acceptable metric units are to beregarded as the standard.

1.4 This standard does not purport to address all OJ thesafety problems, if any. associated wirh its use. I t is theresponsibilitj. of rhe user oft hi s srandard to establish appro-priare safer , and healtli practices and d etermine the app lica-bilitj3 of regitlatorj, litnirations prior to use.

2. Referenced Documents

2.1 .4STM Slandards:D i56 Test Method for Saybolt Color of Petroleum

Products (Saybolt Chromom eter Method)'D 1265 Practice for Sampling Liquefied Petroleum(LP)Gases'

D 1266 Test Method for Sulfurin Petroleum Products(Lam p Method)'

D 1657 Test Method for Density or Relative DensityofLight Hydrocarbons by Pressure Thermohydrometer'

E I l Specification for Wire-Cloth Sieves for TestingPurposes'

2.2 Institute of Petroleum S~andard."IP I8 Sampling Petroleum Gases, Including Liquefied

Petroleum Gases

3. Summary of Test Method3.1 The sample is burned in an oxy-hydrogen burner,or

in a lamp in a closed system in a carbon dioxide-oxygenatmosphere. The latter is not recommended for trace quan-

I This test method i s under the jurisdiction of ASTM Committee D-2 onPetroleum Products and Lubricants and is the direct responsibilityof Subcom-mittee W2.03 on Elemental Analysis.

Current edition approved MarchIS. 1992. Published May 1992.Onginallypublished asD 2784 - 69 T. Last previous edition D2784 - 89.

Anniral Book ofAST.11 Siandurds. Vol 05.01.' 4nniral Book O(AST.11 Slandards, Vol 14.02.

Available from Am encan National Stand ards Institute. W . 42nd St.. 13thFloor. New York. NY 10036.

tities of sulfur due to th e inordinately long combustion timesneeded. The oxidesof sulfur are absorbed and oxidized tosulfuric acid in a hydrogen peroxide solution. The sulfateions are then determ ined by e ither of the following finishes:

3.1.1 Barium Perchlorate Titration-The sulfate istitrated with barium perchlorate using a thorin-methyleneblue mixed indicator.

3. .2 Turbidimetric-The sulfate is precipitated asbarium sulfate and the turbidity of a suspension of theprecipitate is measured with a photometer.

4. Significance and Use4.1 It is important t o have the sulfur contentof liquefied

petroleum gases at low enough concentration to meetgovernment regulations. The presence of sulfur can result incorrosion of metal surfaces. Sulfur can be poisonous tocatalystsin subsequent processing.

5. Apparatus5.1 Oxy-Hydrogen CombustionAssemblj-The two types

listed below are recommended. Any combustion apparatusgiving equivalent results, however, is satisfactory.

5.1.1 Wickbold-Tjpe Combustion Apparatirs.sas shown

in Fig. 1 .5. .2 Modtjed Beckman Burner-Type Apparatus? asshown in Fig. 2. Each of the above types of apparatus shallconsist of three parts: atomizer-burner, combustionchamber, and absorber with spray trap.A blowout safetyport in the combustion chamberis desirable. The remainderof the apparatu s shall consist of a suitable support stand withthe necessary needle valves and flow meters for precisecontrol of oxygen, hydrogen, and vacuum.

5.1.3 Safety Shield-A transparen t shield shall be used toprotect the operator in the event an explosive mixture isformed in the combustion chamber.

5.2 Apparatus for La mp Cornbusrion:5.2.1 Absorbers, Chimneys, and Spray Traps,as required

are described in detail in AnnexA3 of Test Method D 1266.5.2 .2 Mangold System,consistingof a vacuum manifoldwith regulating device, valves. etc. (Fig.2 of Test MethodD 1266) and a dual manifold (burner and chim ney) sup-plying a gas mixture of approximately70 % carbon dioxide(CO,) and 305 oxygen O,) at regulated pressures. The gas

Available from Koehler Instrumen tCo., h i . . 1595 Sycamore Ave.. Bohemia.NY 117 16. with an all-stainless steel burner . or from Atlas Instrumen tCo.. 8902 E.1 I ih St.. Tulsa, OK. For the latter. Hoke No. 993 combination flow meter-vaheassemblies should be substituted for those supplied.

vailable from Scientific GlassblowingCo.. P.O. Box 18353. Houston. TX77023.



--` ` ` `

, ,` ` `

, , ,` ` `

,`

, , , , ,`

, ,` ` ` ` -` -`

, ,`

, ,`

,`

, ,` ---


36/56


37/56

NLET

FIG. 2 Trace Sulfur Apparatus Flow Diagram

0.05 Circu lor HoleP i o n e Surfoce)

4 6 k

O 0Alldimensions in millimetres

FIG. 3 Blast -Type Gas Burner

8. Procedure for Combustion of Sam ple8.1 Con nect the sample cylinder with stainless steel tubing

to the g a s expansion valve. Attach to this a nother sectionofstainless steel tub ing which runs to the v icinityof the burner.Make the final connection to the burner with sulfur-freerubber tubing. Wrap the expansion valve with heating tapeand connect this to a variable transformer. Inserta ther-mometer between the heating tape and expansion valvesothat the therm om eter bulb is in con tact with the valve body.See Fig. 4.

8.2 Turn on the variable transfomer and allow theexpansion valve to reach43C (1 10F). Alternatively theexpansion valve maybe placed in a suitable metal beakerand covered with water maintained at 11O"F.

8.3 Oxy-Hydrogen Combustion-Assemble the appa ratusaccording to the manufacturer's directions (see also14.1).Add to the absorber 25 mL of the hydrogen peroxidesolution.

NOTE : Warning-Before attempting subsequen toperations, theoperator should ( 1 ) be aware of the various hazards that can exist

S t a i n l e s s SteelR e l i e f Va l v e

Attach Burner

Here

S t a i n l e s s S t

C o n n e c t i o ns

PR e g u l a t i n g Val v e M a i n t a i n e d a t\ I O O F by H e o t i n g To p e

(4-f L e n g t h , l 4 0 W, 4 5 V )C o n t r o l l e d b y a Va r i a b l e T r a n s f o r m e r

FIG. 4 Burner Assembly for LPG



- -

` ` ` `

, ,

` ` `

, , ,

` ` `

, `

, , , , ,

` , ,

` ` ` `

- `

- `

, ,

` , ,

` ,

` , ,

` - - -


38/56

~

S T D - G P AS T D 23LiO-ENGL 3777 9 3824b77 0038532 b8T 9

4 ~ 2 7 8 4

through he improper use of hydrogen as a fuel, and 2) Recaution-have the safety shield in place.

8.3.1 Light the burner and insert into the combustionchamber. If necessary, readjust gas flows. Open the bottomvalve of the sample cylinder. Slowly open the expansionvalve until an optim um burning rate is achieved.

8.4 Lam p Combustion-Add to the absorber25 mL of

hydrogen peroxide solution. Assemble the chimney, ab-sorber, and spray trap and connect to theC 0 2 - 0 2andvacuum manifold. Make the necessary vacuum adjustments(see 5.1 of Test Method D1266). Set up a control blankabsorberas in 5.3 of Test Method D1266.

