EPA-AA-EOD/TPB-87/2
Technical Report
Evaluation of Several Methods to Measure
Volatility of Motor Fuels
July 1987
Marcus E. Haubenstricker, Ph.D.
Carl A. Scarbro
NOTICE
Technical reports do not necessarily represent final EPA decisions or
positions. Their publication or distribution does not constitute any
endorsement of equipment or instrumentation that may have been evaluated.
They are intended to present technical analysis of issues using data which are
currently available. The purpose in the release of such reports is to
facilitate the exchange of technical information and to inform the public of
technical developments which may form the basis for improvements in emissions
measurement.
Testing Programs Branch
Engineering Operations Division
Office of Mobile Sources
Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, Michigan 48105
-------
Abstract
The U.S. Environmental Protection Agency is proposing regulations to
reduce the amount of hydrocarbons released to the atmosphere due to the
evaporation of automotive fuels. The new regulations may define upper
volatility limits for these fuels based on seasonal climatic patterns.
Volatility of gasoline fuels is typically quantified by measurement of Reid
vapor pressure (RVP). Although an established procedure exists for the
assessment of this parameter (ASTM D 323), there is question as to the
accuracy of the procedure when evaluating fuels with water-interactive
constituents, e.g., alcohols and ethers. ASTM P 176 is a procedure which has
been proposed as its replacement. It addresses the problem of water and is
known as the "dry" version of ASTM D 323.
This correlation study was designed to evaluate the proposed ASTM P 176
procedure for repeatability and reproducibility. The results of this study
will help establish a basis for enforcement tolerances. Also evaluated was
the equivalency of results from two commercial instruments versus those
obtained by the traditinal "gauges and bath" technique.
Thirty-nine laboratories within the United States and Canada analyzed
duplicate samples of four fuels. The work was performed during the fall of
1986.
The results of the study indicate a level of precision similar to that
determined by ASTM in 1985. Precision may be improved by calibrations or
other enhancements to the technique. It was also determined that the results
from the commercial instruments were comparable to those from the gauges and
bath methods.
-------
-2-
Background
The U.S. Environmental Protection Agency has placed a limit on the lead
content of automotive fuels as part of a process to limit adverse automotive
emissions. To counteract the resultant decrease in octane, refineries have
increased amounts of lighter hydrocarbons in final fuel blends. EPA continued
its original actions by placing limits on other emissions products. As
exhaust by-products decreased, evaporative emissions became more noticeable
for two reasons: Evaporative emissions constituted a larger fraction of total
hydrocarbon emissions as exhaust hydrocarbon emissions decreased, and fuel
blends became more volatile due to the increases in amounts of lighter
hydrocarbon fractions.
In an additional effort to control evaporative emissions, EPA may develop
upper limits for fuel volatility. If adopted, these limits would likely be
geographically based on seasonal climatic trends.
Volatility is historically associated with the results of a vapor pressure
assessment using the Reid method. The American Society for Testing and
Materials (ASTM) adopted this test in 1930, and presently documents it in
their procedure D 323. A reference to Reid vapor pressure (RVP) implies this
procedure, which prescribes a closed cylinder and the determination of the
headspace pressure of a sample which has been heated from 32°F to 100°F. The
units of measurement are generally expressed in pounds per square inch (psi).
Because of various market economics, short-chained alcohols and ethers are
being used as fuel additives. In some cases, these additives comprise a
significant percentage of a fuel. Due primarily to chemical hydrogen bonding,
these substances interact with water, resulting in lower observed vapor
pressures. To prevent these potential inaccuracies, the RVP should be
determined under dry (non-aqueous) conditions.
In ASTM P 176, a proposed general specification for gasoline, ASTM has
outlined a new procedure for measurement of volatility. This procedure is
identical to ASTM D 323 except that safeguards are taken against inclusion of
water. ASTM conducted a correlation study in 1985 to determine the
statistical acceptance of the procedure. The study included 17 laboratories,
18 fuels, and 271 observations. The results of the study produced no
statistical difference between the "wet" (ASTM D 323) and "dry" (ASTM P 176)
methods.
Purpose
The purpose of this study is to evaluate the repeatability,
reproducibility, and accuracy of several methods to measure RVP as they are
applied at a number of laboratories throughout North America. The results
will be used in the development of regulations of fuel volatility.
-------
-3-
Program Design
Equipment and Methods:
In addition to the ASTM P 176 method using the traditional gauges and
bath, there was also a need to assess the capability of more automated
methods. Two types of instruments are in general use — the German-built
Herzog semi-automatic unit distributed by UIC, Inc., Joliet, IL; and a micro
semi-automatic unit developed and manufactured by Southwest Research Institute
(SwRl), San Antonio, TX.
Laboratories:
In order to obtain reasonable results for each of the several test
methods, a number of laboratories would need to participate. Twenty to thirty
was an early estimate. Laboratories which participated in the prior ASTM
survey were identified as candidates. Other laboratories known to EPA
personnel would also be invited to participate.
Fuels:
Pour fuels were chosen for the survey. Two of these are used at EPA's
Motor Vehicle Emissions Laboratory (MVEL) as test fuels. One was a high
octane, non-oxygenated, unleaded test gasoline with a nominal vapor pressure
of 9.0 psi and is used by EPA and the automotive industry as an emission
certification fuel. It was procured from Howell Hydrocarbons of San Antonio,
TX. The other test fuel was a "commercial" unleaded test gasoline procured
from Phillips Chemical Company of Bartlesville, OK. Its RVP was approximately
11.5 psi.
The third fuel was cyclopentane (90$ pure), procured through Curtin
Matheson (order number NCX2414-1). The theoretical RVP of this hydrocarbon
compound is 10.15 psi. This particular fuel component was chosen because of
its purity and therefore stable vapor pressure.
The fourth fuel was a commercial grade gasohol (nominally 10$ ethanol) and
was procured from a local service station. The vapor pressure of this fuel
was over 12 psi. The procured sample was stored in a closed drum at below
60°F until bottled.
Duplicate samples of each fuel were to be sent to each laboratory. The
containers were coded but the contents were not identified.
Sample Handling:
Sample handling was a critical factor in the program. Every effort was
made to ensure that samples of each fuel were identified and that they were
packed properly for shipment. Concern for possible changes in vapor pressure
during the process was addressed by a statistical analysis of results versus
the time between bottling and analysis.
-------
-4-
Data Processing:
The results from each laboratory were evaluated from the standpoint of
repeatability, reproducibility, and accuracy using accepted statistical
techniques.
Conduct of the'Program
During June and July, a number of laboratories were contacted. Almost all
agreed to participate. These contacts also resulted in unsolicited
responses. Ultimately, a total of 49 laboratories agreed to participate.
Most of the laboratories were equipped for the gauges and bath method.
Almost as many used a Herzog instrument, and a few had the SwRI unit. Three
offered to perform the analyses using a gas chromatograph. A listing of the
laboratories and their type of equipment is contained in Attachment A.
Attachment B is a copy of the letter used to describe the program and formally
solicit participation.
Acceptable containers for storage and shipment of the samples were
obtained from All-Pak Corporation of Pittsburgh. These shippers met DOT
exemption E-9168 for hazardous materials and were comprised of a one-quart
PVC-coated flint glass bottle with teflon cap, a foam-filled metal can to
contain the sample bottle, polypropylene tape to seal the bottle and can, a
plastic bag to further seal the can, a copy of the DOT exemption and a
specially constructed cardboard shipping box, and shipping tape to seal it.
The teflon-lined caps were replaced by poly-seal caps to ensure an even better
aeal.
The four fuels were bottled on October 7, 9, 16, and 21 and boxed for
shipment. Each type of fuel was bottled in the course of a single day by a
bottom-fill method through 1/4" tubing. Each bottle was filled to 15%
capacity and securely capped.
When the samples were packaged for shipment, a special seal (Attachment C)
was placed over the can before it was placed in the plastic bag. This was
done to direct the samples to certified analysts before being opened. Also
placed in sample box #1 was a copy of the ASTM vapor pressure procedure
(Attachment D), a data report sheet to be filled out by the analyst
(Attachment E), and a sample transfer tube fabricated for the one-quart sample
bottles.