8.4.1 Open the bottom valve of the sample cylinder.Slowly crack thegas expansion valve. L ight the b um er withan alcohol lamp, and insert the burner into the combustionchamber (chimney).

8.5 Bum a quantity of sample in accordance with Table1.NOTE-in burning materials with sulfur concentrations greater

than 50 pug, estrict sample sizes to give qua ntities that will not containmore than 250 pg of sulfur for the turbidimetric finish or more than 150pg for the barium perchlorate finish. Alternatively, aliquots of theabsorber solutions which do not contain more than these maximumscan be used.

NOTE -Minor adjustment of the gas flow rates can be necessary tomaintain those recommended by the manufacturer.

8.6 After a suficient quantityof sample has been burned,turn the bottom valve of the cylinder off. Allow theremaininggas in the tubing andg a s expansion valve to b umitself out. Turn the heatoff on the g a s expansion valve.Disconnect the tubing from the sample cylinder and reweighthe sam ple cylinder to the nearest0.05 g. Leave the absorbersolution in the assembled lamp unit. The same absorbersolution will be used for the scavenger-rinse burning. Allowthe expansion valve to cool to ambient temperature.

8.7 If the oxy-hydrogen burner permits, flush the tubingand valve with 10 mL of scavenger-rinse and bum withoutdisconnecting the tubing. Otherwise disconnect the tubingand bum in the normal liquid mode.For the lamp burningcollect the rinsings in a standard lamp sulfur (see TestMethod D 1266) flask. Insert a standard burner equippedwith a wick into theflask and carry out the combustionasdescribed in Section 7of Test Method D1266.

8.8 For he oxy-hydrogen burners, when all of the rinsingsare consumed, shut down the burneras recommended by themanufacturer.

8.9 After the rinsings have been burned in the lamp,remove it, turn off the C 0 2 - 0 2supply, and turn off thevacuum pump.

8.10 For oxy-hydrogen blank determinations bum a hy-drocarbon stock with a very lowor nondetectable sulfurcontent. Make at least twoof these prior t o th e analysis ofsamples with trace sulfur contents to en sure that t he blanksare small and constant. Sub tract from the total sulfur figuresany blankso obtained. The remaind er is the net m icrograms

TABLE 1 Sample Sizes

of sulfur from the sample. Likewise subtract any sulfurobtained in the lam p combustion blank from the total figure.

8.1 I Disconnect the spray trap from the vacuum line andthoroughly rinse the spray trap and chimney with about 35mL of distilled water, collecting the rinsings in the absorber.It is important that any materials clinging to these partsbetransferred to the absorber to avoidlow values for sulfur

content.BARIUM PERCHLORATE TITRATION FINISH

9. Reagents9.1 Ion-Free Water-Distill deionized water and store in

tightly capped, highdensity polyethylene bottles.9.2 Hydrochloric Acid, Standard Alcoholic(O. 1 M)-

Dilute 20 mL of aqueous 0.5 M HCl with 80 mL ofisopropanol.

9.3 Inhibited Thorin-Methylene Blue Mixed IndicatorSolulion-The indicator is made upas two solutions andthese mixed together in equal volumes once per week asfollows:

Solution A: 0.8 g thorin?0.29 g potassium bromate,water to make 500 mL,

Solution B O 16 g methylene blue,0.2 mL of 0.5 M HCl,water to make 500 mL.

9.4 FleisherS Methyl Purple Indicato r Solution.o9.5 Barium Perchlorate(0.005 M)-Dissolve 1.95g of

barium perchlorate trihydrate" in200 mL of water and add800 mL of isopropanol. Adjust the apparent pH to about3.5with perchloric acid, using a pH meter.

9.6 Perchloric Acid," 70 %.9.7 Sodium Hydroxide, Standard Solutions(0.03 M)-

Prepare by mixing7 parts of water with 3 parts of standardO.1 M sodium hydroxide (NaOH) solution. Concentrate400mL of 0.03 N NaOH solution by evaporating to30 mL, anddetermin e any sulfate present in accordance with AppendixA 1, Turbidimetric Procedure for Sulfate of Test MethodD 1266.If sulfateis found, corrections must be made for anysulfur introduced by the reagent in the alkali titrationfollowing comb ustion.

9.8 Methylene B lue.

10. Preparation of Working Curve10.1 Into separate 30-mL beakers pipet eachof the

aliquots of the stan dard sulfate solution given in Table2. See6.3. To each aliquot add sufficient water to m ake3.4 mL, 12mL of isopropanol (total volume15.4 mL) and 3 drops ofmixed thorin-methylene blue indicator solution. Titrateasindicated below.For each sulfur level given in T able2, titratethree of the corresponding aliquots. Plot the millilitres oftitrant used versus micrograms of sulfur. Draw the beststraight line through points. Check a t least two points on thecurved at least every10 days.

~~

S a m p Size, gSuMirContent, ppm

TurbidimetricFinish BariumPerchlorate Finish

1 to 55 to 10

10 to 50

4520

5

3010

3

Available from Hach Chemical Co., A m a . IA.'OAvailable from Reisher Chemical Co., Benjamin Franklin Station.

''Avalable from G. Frederick, Smith Chemical Co., P O ox 23344,Washington, DC 20044.

Columbus. OH 43223.



--` ` ` `

, ,` ` `

, , ,` ` `

,`

, ,

, , ,`

, ,` ` ` ` -` -`

, ,`

, ,`

,`

, ,` ---


39/56

TABLE 2 Preparation of WorkingCurve~~ ~

Sulfur, gg 40 80 120 240 300AligUOtS,rnL 0.40 0.80 1.20 2.40 3.00

I I . Procedurefor Analysisof Solutions11.1 Quantitatively transfer the absorber contents to a

500-mL Erlenmeyer flask, using ion-free water for rinsing.Add 2 drops of Fleishers methyl purple indicator so lution tothis solution and titrate to a faint green end point with 0.03M NaOH solution (Note 12). Addi mL more of the 0.03MNaOH solution to the solution and reduce the volume to 2 to3 mL by evaporation on a hot plate in sulfate-free environ-ment. (Warning-see Note 11):DO NOT BOIL DRY . Coolthe solution to room tem perature and m easure its volume ina 10-mL graduate (Note 12). Adjust the volume to 3.0 mL byadding ion-free water.

NOTE10: Warning-Do not boil.NOT E I-The volume of sodium hydroxide should not exceed 2

mL. More indicates that the sulfur or halogen content is excessive orthat there is a serious air leak in the apparatus.

NOTE12-For high or completely unknown sulfur contents, theconcentrated absorbent can be quantitatively transferred to a 5-mLvolumetric flask, adjusted to 5 mL, and aliquots used. Each aliquot isthen subsequently made up to 3 mL as in 1 . I . Continue as in I 1.2.

11.2 Transfer the absorbent to a 30-mL beaker, rinse thegraduate an d the 500-mL boiling flask successively with two6-mL portions of isopropanol, and add the rinses to thebeaker.