A contract was awarded to United Parcel Service for shipping the samples
to laboratories in the United States. Shipments to Canadian laboratories were
arranged through Consolidated Preightways. All samples left MVEL by the end
of October. Overall, 520 samples were sent to a total of 49 laboratories.
Results began to return during the first week of November, and by the end
of November approximately 25 laboratories had responded. Ultimately, 39
laboratories submitted results which accounted for 80$ of the samples which
were sent out.
-------
-5-
Results
Attachments P, G, and H are listings of the results from each reporting
laboratory for the P 176, Herzog, and SwRI techniques, respectively.
Anonymity of the laboratories has been maintained by use of the code numbers.
Statistical analyses of the data for repeatability and reproducibility
began in mid-December. The methods employed were identical to the methods
used by ASTM on the original survey in 1985, and are found in "Manual of
Determining Precision Data for ASTM Methods on Petroleum Products and
Lubricants," (RR D-2-1007). According to ASTM:
"Repeatability is a quantitative expression of the random error
associated with a single operator in a given laboratory
obtaining repetitive results with same apparatus under constant
operating conditions on identical test material. It is defined
as the difference between two such results at the 95$
confidence level.
Reproducibility is a quantitative expression of the random
error associated with operators in different laboratories each
obtaining a single result on an identical test sample when
applying the same method. It is defined as the difference
between two such single and independent results at the 95$
confidence level."
After evaluating the data, it was observed that some laboratories reported
consistently different results than others. Assuming that calibration or some
other systematic difference was the cause, it was decided to normalize the
data based on the accepted RVP of the cyclopentane. The average value of the
raw data for cyclopentane for each laboratory was therefore divided by 10.15
psi. The ratio was then applied to the raw data for each fuel for each
laboratory and the results recalibrated. This normalization resulted in
generally improved values, for reproducibility. The summary of all results is
contained in Attachment I. The statistics for the calculations of
repeatability and reproducibility for each test method are found in
Attachment J.
Concerns about the effect of storage time on vapor pressure were addressed
by evaluation of vapor pressure vs. time as shown on Attachment K. For this
purpose, the total vapor pressure of all eight samples was plotted against the
number of days between the date of bottling and analysis date.
Discussion
The statistical results generated in this EPA survey are similar in
magnitude to the results of the 1985 ASTM RVP survey. Normalization to
cyclopentane improved the calculated reproducibility values for all three test
methods.
-------
-6-
In 1985, ASTM deleted data above and beyond the criteria used in their
RR D-2-1007. This is considered acceptable if the data is judged as
contributing to the majority of the variance for a sample or a lab. Deletions
did not extend beyond ASTM RR D-2-1007 directives in the present EPA aurvey.
Conclusions
The statistical calculations of the correlation data indicate similar
precision as determined by ASTM in their 1985 study for the same three dry-
methods (bath and gauges, Herzog, and SwRl). All of the studies indicate
worse precision than the ASTM published values for ASTM D 323 (repeatability =
0.25 psi, reproducibility = 0.55 psi). The precision does not change due to
the magnitude of the sample measurement, and the cyclopentane
displays the same variance as the other fuels. The variability appears to be
due to the methods of RVP analysis.
Based on the correlation coefficient and the slope of the best fit line,
there appears to be no relationship between storage time and vapor pressure.
Recommendations
Improvement due to normalizing the data indicates that differences in
technique used by the individual labs may have contributed to the
reproducibility of the results. Quality assurance techniques, such as
calibration using pure components, may enhance accuracy and reproducibility of
all the test methods and should be considered.
1818c
-------
ATTACHMENT A
EPA Vapor Pressure Correlation Testing Laboratories
Laboratory Number Manual P 176 Gas
Samples 'Tank Herzog SvRI Chrom.
Andre Jasmin 16
Ultramar Canada Inc.
CPP 2055
165 Rt. Des lies
St. Romuald, Quebec G6W 5M4
(418) 837-3641
E. Lyn Turner 8
AM Laboratories
6207 Tri-port Court
Greensboro, NC 27410
(919) 854-0747
Richard Yurek 8
E.W. Saybolt & Co., Inc.
400 Swenson Drive
Kenilworth, NJ 07033
(201) 245-3100
O.L. Salter 8
Charles Martin Analytical Laboratory
P.O. Box 1558
Pasendena, TX 77701
(713) 477-1542
Roger Penstermaker 8
Phillips Petroleum Co.
1186 Adams Bldg.
Bartlesville, OK 74004
(918) 661-3508
Harold Honaker 8
Ashland Petroleum Co.
Automotive/Product Application Lab
P.O. Box 391
Ashland KY 41114
(606) 329-5495
T. Hok Gouw 24
Chevron Research Co.
P.O. Box 1627
576 Standard Avenue
Richmond, CA 94802
(415) 620-2417
-------
ATTACHMENT A (cont.)
EPA Vapor Pressure Correlation Testing Laboratories
Laboratory Number Manual P 176 Gas
Samples Tank Herzog Swrl Chrom.
Leo Zafonte 8 1
GARB
Haagen-Smit Laboratory
9528 Telstar Avenue
El Monte, CA 91731
(818) 575-7054
Baron Munchausen 8 1
Litton Core Laboratory
P.O. Box 34282
8210 Mosley Road
Houston, TX 77234
(713) 943-9776
Bob Snavley 16 1 1
Koch Refining Co.
P.O. Box 2608
Corpus Christi, TX 78403
(512) 289-8511
Larry White 8 1
Diamond Shamrock Refining/Manufacturing Co.
P.O. Box 490
Three Rivers, TX 78071
(512) 786-2536
David Brown 16 11
Phillips 66 Company
P.O. Box 866
Sweeney, TX 77480
(409) 491-2265
Jack Gilliland 16 1 1
Chevron USA
Quality Control Laboratory
P.O. Boc 1272
Richmond, CA 94802
(415) 620-2100
C. Iverson 16 11
Phillips 66 Company
Refinery - NGL Laboratory
P.O. Box 271 SPUR 119
Borger, TX 79007
(806) 273-2831
-------
ATTACHMENT A (cont.)
EPA Vapor Pressure Correlation Testing Laboratories
Laboratory
Dave Pillon
Shell Canada Limited
Sarnia Manufacturing Centre
Refinery Lab
Corunna, Ontario NON 1GO
(519) 481-1301
Ian Taylor
Shell Canada Limited
Oakville Research Centre
3415 Lakeshore Road, West
P.O. Box 2100
Oakville, Ontario L6J 5C7
(416) 827-1141
Terry Boyle
Iringin Oil
P.O. Box 1260
St. Johns
Mew Brunswick, Canada E2L 4H6
(506) 633-3000
Calvin LaBauve, Jr.
Murphy Oil USA, Inc.
Meraux Refinery
P.O. Box 100
Meraux, LA 70075
(504) 271-4141
William Marshall
NIPER
P.O. Box 2128
Bartlesville, OK 74005
(918) 337-4345
Rick Schomaker
Oldsmobile Division
920 Townsend Street 40-2
Lansing, MI 48921
(517) 377-5019
Number
Samples
8
Manual
Tank
P 176
Herzog
SwRI
Gas
Chrom.
8
16
8
8
-------
ATTACHMENT A (cont.)
EPA Vapor Pressure Correlation Testing Laboratories
Laboratory
Rich Macintosh
Petro-Canada
Trafalgar Refinery
Oakville, Ontario L6J 5B5
(416) 825-1703
Jane Jackson
Alberta Research Council
P.O. Box 8330, Station P
250 Karl Clark Road
Edmonton, Alberta T6H 5X2
(403) 450-5100
Tony Novak
ARCO Petroleum Products Co.
1801 E. Sepulveda
Carson, CA 90745
(213) 584-8201
ftoy Sabourin
National Research Council of Canada
Montreal Road, Bldg M-9
Ottawa, Ontario K1A OR6
(613) 993-2186.
Jeff Mann
AC Spark Plug Division, CMC
1601 N. Averill Avenue
Flint, MI 48556
(313) 257-8246
Dr. Leo Duffy
AMOCO Corporation
Amoco Research Center
P.O. Box 400, Bldg 600
Naperville, IL 60566
(312) 420-5225
Number
Samples
16
Manual
Tank
P 176
Herzbg
SvRI
Gas
Chrom.
16
24
8
16
-------
ATTACHMENT A (cont.)