11.3 Pipet 0.40 mL of the standard sulfate solution (40 pgof sulfur) into the beaker. Add 2 drops of the thorin-methylene blue mixed indicator solution. Adjust the re-sultant gray-green color by adding 0.1M HC1 dropwise tothe solution until the color changes to bright green.

11.4 The 2-mL buret containing standard bariumperchlorate should have its tip positioned just below thesurface of the solution in the beaker. The solution must bestirred by a small baron a magnetic stirrer or with a smallpropeller stirrer.A white background and good white lightmay be helpful in obtaining a precise end point. Add thebarium reagent at a steady rate ofO.1 mL in 5 A ) s until theend po int is indicated by a rapid, though slight, color changefrom green to a bluish gray (Note 13). Shutoff the buret atthe point of greatest rate of color change (Note 14).

NOTE 13-It is helpful to match end point colors with solutionssaved from prior standardization titrations performed within the last 15min and weil stirred to prevent drop-out of the colored barium sulfate

precipitate. People having al ow

red-green color sensitivity find thatusing the blue light of Method D 156, sharpens the end point veryconsiderably.

NOTE 14-The end point can be. checked by again adding 40 pg ofsulfur (0.4 mL standard sulfuric acid) and retitrating to the end point.

11.5 From the w orking curve, find the total sulfur titratedto the nearest1 pg. Subtract the 40 pg added.

11.6 For blank determinations, repeat the operations in8.3 and 8.7, and bu rn a hydrocarbon stock with a very low ornondetectable sulfur content. Burn for the same length oftime as the sample in th e norm al liquid mode. Subtract fromthe sulfu r figures in 11.5 any blankso obtained. This is thenet micrograms of sulfur from the sample.

TURBIDIMETRIC FINISH

12. Apparatus12.1 Photometer-Preferably a spectrophotometer having

an effective band width of about50 nm, and equipped with ablue-sensitive phototube foruse at 450 nm , or alternativelyafilter photometer equipped with a color filter having amaxim um transmission at approximately 450 nm.

12.2 Absorption Cells, having an optical path length of5cm. With use the cells may become coated with a film.Toremove this film, wash the cells with a detergent, using asoftbrush. After cleaning, rinse thoroughly with water.

NOTE15-The procedure as written assumes an absorbance changeof about O. 10 for each 100 pg of sulfur in 50 mL of solution measured ina 5-cm cell. Photometers employing cellsof shorter optical paths will notgive the precision of measurement stated in this method.

chloride dihydrate a s specified in 13.2.

13. Reagents13.1 Alcohol-Glycerin Mixture-Mix 2 volumes of dena-

tured ethyl alcohol conformingto Formula No. 3A of theU.S. Bureau of Internal Revenueor ethyl alcohol (99% byvolume) with1 volume of glycerin.

13.2 Barium Chloride Dihydrate (BaCl, 2H,O)-Crystalspassing a 20-mesh (850-pm ) sieve and retained on a 30-mesh(600-pm) sieve conforming to SpecificationE 1 1.

NOTE16-The crystal size of the BaC1,.2H 20 is an importantvariable that affects the development of turbidity.

13.3 H ydrochloric Acid( 1+12)-Add 77 mL of concen-trated hydroch loric acid (HC1, relative density 1.19) to a1 Lvolumetric flask and dilute to the mark with water.

14. Calibration

14.1 Only by the most scrupulous care and attention todetails can reliable results be obtainedby this method. Beforeusing new glassware and thereafter as required, clean theglassware with concentrated nitric acid. Rinse three timeswith tap water and follow with three rinses of deionizedwater. Reserve the glassware for use in this method alone.

14.2 Into 50-mL volumetric flasks introduce,by means ofthe buret, 0.25,0.50, 0.75, 1.00, 1.50, 2.00, 3.00, and5.00mL of standard sulfate solution (1 m L= 100 pg S ) . See 6.3.Add 3.0 mL of HCl (1+12) to each flask, dilute to volumewith water, and mix thoroughly. Prepare a reagent blankstandard ina similar way, om itting the stand ard sulfate.

14.3 Pour the en tire contents of each flask into a 100-mLbeaker. Add by means of a pipet10 0.1 mL of alcohol-

glycerin mixture and mix for 3 minon the magnetic stirrer.Select a s tirring speed just below that which might cause lossof sample through splashing. Maintain this speed throughoutthe entire procedure.

14.4 Allow the solution to stand undisturbed for 4 min.Transfer to an absorption cell and measure the initialabsorbance, using w ater as reference.

14.5 Return the solution to the beaker and add 0.300.01 g of BaCl,-2H,O crystals, eithe r by weighing thisamount or by using the scoop. Stir with the magnetic stirrerfor exactly 3 min. Allow to stand for an additional 4 min,transfer to the cell, and again measure the absorbancerelative to w ater.

12.3 Scoop, capable of dispensing 0.30k 0.01 g of barium



- -

` ` ` `

, ,

` ` `

, ,

, ` ` `

, `

, , , , ,

` , ,

` ` ` `

- `

- `

, ,

` , ,

` ,

` , ,

` - - -


40/56

14.6 Following steps described in 14.3, 14.4, and 14.5,obtain a reagent blank reading by subtracting the initialabsorbance of the reagent blank standard from that obtainedafter addition of BaC 12.2H2 0. This reading should no texceed0.005.

14.7 Obtain the net absorbance for each standard bysubtracting the initial absorbance and reagent blank readingfrom the absorbance o btained in accordance with 14.5. Plotthe net absorbance of each standard against micrograms ofsulfur contained in 50 mLof solution and draw a smoothcurve through the points.

14.8 To detect possible shifts, check the calibration curvedaily by making single determin ations.

15. Procedure for Analysis of Absorber Solutions

15.1 Drain the absorber solution into a 250-mL beakerand quantitatively rinse the absorber collecting the rinsingsin the beaker.

15.2 Reduce the volume of the absorber solutionstoabout 25 mL by evaporation on a hot plate. Quantitativelytransfer the resultant solution to a 50-mL volumetric flask,rinsing the beaker with several small portions of water. Add 3mL of HCl (1+ 12) to the flask, make up to volume withwater, and mix thoroughly.

15.3 Into a 100-mL beaker pour the entire contents of the50-mL volumetric flask containing the solution tobe ana-lyzed. Proceed as directed in 14.3, 14.4, and 14.5.

NOTE17-Should the blank reading exceed 0.02 0, the precision

obtainable will be impaired. If so. make an analysis of the reagents aloneto determine whether the atmosphere or reagents are at fault. Place 30mL of the H202 i 5 %) in the 50-mL volumetric flask, dilute to themark with HCI ( 1+ 2 15 ) , and proceed as described in 14.6. If this reagentblank reading exceeds 0.010,results should not be Considered reliable.

15.4 Obtain the net absorbance of the analysis solution bysubtracting the initial absorbance a nd th e net absorbance forthe oxy-hydrogen combustion blank o r the lam p combustion(depending upon the apparatus used for combustion) fromthat obtained after the addition of BaCI,2H,O.