EPA Vapor Pressure Correlation Testing Laboratories
Laboratory Number Manual P 176 Gas
Samples Tank Herzog SwRI Chrom.
Michael Davis 8 1
Department of the Army
Corpus Christi Army Depot
Mail Stop 27
SDSCC-QLC
Corpus Christi, TX 78419
(512) 939-3555
Bob Pateraon 8 1
Corpus Christi Petrochemical Co
P.O. Box 10940
1501 McKenzie Road
Corpus Christi, TX 78410
(512) 241-6450
Charett Navarrete 8 1
Curtis & Tomplins Lits.
401 Canal Street
Wilmington, CA 90748
(213) 549-6727
Sue Frederick 8 1
EXXON Company, USA
P.O. Box 9000
E. 22nd Street
Bayonne, NJ 07002
(201) 925-6976
Edwin Calvin 16 1 1 0
EXXON Research and Engineering Co.
1900 E. Linden Avenue
Bldg. S105
Linden, NJ 07036
(201) 474-2516
Ted Eckman 8 1
General Motor Corporation
Milford Proving Grounds
Bldg 31
Milford, MI 48042
(313) 685-6032
Syd Feldesteen 8 I
Chrysler Corp; Fuels & Lube Lab
12800 Oakland Avenue
Bldg. 138-2W, Code 5830-u
CIMS 418-17-07
Highland Park, MI 48203
-------
ATTACHMENT A (cont.)
EPA Vapor Pressure Correlation Testing Laboratories
Laboratory Number Manual P 176
Samples Tank Herzog SwRI
Bill Okamoto 8 1
Ford Motor Co.
P.O. Box 2053
SRL Rm 3198
Dearborn, MI 48121
(313) 594-1016
R.L. Akers 8 1
Marathon Petroleum Co.
1300 S. Fort Street
Detroit, MI 48217
Paul Michaluk 8 1
Petro-Canada R&D
2489 N. Sheridan Way
Sheridan Park
Ontario, Canada
L5K 1A8
(416) 896-6805
R.E. Campell 16 1 1
Port Arthur Product Control Lab
Texaco USA
P.O. Box 712
Port Arthur, TX 77641
(409) 982-5711
Clark Ellison 8 1
Coastal Refining & Marketing, Inc.
P.O. Box 521
1300 Cantwell Lane
Corpus Christi, TX 78403
(512) 887-4252
Campos Roschetzky 8 1
Champlin Petroleum Co.
P.O. Box 9176
1801 Nueces Bay Blvd
Corpus Christi, TX 78408
(512) 887-3362
Carl Scarbro 16 11
USEPA
2565 Plymouth Rd
Ann Arbor, MI 48105
(313) 668-4209
-------
ATTACHMENT A (cont.)
EPA Vapor Pressure Correlation Testing Laboratories
Laboratory Number Manual P 176 Gas
Samples Tank Herzog SwRI Chrom.
Frank Gallagher 16
U.I.C.
P.O. Box 865
Joliet, IL 60434
Marvin Jackson 8
General Motor Corporation
Research Lab
3400 Mound Road
Warren, MI 48090
(313) 947-1778
Dale Burton 8
Automotive Testing Laoboratories, Inc.
Route 33, Bldg 40 AT TRC
P.O. Box 289
East Liberty, OH 43319
(513 666-4351
Doug Probert 8
Shell Canada Limited
Scotford Refinery
P.O. Bag 23
Fort Saskatchewan, Alberta T8L 3T2
Bill Finley 16
Texaco USA
P.O. Box 37
Convent, LA 70723
(504) 562-7681
J. Weyman Carson 8
Texaco Chemical Co.
P.O. Box 968
Port Arthur, TX 77641
(409) 982-5711
Dale Fuller 8
Valero Refining Co.
5900 Up River Rd.
P.O. Box 9370
Corpus Christi, TX 78469
(512) 289-3254
I846c
-------
Attachment B
ENGINEERING OPERATIONS DIVISION
Evaluation of Several Methods to Measure Volatility of Motor Fuels
Basic Letter of Inquiry
(date)
(Name of Laboratory)
Dear (Name)
As a part of its effort to reduce levels of ozone, the EPA is considering
ways to restrict the amount of hydrocarbons released into the atmosphere. One
approach would be to set limits on the volatility of gasoline. A possible
mechanism for enforcement would be through the measurement of vapor pressure.