15.5 Convert net absorbance to micrograms of sulfur byusing the calibration curve.

16. Calculation

follows:

Sulfurcontent, pg/g = A / Wwhere:A = micrograms of sulfuras obtained in 1 1.6 or 15.6, andW = grams of sample burned.

sulfur.

grains of total sulfur per i o0ft3 as follows:R for propane) = 0.083sR (for butane) = O.11 1SR (for propane-butane mixtures)

where:R = grains of total su lfur per 100 ft3 of gas at 15.6'C (60'F)

S = sulfur content, pg/g, an dG = relative density of the m ixtu re at 15.6/ 15.6'C (60/60'F) .

NO TE 18-The derivatives of constants used in the above equationsare based o n the following properties of propane and butane:

Specific volume for propane (of the real gas at 60'F

Specific volume for butane (same conditions as

16.1 Calculate the amount of sulfur in the sample as

(1)

16.1. Roun d the result of the test to the nearest1 pg/g of

16.2 Alternatively calculate the concentration in units of

(2)

= S[0.366(G- 0.5077)+ 0.0831

and 0.10132 MPa (760 mm) Hg,

8.45I5

6.3120and 14.696 psia), ft3/lbof gas

above)

NOT E 19-If the relative density of the mixture is not known,determine it by Test Method D 1657.

NOTE 0-Multiply by 2.2883 to convert grains per cubic foot tograms per cubic metre. Multiply by 35.31 to convert grains per cubicfoot to grams per cubic metre.

17. Precision and Bias17.1 The precision of this test method has not been

determined. The responsible subcommittee is attempting toattract volunteers for an interlaboratory study.

17.2 The bias of this test method cannotbe determinedsince appropriate reference material containing a knownlevel of sulfur in liquified petroleumgases is not a vailable.

The American Society lor Testing and Materials takes no position respecting th e validity of any patent righ ts asseI1ed in conn ectionwith any item m entioned in this stan dard. Users ot this standard are expressly advised that determination ot the validity d ny suchpatent rig hts , and th e risk of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the r es p s i b ie technical commirree and must be reviewed every tiv e years andnot revis ed, either reapproved or w ithd rawn . Your comm ents are invit8d either for revision of this standard 01 or additional standards

and should be addressed to ASTM Headqu arters. Your comments w i// receive careful consideration at a meeting of the responsibletech nical comm ittee, wh ich you may atten d. If you feel that your comments have not received a lair hearing you should m ake yourviews known to the ASTM Commirree on Standards, 7976 Race Sr.. Philadelphia, PA 19703.



--````,,```,,,```,`,,,,,,,````-`-`,,,,`,`,,`---


41/56

S T D - G P AS T D 2 1 4 0 - E N G L 1997 3 8 2 4 b 9 9 0038535 377 =

GPA STANDARD 2174-93

OBTAINING LIQUID HYDROCARBON SAM PLES FOR ANALYSIS BY GASCHROMATOGRAPHY

1. S C O P E

1.1 The specific purposeof this method is to describe theequipment and procedures for obtaining representative samplesof natural gas liquids and the subsequent preparation of thosesamples for laboratory analysis by gas chromatography. Theprocedures described in this method may be used for obtainingsamples for analysis by me thods other than gas chrom atography.A S A M P L E P R O B E M U S TB E U S E D TO O B TA I N AREPRESENTATIVE SAMPLE.

2. O U T L I N EOF M E T H O D S

2.1 A h y d r o c a r b o n f l u i d s a m p l e i s t r a n s f e r r e d u n d e rpressure from a source into a sample container by one of the

following methods:

a. Floating Piston Cylinder Method

b. Water Displacement Method1 ( total H 2 0 removal- 80replaced by hydro carbons;20% displaced for outage)

c. Water Displacement Method2 (part ia l H 2 0 removal-70% replaced by hydrocarbons; 20% displaced for outage;10%remaining in cylinder)

at the bottom. It is strongly recomm ended that the samples beobtained under the supervision of a person knowledgeablei n the

phase behavior of hydrocarbon systems and experiencedin allsampling operations.

3.4 T h e s c o p e o f t h i s m e t h o d d o e s n o t i n c l u d erecommendationsfor the location of the sampling point in a lineor vesse l , a l though the impor tance of the proper sampl inglocation cannotbe overemphasized.

3.5 A cer ta in amountof information about a sample isnecessary before it can be intelligently handledin the laboratory.Essential information includes the sample source, sample date,cylinder identification, sample source pressure and temperature,ambient temperature, typeof analysis required, and the sampling

method used. The re may be addit ional related facts such asfield-determined results and operating condit ions which willassist in the evaluation of the analytical data.This informationmust accompany the filled sam ple cylinder.

3.6 If the hydrocarbon fluid samples are to be transported bycommon carrier within the United S tates, the sample containersmust meet the specifications of and be labeled and packagedacco rd ing to t he Haza rdous Ma te r i a l s Regu la t i ons o f t heDepartmentof Transportation.

d . Ethylene Glycol Displacement Metho d ( to ta l g lycolremoval - 80% replaced by hydrocarbons;20 displaced foroutage)

3.7 This method assumes all procedures begin with clean,l eak - f r ee s ample cy l inde r s . Cy l inde r s mus t be t ho rough lycleaned prior to samp ling with an approp riate volatile solventorby fo l lowing manufac turers recomme ndat ions . The use of

NOTE 1: Methodsb, C, and d may not be aPP1icable to theanalysisof certain samples containing reactive non-hydrocarbons such assulfurcompounds, carbon dioxide, etc.

detergentlwater solutionsor steam is not recommended for thecleaning of floating piston cylinders.

3. G E N E R A L INFORMATION 3.8 For f loating piston cylinders, i t is desirablein mostcases to use an inert charge gas for the piston cylinder whichis not present in the sampleso that a leak in the cylinder itselfcan be easily detected during the analysis. The use of naturalgas as a back pressure fluid is not recommended because a leakacross the piston may result in compromising the sample withhydrocarbons (or other components) which were not actually

3.1 The o bjective of any sampling operation is to secure, ina suitable container, an adequate portion of a hydrocarbon fluidunder pressure, having the same co mposition as the stream beingsampled.

present at the source. A m ethod to insure a leak-free cylinder isto pressure tes t both s ides of the p is ton cyl ind er. This i saccomplished by pressurizing the displacement chamber withinert gas to a typical operating pressure with valveC opened(see Fig. i ) . Next, close valvesC and D and note the pressure

3.2 Particular emphasis should be givento the necessity ofobtaining accurate, representative samples for analysis sinceanalyses, regardlessof the care and accuracy of the laboratorytests, may be useless if the samples are not valid.

reading-on gaug e N. After a period of five minutes, a decreasein pressure as indicated at g auge N and a corresponding increasein pressure at gaugeM would indicate a leak across the piston.A pressure dro p only at g aug e N indicates a leaky fitting, valve,gauge, or rupture disc. Th e inert gas should be removed fromthe displacement end and the above procedure repeated for thesample side. It is also possible to analyze the inert gas sideof

3.3 It is not possible, nor is it the intent of this method, toprovide a procedure that wil l be applicable for al l samplings i t u a t i o n s . T h e s a m p l e s o u r c e h e r eis a s s u m e d t o b e ah o m o g e n e o u s , s i n g l e - p h a s e l i q u i d . A l l s a m p l e s m u s t b eobta ined us ing a probe des igned to secure product f rom thecenter one-third of the f lowing s tream. The loca tion of theprobe fitting should be on the topor side of the line rather than



-- ` ,, ` ,,, ` , ,,,,, ,, ` - - ,, ,, , ,, ---


42/56

SAMPLESOURCE

PROBE H

R 1L

SAMPLE

FLOATINGPISTON

INERT GAS

IYLINDER+

INDICATOR

N

ROD

Figure 1. Typic al Visual Indicator Samp ling System

the cylinderto determine hydrocarbon concentrationif leakageoccursin that direction. Cylinders should be tcstcd at least on anannual basis.

3.9 Carc mustbe exercised when sampling liquids having avapor pressure higher than atmospheric pressureto prevent

flashingof lighter components when transferring product fromthe source to a sample cylinder,or from a primary samplecylinder to a secondary cylinder. For floating piston cylinders,precautions should also be taken to ensure that the inert gaspressurein the piston cylinder never drops below the samplingpressureor the products vapor pressure at the existing sourcetemperature, thus preventing flashingof the sample.Theoretically, the productin the vapor s tate couldbe returnedtoliquid phase, but the physical restrictionsof the piston cylinder(e.g., rupture disc, valve dead space, and stirring assembly

design) may preventthe flashed product from returningLO theoriginal homogeneous mixture.

3.10 Duplicate Samples

3.10.1 When resampling is difficult or impossible,it is

advisable to take duplica te samples as a precaution againstaccidentalloss. These duplicate samples mustbe taken usingthe same sampling method and from the same sampling locationThe sample sequence shouldbe noted on thc sample informationt&

4. APPARATUS

4.1 Sample Containers



--` ` ` `

, ,` ` `

, , ,` ` `

,`

, , , , ,`

, ,` ` ` ` -` -`

, ,`

, ,`

,`

, ,` ---


43/56

S T D * G P ASTD ZLqO-ENGL 1777 3824b77 0038537 L b 3

4.1.1 Floating Piston Cylinders

4.1.1.1 The containerrequired for this method is constructedof metal tubing, honed and polishedon the inside surface. Thecylinder is preferably closed with removable end capsto provideaccess to remove and service the moving piston. The end capsare drilledand tapped for valves, gauges, and relief valves. Thecylinder is designed consistent with the maximum pressureanticipated during sampling and tobe nonreactive to materialsbeing sampled, the pressurizing fluid,the cleaning solvents andthe expected corrodants. The volume of the cylinder willdepend on the amount of sample needed for the laboratoryanalysis.

4.1.1.2 The cylinder itself contain s a moving pistonequipped with O-rings, teflon rings, or other devicesto effect aleak-free seal between the sample and the pressurizing fluidwhile allowing it to move freelywithin the cylinder. The use ofguide rings is recommended to assure smooth piston travel. Thepiston and sealing device must be non-reactive tothe sample, thepressurizing f lui d, the cleaning solvents, and expectedcorrodants.

T I O R Some types of lubricating fluid s will absorbappreciable amounts of the Cd + fraction from the gas beingsampled, thus compromising the integrityof the sample. It isrecommended that non-absorbing lubricating greases, such as,but not limited to DuPont s Krytox AC orAD, e wed to preventthis difficulty. Further details regarding non-absorbinglubricating greases can be obtained from the piston cylinderm a nufac turers .

4.1.1.3 All valves and safety devices must meet theappropriate material and pressure requirements for safe design.The pressure reliefs may be of spring or rupture disc types.These allow apartial or completeloss of contents due to thermalexpansion or over-pressurization.Should relieving occur, thesample shouldbe considered compromised.

4.1.1.4 Some piston-type cylinders are fabricated from non-magnetic materials such as the 300 series stainlesssteel. Thepiston likewiseis fabricated of stainless steel but has magnetsattached to the precharge side of the piston. Asthe piston movesthe'length of the cylinder, the magnetic field generated by themagnets flips a series of bi-colo red flags. This system , or

systems of similar configuration, indicates the piston positionand the volume of product in the cylinder.

4.1.1.5 Some piston-type cylinders are fabricatedwith a rodattached to the piston which extends through the end cap on theinert gas back pressure chamber with appropriate sealing devicesto prevent the inen gas from leaking. The traveling rod providesan indication of the piston position and the volume of theproduct sample in the cylinder.Again, some variations of thisstyle may exist.

4.1.1.6 Othertypes of floating piston cylinders are availablewhich haveno visuai method of determining the sample volumedirectly. For these cylinders, a displacement cylinder, cylinderY Fig. 2), is fabricated from metal tubing. This cylinder mustbe designedto meet the same pressure requirements as the pistoncylinder and havea volume of no morethan 80 percent o thepressurizing volume of the piston cylinder.

4.1.2 Double Valve Cylinders

4.1.2.1 Metal sample con taine rs of a typewhich insuremaximumsafety and whichare corrosion resistantto the productbeing sampled should beused. Stainlesssteel containers arerecommendedto minimize problems of surface adsorption ofheavy components (hexanes and heavier components) andtominimizethe reaction of carbon dioxide or other contaminantswith the container. Sample containers and valves must have aworking pressure equal to or exceeding the maximum pressureanticipated in sampling, storage, or transportation of the samplecontainer . Soft-seated valves are preferable to thosehavingmetai-to-metai seats. Thesize of the container depends upon theamount of sample required for the laboratory tests to beperformed.

NOTE 2: DOT regulations regarding the use of pressure reliefdevices on these cylinders must be followed.

4.2 Sample Transfer System

4.2.1 Transfe r lines , valves, and gauges in the transfersystem shall be designed consistent with maximum anticipatedpressure andbe resistant to all expected corrodants. (Stainlesssteel is preferred.) The transfer lines should have a minimumdiameter of 114 inch (6.35 mm) andbe as short as is practical.The useof filters and dryersis discouraged.

4.3 Composite SamplingMechanism

4.3.1 A composite sampling mechanism is a devicewhich isused to obtaina representative samplefrom a flowing productstream over a given periodof time. The unit consists of asample probe, eithe r a flow-through sample injection valve,whether automatic or manual(see Figures 3 and 4)or a probe-mounted sample pump (Fig.5) and utilizes a floating pistoncylinderas its collection chamber. The system must include amethod to mix the product samplein the collection chamber.This is necessary regardlessof whether the originalfloatingpiston cylinderis removed andused for laboratory analysis oratransfer of the product sampleis made into a secondary cylinderas described herein. (Figure6 depicts a typical automaticsamp ler and the various parts required. Figure5 depicts atypical proportional sampler using an injection pump and thevanous parts required.)