The purpose of this letter is to confirm our telephone conversation
regarding the involvement of your organization in a program to evaluate the
accuracy, reproducibility, and precision of ASTM's proposed procedure for
measurement of vapor pressure. This procedure is contained in ASTM P 176,
"Proposed Specification for Automotive Spark-Ignition Engine Fuel". Details
can be found on page 1059 of the 1986 Annual Book of ASTM Standards. We are
also interested in evaluating any comparable methods such as those using the
Herzog unit from UIC, the automated instrument from Southwest Research
Institute, or other methods, e.g., gas chromatography. (Merge*)
Each of the twenty or more laboratories in the program will receive a set
of eight blind samples for each measurement technique. Included in the set
are several examples of gasoline, a pure component, and an alcohol blend.
They will be shipped in one-quart bottles using DOT-approved containers and
should arrive within two weeks. Instructions, a data sheet, and a sample
transfer connection will accompany the shipment.
I will collect and analyze the data. A copy of our report will be
forwarded to you upon completion of the program. In addition to the results
of your analyses, we are interested in your opinions concerning sampling
procedures aa they apply to dispensers found at typical gas stations and
"ideal" containers for vapor pressure samples.
The timing of this project is somewhat critical to the schedule for the
overall program. As a result, your results must be here by October 20 so that
we can analyze the data and make our recommendations on test methodology,
outpoints for enforcement actions and directions for further evaluations. If
you will have trouble meeting this deadline, please contact me.
The sentence merged here confirmed the type(s) of equipment to be used by
the particular laboratory.
-------
-2-
We appreciate your involvement in the program and welcome your suggestions
and comments. The success of this program will help develop better
communication on technical issues between EPA and the private sector. If you
have any questions, please feel free to call me.
Sincerely,
Marcus Haubenstricker, Ph.D.
Fuels and Chemistry Services
Testing Programs Branch
EPA Motor Vehicle Emission Laboratory
2565 Plymouth Road
Ann Arbor, Michigan 48105
156lc
-------
Attachment C
ENGINEERING OPERATIONS DIVISION
Evaluation of Several Methods to Measure Volatility of Motor Fuels
Seal for Certified Analysis and Sample Label
Seal
CAUTION: This container contains an EPA vapor pressure correlation sample
and is not to be opened except by a qualified analyst
experienced in gasoline vapor pressure measurement. If leakage
is evident please note the sample number and call Carl Scarbro
or Dr. Haubenstricker at (313) 668-4209 or (313) 668-4378 for a
replacement. The sample bottle must be chilled to 32-34°P
before opening. Please note all precautions in ASTM P 176
before opening the container and performing the analysis.
Sample Label
Sample No: 12345678
Method/Equipment:
( ) Water Bath and Gauges
( ) Herzog Instrument
( ) SWRI Instrument
( ) Gas Chromatograph
-------
A2.4. J ripfi Kltrr <* ttulh.
A2.4.4 Rrfrafliimrtcr.' temperature compcnulcxl.'
hand held. 0 lu 10 * Oiii (I.3330- 1.3479).
A 1.5 Rragrufi j
A2..VI I'tailv t) Kraft-nit— Hragriil-grndc clirmi-
ral* will he used iff ill lesls. Unless iilhrrwi.tr indicated.
il il intended thai all reagents shall conform lo ihc
specifications of the Committee on Analytical Kcngrnls
of (he American Chemical Society, where luch jnccifi-
nitons art available.'* ()lher grades may be used, pro-
vided il it firs! ascertained lhal Ihe reagent is ofsuffi-
cienl purity lo prrmil its use wilhoul lessening the
accuracy of Ihc determination.
A2.5.2 Purity uf Holer—Unless otherwise indi-
cated, reference lo water shall be understood to mean
distilled water of water of equivalent purity. See Spec-
ification I) 119). Type IV.
A2.5.3 .-firriMir (Danger—I-ilremely flammable.
See Anne* A I.))
A2.S.4 * llcplinr (Wirnliig—flammable. Harmful
if inhaled. See Anncn AIM conforming lo material
lived in 7CM Methods I) 3KI. I)hi I. D 2699. and
1)2700. or (Klroleum spin! MI/KO (U'.rnlng-ll.ini-
nuhle. Vajior harmful See Annci AI 6) conforming
lo II* specification or equivalent. Commercial grade
' quality il adequate foi these solvents.
A l.t l*rtp*rallo
. vi lhal II
lomeler
until opt
.» r,
room ten
Attachment D
ENGINEERING OPERATIONS DIVSION
EvaluaU-on of Several Methods to Measure
Volatility of Motor Fuels
P 176 Procedure for Measurement of Vapor Pressure
nrier
0 V
llrii.
table
dilioni on idrntiral trsl material would, in the long
run, and in Ihr normal and correct o|>er«lion of Ihcirsl
method, eircrd Ihe following values only in one cast
in twenty: I
and pipel filler and Moppet Ihe cylinder
A2.7. 3 Invert the cylinder, holding Ihe shipper wiih
a finger, and allow the waler 10 flow into Ihe sample.
Keeping Ihe cylinder in a lioii/onl.il altitude, shade it
vigorously for exactly 2 mm. two lo three sliokes |*r
second, using 12) lo 250 mm (5 lo 10 in.) strokes.
A2.7.4 Immediately place Ihe cylinder in a vertical
position on a tiliuiion-frec suifncc and allow the run
lenls lo settle undisturbed for 5 mm.
A2 7.5 Open the rover plate of the refriictnmeier
and using a soft clolh, or soil tissue paper moiMcneil
with waler. wipe Ihe prism and covei plale Oiy hnih
surfacn with a soil, dry clolh or tissue. Then close Un-
cover plale.
A2 7.6 Air dry I lie pi|x:i and install the pipel Win
Krmove Ihe *lop|ici from llie giadiialrd olimlci. lilt
Ihe cylinder. ,ind insrn Ihe pi|X-| Milh Hie Up in Ihr
water phave. Squee/e Ihe pi|>el fillef and disrhaigc -in
through Ihe pi|>rl lo puige any I'uel lhal may havr
enlcreil the pipel lip. Draw up mlo Ihc pipel ahoul one-
half of Ihc waler phase in Ihe bottom of the cylinder
1 ry to avoid drawing in any emulsion.
Noil A2.2 — If the water is totally arrvoihed into Ihe
furl phase, this indicates lhal the sample contains i
high coordination (greater than 25 volume %) uf
elhanol or heavier alcohol.
A2.7.7 Remove Ihe pipel from the giadualed cylin-
der. dry Ihr outside of Ihc pi|X'l wiih lissur lo remove
any fuel on Ihe suiface. and then disihaicc a few diop*
of waler phase to purge any fuel in Ihe lip. Till Ihe
lefracloineler with Ihe eyepiece up. Wulioul lilting the
rover plate, place the pipel Up at Ihe top of the co»er
plale between Ihe hinges ami diM haigc enough of Ihe
waler phase lo rover al leasi half of the prism.
A2.7.H lininc-di.ik-ly poinl Ihe instrument toward a
window or oilier source of h^lil s» the light enters Ihe
piism fmni above Look lliroiiKh Ihe eyepiece and take
Ihe reading al the poinl where the dividing line between
light and dark crosses the scale.
Null A2 1— lake Ihe reading within III s of dis-
chatging Ihc walei phase onto Ihe refrat lomeler pn*ni.
Ihc reading may change with lime as constituent*
evaporate or Ihe sample otlici-wise changes in compo-
sition.
A2.7.9 Record (he scale leading.
1 the AO In.luMii.il Muid leMei M.»lfl HH«I A
I rnificraiurc Ciiiiirvnuiril 1 1 j nd Krlijchinieirr ha> net" I
tuiuMr Al*ti an oiuivjlt-nl mjy he iivtl.