- -

` ` ` `

, ,

` ` `

, , ,

` ` `

, `

, , , , ,

` , ,

` ` ` `

- `

- `

, ,

` , ,

` ,

` , ,

` - - -


44/56

U M P L E

PISTON

INERT GAS

h

AIR(ATMOSPHERIC

PRESSURE)

CX NDERY

4.4 Sample Filters

4.4.1 The sample filter is an optional deviceused to protectthe sampling valve from scoring dueto h e presence of foreigncontaminants suchas metal shavings, dirt, etc., in aNGLsample. The filter should beof a small total volume,of aninline-type design, and containa replaceable/disposable element.

Figure 2. Displacement Cylinder Sampling System

5. SAMPLING PROCEDURES

5.1 Floating Piston C ylinder Method

5.1.1 Technique of using a cylinder equipped with visualindicator or liquid phase samples of spot or instantaneoustype (see Figure 1 )

CAUTION; A filter may introduce error if not handledprop erly. The filt er should be clean and fre e of any residualproduct fro m previous samplesso that a buildup of heavy-endhydrocarbon components doe s not result. (Th is may beaccomplished by a heatinglcm ling process or inert gas purge,etc .) The filter element should be15 micron size or larger sothat during the purging process NGL is not flas hed , causingfractionation and bubble formation.

5.1.1.1 With sample side of cylinder evacuated(fromcleaning operation) and valve C open ,fill displacement endwithinert gas to a pressure at least10 psi (69 P a ) above samplingpressure. Close valveD.

5.1.1.2 Open valveA at sample source and thoroughly blowoutany accumulated material. Close valveA at sample source.

40

right Gas Processors Associationded by IHS under license with GPA Licensee=UK LOCATION/5940240005


- -

` ` ` `

, ,

` ` `

, , ,

` ` `

, `

, , , , ,

` , ,

` ` ` `

- `

- `

, ,

` , ,

` ,

` , ,

` - - -


45/56

I

r ??- -+- - - - 4 7 3

i

- /T R U M SAMPLE-PROBE

Figure3. Typic al Sample Prob e Installation

on Orifice Flange5 .1 .1 .3 Connect piston cylinder to sampling source at

valve A.

5.1.1.4 With valvesB and C closed , open source valve A tofull open position. Observe sample pressure on gaugeL. Crackvalve B and fitting at valveC to purge line. Do not allowpressure L to drop below original sample pressure. Discontinuepurging after a sufficient time and only when liquid productispresent. If the product flashes without leaving a liquid residue atvalve B and the fitting at valve C, the operator must usejudgment in determining when to discontinue purging. Close

valve B and tighten fitting at valveC.

NOTE 3: If the diameterof the indicator rod is of significant sizecompared to the piston, then thepressure on the inertgas side will beslightly higher than the sample side,Le., gauge N will read higher thangauges L and M. This comment appliesto all subsequentcommentsregarding equal pressures on gaugesL. M and N.

5.1.1.5 Adjust pressure on gaugeN to equal pressureL byreleasing adequate inert gas through valve D.

5 . 1 . 1 . 6 With valve D closed, s lowly open valveCcompletely. There shouldbe no pressure change on gaugeNand pressure at gaugesL, M and N should allbe equal.

5.1.1.7 Partially open valveD, lowly allowing inertgas tovent to the atmosphe re. To prevent flashing, do not allowpressure M to drop below sampling pressure. Continue theoperation until the indicator designates the cylinder contains80% by volume of product (following manufacturer?sinstructions).

i I~~ ~~ ~

Figure4. Typic al Samp le Prob e Installation

for a Pump5.1.1.8 Close valves D, C, and A in that order. Open valve

B to release pressureon t h e sam ple line. Disconnect thecylinder from sample source.

5.1.1.9 Do not take outage or reduce pressure on pistoncylinder. Check valves C and Dfor leaks, cap valves to protectthre ads , prepare sam ple information tag and box fo rtransportationas per Department of Transportationor applicablerequirements.

5.1.2 Technique of using cylinder not equipped with visual

indicator fo r liquid phase samples , via displacement cylinder,of?spot? or ?instantaneous? type (see Figure 2)

5.1.2.1 With sample side of the piston cylinder evacuated(from cleaning operation) and valve C open,fill displacementend with inert gas to a pressure at least10 psi (69 kPa) abovesampling pressure.Close valve D.

5.1.2.2 Open valve A at sample source and thoroughlyblowout any accumulated material. Close valve A at sample sourcc.

5.1.2.3 Connect piston cylinderX to sample sourceat valveA and displacement cylinderY to piston cylinder. DisplacementcylinderY shouldbe filled withair at atmospheric pressure.Forproper cylinder sizing referto Section4.1.1.6.

5.1.2.4 With valvesB and C closed, open sample sourcevalve A to full open position. Observe sample source pressureon gauge L. Crack valveB and fitting at valve C to purge line.Do not allow pressureL to drop below original sample pressure.



-- ` ,, ` ,,, ` , ,,,,, ,, ` - - ,, ,, , ,, ---


46/56

Minimum5 Diameters

Productflow

meter

totalizer_ _ _

Center-stream/ sample probe

injection pump

f

1 P P ~ III

4 114 to J2

IPulse : tubing

dividercircuit

iI 1 puise per: "X" bbis. II

I

Power IAC - - - - - _- - -Power

Pressureregulator

Pressure

I

I Pressure

I

recomer

Check valve-ample transfer tapk

Bleed valve

Figure5. Typical Proportional Sampler Using an Injection Pump



--` ` ` `

, ,` ` `

, , ,` ` `

,`

, , , , ,`

, ,` ` ` ` -` -`

, ,`

, ,`

,`

, ,` ---


47/56

PRODUCTFLOW METER

; L - - i m T f pTOTALIZER

1 PPB iDIVIDER

1 PULSE PER '"x" BELS. iPOWER

NTERPOSER

ACPOWER

- - ) - - -

a-

ILOW-TH RU

SAMPLEINJECTION VALVE

- _ - -- - - iII

xK

PRESSURESUPPLY

INERT GAS

PISTON

SAMPLE

- -

STREAM3 SAMPLE PROBE

1 - FILTER A

SAMPLE MIXING

PRESSURERECORDER

TUBING

. SAMPLETRANSFER TA P

VALVE

DEVICE

Figure 6. Typical Autom atic Pr opo rtion al Sampler



- -

` ` ` `

, ,

` ` `

, , ,

` ` `

, `

, , , , ,

` , ,

` ` ` `

- `

- `

, ,

` , ,

` ,

` , ,

` - - -


48/56

Discontinue purging after a sufficient time and only when liquidproduct is present. if the product flashes without leaving aliquid residueat valve B and the fitting at valveC, the operatormust use judgment in determining whento discontinue purging.Close valveB and tighten fitting at valveC.

5.1.2.5 With valveE closed, open valve D and observepressureon gauge N. Adjust pressureN to equal pressureL byslowly venting inert gas through valveE. Close valveE.

5.1.2.6 With valveE closed, slowly open valveC to fullopen position. There shouldbe no pressure change indicated ongaugeN and pressure at gaugesL, M and N should allbe equal.

5.1.2.7 Close valve D. Open valve E and properly ventpressure through valveF. Close valveF.

5.1.2.8 Slowly open valveD, allowing inert gas fromcylinderX to flow into cylinderY. Observe gaugeM so as notto allow pressureM to drop. Continue operation until-pressureof all three gauges equalizes. At this point, a volume equal tocylinder Y has been displaced into cylinderX by the productsample. Sample cylinderX now contains80% by volume ofsample, leaving sufficient inert gas space to insure safe storageand transport.

5.1.2.9 Close valvesD, C, and A in that order. Open valvesB and F. Disconnect displacement cylinderY. Disconnectpiston cylinderX from sample source.

5.1.2.10 Do not take outageor reduce pressureon pistoncylinder X. Check valves C andD for leaks, cap valves toprotect threads, and prepare sample information tag and box fortransport as per Department of Transportationor applicablerequirements.

5.1.3 Technique of using floatin g piston cylinders fo r liquidphase samples of composite type via manual or automaticsystems (see Figures5 and 6 )

5.1.3.1 Automatic proportional samplers take small samplesof the flowing stream proportional to the flow rate. Timeincrements maybe used only when the flow rateis consmt.

5.1.3.2 Thorough purging of samp le lines, pumps and

connections to the sample cylinderis necessary to avoidcontamination of the samp le. Sam pler system s should bedesigned to minimize stagnantareas which could resultin thesample not being representativeof the fluid source.

5.1.3.3 Precautions shallbe taken to avoid vaporization insample loop lines when operating near the equilibrium pressureof the liquid. In some instances, it maybe necessary to insulatethe sample line and sample container,or control the pressureortemperatureof sample containers containing volatile materials.

5.1.3.4 All samples must be obtained using a probe. (Referto Section 3.3 for specifications and location.)A bypass devicethat causes a differential pressure, such asan orifice plateor asmall pump, is used to supply fresh liquidto a location forfilling the sample container, either manuallyor by use of sampleinjection valves. The bypass sam ple line utilizes short lines ofasmall diameter connecting the fresh sample source to thefloating piston cylinder receiver (see Figures 3 and4).

5.1.3.5 The floating piston cylinder should be connected tothe purged sample line via the composite sampling device. Inertgas pressureor appropriate pressurizing fluid should be appliedto the cylinder to force the piston to the starting end of thesampling cylinder. The inert gas pressure shouldbe maintaincdat a pressure which exceeds by200 psi (1379 kPa) theequilibrium vapor pressure of the fluid sampled under expectedvarying temperature conditions. At pressure differences lessthan 200 psi (1379 kPa) the chance forerroneous samplesincreases.

5.1.3.6 The automatic sample injection valveswill requireadjusting to obtain incremental samples at a rate such that thefloating piston cylinder will have adequate capacityto hold thesample during its period of sampling. Care shouldbe exercisedto prevent overfilling of the sample containerso that relievingdoes not occur, thus changing the composition of the sampleandcreating a serious safety hazard.

5.1.3.7 Increasingor decreasing the volume of sampledesired can be accomplished by changing the volumeof theincremental sample obtained with the injection valve and/or bychanging the frequency ofthe proportioning devicewhichcontrols the injection valve.

5.1.3.8 The sample obtained bythe sampler apparatusandstored in the floating piston cylinder mustbe mixed thoroughlyby either a mechanical deviceor a liquid pump and circulatingsystem,

5.1.3.9 The floating piston cylinder usedi n the abovesystem may be removed from the composite sampling deviceafter the desired product hasbeen sampled.

CAUTION: Under no circumtan ces should the cylindcrbefilled beyond 80 of ts capacity with sample.

5.1.3.10 Do not take outageor reduce pressure on thecylinder. Check valves for leaks,cap valves to protect threads,and prepare sample information tag and boxfor transportas perDepartment of Transportationor applicable requiremenis.

5.1.3.11 If it is not possible to disconnect the primaryfloating piston cylinder fromthe automatic system, the samplemay be transferredto a secondary floating piston cylinder. Oncethe samplein the primary cylinder is adequately mixed, proceed



--` ` ` `

, ,` ` `

, , ,` ` `

,`

, , , , ,`

, ,` ` ` ` -` -`

, ,`

, ,`

,`

, ,` ---


49/56

S T D - G P AS T D Z LV O - E N G L 1997 3824b99 0038543 4b5

as in step 5.1.1, treating the primary cylinder as a flowingsource. ( In this particular transfer situation,as sample iswithdrawn, the master sample container wil l part ial lydepressurize. The product vapor pressure althe existing mastersampler temperature mustbe maintainedor exceededto preventflashing.)

5 .2 WaferDisplacement and Ethylene Glycol DisplacementMethods - Total displacement

5.2.1 A double valve sample cylinder filled with cleanethylene glycol or water and a vesselto measure the displacedliquid are requiredfor these two methods (see Figure7) .

Ethylene glycolmust be the glycol t y p used. Water must havea pH between5.0 and 7.0. Water with apH greater than 7.0must be acidified to bring the pH within the proper range.(0.1N H2SO4is a satisfactory acid solution.)

5.2.2 The total volume ofthe sample cylinder must befound, then80% of the total volume shouldbe calculated. (Forexample, if the total volume of the cylinderis 500 ml, then 80%of the total volume is400 mi.)

5.2.3 Open valveA at sam ple source and thoroughly blowout any accumulated material. Close valveA at sample source.

L

SAMPLESOURCE

[ iYLINDER

rlRADUATEDCYLINDER

b

Figure7. Liquid Displacement Sampling



--` ` ` `

, ,` ` `

, , ,` ` `

,`

, , , , ,`

, ,` ` ` ` -` -`

, ,`

, ,`

,`

, ,` ---


50/56

S T D * G P A S T D 2 L 4 0 - E N G L1997 3 8 2 4 b 9 9 00118544 3 T L m

5.2.4 Connect the sample cylinder to sampling source atvalveA exactlyas shown in Figure7.

5.2.5 With valvesB, C, and D closed, open sample sourcevalve A to full open position. Observe sample pressure ongaugeL. Crack valveB and fitting at valve C to purge line.Donot allow pressureL to drop below original sample pressure.Discontinue purging after a suffic ient time and only when liquidproduct is present. If the product flashes without leaving aliquid residue at valveB and the fitting at valve C, the operatormust use his judgement in determining when to discontinuepurging. Close valveB and tighten fitting at valve C.

5.2.6 With valve D still closed, slowly open valve Cto fullopen position. Pressure at gaugesL and M should be equal.

5.2.7 Slowly open valveD to allow a slow discharge oftheglycol or water displacement liquid into the measuring vessel.To prevent flashing, do not allow pressureM to drop below

sampling pressure. Con tinue operation until80% by volume ofthe displacement liquid has been displaced by productin thesample cylinder.

5.2.8 Close valves D, C andA in that order. Open valveBto release pressure on sam ple line.

5.2.9 With sample cylinder still attached to source valveA,slowly open valve D to drain the remaining 20% displacementliquid from the sam ple cylind er. After15 ml has beenwithdrawn, the operator should reducet h e drain rate andcarefully extract the remaining wateror glycol. (Thedisplacement liquid may become cloudy just priorto the end

point, although the effect is less noticeable with glycol.)At thistime, reduce the rate to 10 ml/min maximum. Whenthe firstindication of product appears, usually a flashing bubble, closevalve D immediately with no furtherloss of product.Disconnect cylinder from sample source.

5.2.10 Check valves C and D forleaks, cap valves to protectthreads, prepare sample information tag and box forUansportationas per Deparunent of Transportationor applicablerequirements.

5.3 Waier Displacement Method - Partial Displacement

5.3.1 A double valve sample cylinderfilled with clean waterand a vesselto measure the displaced liquid are required for thismethod (see Figure 7). The water used mustmeet the samerequirementsas in 5.2.1.

5.3.2 The total volume ofthe sample cylinder must befound; then 70% and20% of the total volume should becalculated. (For example,if the total volume of the cylinder is500 ml, then 70% of the total volume is 350ml and 20% is 100mi.)

5.3.3 Open valve A at sample source and thoroughly blowout any accumulated material. Close valveA at sample source.

5.3.4 Connect the sample cylinderto sampling source atvalveA exactly as shownin Figure 7.

5.3.5 With valvesB, C, and D closed,open samplesourcevalve A to full open position. Observe sample pressure ongaugeL. Crack valveB and fitting at valve Cto purge line. Donot allow pressureL to drop below original sample pressure.Discontinue purging aftera sufficient time and only when liquidproduct is present. If the produ ct flashes without leaving aliquid residue at valveB and the fitting at valveC, he operatormust use his judgement in determining when to discontinuepurging. Close valveB and tighten fitting at valve C.

5.3.6 With valve D still closed, slowly open valve C to fulopen position. Pressure at gaugesL and M shouldbe equal.

5.3.7 Slowly open valve D to allow a slow discharge of thewater into the measuring vessel.To prevent flashing, do notallow pressureM to drop below sampling pressure. Continueoperation until 70% by volume of water hasbeen displaced byproduct in the sample cylinder.

5.3.8 Close valvesD, C, and A in that order. Open valveBto release pressure on sample line.

5.3.9 With sample cylinder still attached to source valve Aslowly open valve D to drain another20% of the water from thesample cylinder. (At this point, 10% ofthe water still remainsinthe sample cylinder.) Disconnect sample cylinder from sampl

source.

5.3.10 Check valves C andD for leaks, cap valvesto protectthreads, prepare sample information tag and box fortransportationas per Department of T ransportation or applicablerequirements.

6 . LIQUID PHASE LABORATORY HANDLING ANDPREPARATION

6.1 Recommended Proceduresfor Liquid Ph ase SamplesPrior to injection of the Sample into the Test Device

6.1.1 Floating Piston Cylinders

6.1.1.1 For floating piston cylinders refer to Figure8 andproceed as follows: Connect a source of inert gas to valveA sothat pressure may be applied to the sample via the floatingpiston. Apply a pressure not less than200 psi (1 379 kPa) abovethe vapor pressureof the sample at the temperature of thesample injection valve by opening the valve on the inert gasource ar d valve A. The inert gas pressure isto be maintained



--` ` ` `

, ,` ` `

, , ,` ` `

,`

, , , , ,`

, ,` ` ` ` -` -`

, ,`

, ,`

,`

, ,` ---


51/56

S T D n G PA S T D 2 1 4 0 - E N G L L997 3 8 2 4 b 7 7 0 0 1 8 5 4 5 238

I

on the sample container at all times during the analysis via aprcchargc regulator or similar device.

FLOATING PISTON

6.1.1.2 The sample should be thoroughly mixed per thecylinder manufacturers recom mendations , whetheri t be bymixing dasher, mixing ball or equivalcnt, immediately priorto each analysis.

6.1.1.3 Connect the sample end of the cylinder, valveB, tothc inlet of the chromatograph liquid sample valve.A l lconnections and tubing areto be made of material impervious tothe sample composition with the smallest diameterand shortestlength of plumbing practica l, thereby minimizing dead space.All tubing between the sample cylinder and liquid sample valveshould be the same diameter.

6.1.1.4 With valve C closed , open valve B tof i l l the samplevalve and associated lines.

6.1.1.5 Slowly crack valveC to purge the sample valve andassociated lines. Be sure sufficient volume is displaced to purgeunmixed areasi n the sample cylinder as wcll as the sampleinjection system. When the purge is com plete, close valve C.

CAUTION: Extreme care must be used to ensure that noflashing of sample occurs in the inlet sampling line and valvesystem. Any measurements should be taken al sample purgevalve C , never at sample valve B . The sample line and valvesystem should remain at 200 psi (1379 kPa) above the vaporpressure of the product.

6.1.1.6 Operate the liquidsample valve either manuallyorautomaticallyto inject the liquid sample into the carrier gas flowimmediately aheadof the chromatographic column. The sampleinjection valve shouldbe actuated quickly and smoothly to placethe sample onto the column all at once andto ensure continuouscarrier gasflow through the column.

6.1.1.7 When sample injection is complete, close valvesBand A in that order, then close valve on inert gas source. Slowlyopen valve C and vent any remaining unused samplein thesystem through an appropriate vent.

6.1.1.8 Disc onn ect the cylin der from the inlet of thechromatograph liquid sample valve and the inert gas source.

6.1.1.9 I f the sample is tobe reused, do not take an outage

1N G R q b [RESSURE EEDLE VALVE

GAS

RELIEFVALVE

PRESSURE

CYLINDERNEEDLE VALVE

GAS

RELIEFVALVE GAS

1 r

INERTGAS 1GL1

CARRIER CARRIER GAS

CHROMATOGRAPHLIOUIDALVEM P L I N G111.

CYLINDEROUTLETVALVE

RELIEFVALVE

NEEDLEVENT VALVE

o

Figure 8. Repressuring System and Chromatograph Valving w ith Floating Piston



--````,,```,,,```,`,,,,,`,,````-`-`,,`,,`,`,,`---


52/56

or reduce pressure on the piston cylinder.Check valves forleaks, cap valves to protect threads, andstore the cylinder in asuitable location.

6.1.2 Double Valve Cylinders

6.1.2.1 For doublevalve displacement cylinders, referto

Figure 9 and proceedas follows: Connect the sample cylinderYto cylinder X so pressurizing fiuid can enter into the bottom ofcylinderY. (The pressurizing fluid shouldbe same type usedtosecure the product.) With this configuration the hydrocarbonsample will be taken from the upper portion of the cylinder.Open valve A and pressurize cylinderX w i t h inert gas.Maintain a pressure atleast 200 psi (1379 kPa) above the vapor

pressure of the hydrocarbon sample at operating conditions.Open valvesB and C to admit pressurizing fluid into samplecylinderY.

6.1.2.2 Mix the sample thoroughly by gently invertingcylinder Y several times. Fix cylinderY in a vertical position bymeans of a ringstandor simila