""KiMjrnl ChrinH ah. Amriii-jn I lirniiial Si«-«l> S|«f'»
i-Jlinnl.- Am. < 'hrniii »l S.IT . WaUunflon. IX 1 1" IUM""""|
cm Ihe lr*lm| u( reu|>.riiii nol hMeil hy Ihr Amrnun I **"'""'
S<»ieir. *ee -Krjfrni Chciuu-alt ami Sumlinh." hi <«""*
R.'S.n. I). Vjn N,,MunJ( o . liu . New V«nl. NV.iml -|l»n«"
Suie* l*)uiinai-«iiicia ~
•rcial
«e in
rfore
concluding lhal an alcohol is present, a minimum
leading of 0.5* llrii mini be observed. However, it is
still possible lhal a sample having a rrfrarlomclrr read-
ing of leu than 0.5* llrii may contain a low level of
alcohol.
A 2.9 Precision tnd DUs
A?*) I Ihe prrrision of this proposrtl trsl method
as deirrminrd hy stalisliral e»aminalion of mlerlalxv
ralory results is as follows:
A2.9 !(/) Hr/irainliilily—Tht diffeiencr hriwrrn
successive lest rrsiills. obtained hy Ihr same operator
with the same Apparatus under conslanl operating con-
Range
0-1.0 vol % ukoliol
Rr|erilion of
Ihe lest method, eiceed the following values only in
one case in Iwenly:
Range
0-1.0 vol % alcohol
Rrproduribihly
0.251 (Rrfmrlomrtcr
Reading)*
A).
Ir.SI MKTIIOI) KOK VAPOR
HIM. (DRV MKIIIOD)
A2.1.2 nia\ — There being no criteria for measuring
bias in these leM product combinations, no statement
ol bias ran he made
Of SPARK-IGNITION KNGINK
A 1.1 S*-«|>*
A.I. 1.1 I his pro|Nisrd lest melhod rovers Ihe deter-
mination nf ihe absolute vapor pressure (Nolr A J.I) of
gasolines and gasohne-otygenalr blends.
Noll A.I.I — Dccause theeilcrnal almospherir pies-
sure is counteracted hy Ihe atmospheric pressure ini-
tially present in Ihc air chamber, Ihe "vapor pressure"
is an absolute pressure al I GOT (37 8"C) in pounds-
force per square inch or (kilopascals (kPa - kN/rn').
1 his vapor pressure differs from Ihe true vapor pressure
of the sample due lo some small sample vapori/jlion
and Ihe presence of air in Ihe confined space.
A.VI.2 Ihe values staled in inch-pound units are
siandaid.
A J.I .Summary of Method
A.V2.I The fuel chamber of Ihe vapor pressure ap-
paratus is filled wiih Ihc chilled sample and connected
lo ihc air chamber al I DOT (37.8'C). The apparatus is
immersed in a hath al 100*1- and is shaken periodically
uniil a constant pressure is observed on Ihe gage at-
tached lo Ihe apparatus. 1'hr gage reading, suitably
coricclcd. is reported as the vapor pressure.
AJ..1 Significance and IJ*c
A.V3 I Test Method l> 323 cannot he used to deter-
mine the va|>or pressure ofgasolinr-onygenaie blends
which rontain wairr-eilrarlablc oiygenales because the
fuel sample comes into contact with waler. This pro-
nosed lesl method is a modification ofTcsl Method
I) J2J where contact with waler haj been eliminated.
A.1.4 Apparajm
AJ.4.1 Ihc tonsliuclion of the required apparatus
is drs. lined in Anne« AI of Test Melhod U J2J.
A.V5 HeigenM
A.I 5. 1 I'linly <>( Rrnxrnli—lJtt rea|;enl glide
chemicals in all icsis. Unless otherwise indicated, it is
inlrndrd lhal all reagents conform to the specifications
of Ihe Commillce on Analytical Reagents of Ihe Amer-
ican Chemical Society where such specifications are
available.' Olhrr grades may be used, provided il is first
ascertained that the reagent is of sufficiently high purity
lo permit its use without lessening Ihe accuracy of the
determination.
A3. 5. 2 Atelonr (Danger — tutremely flammable.
See Anno Al.l)
A3. V 3 Naphtha (Danger— r»liemely flammable.
See Anne. A I 2).
A3.6 Handling of Sumplei
A3. 6. 1 The enlrcme sensitivity of vapor pressure
measurements lo losses through evaporation and Ihc
resulting changes in composition is such as lo require
Ihe utmost precaution and Ihe most meticulous cart in
Ihe handling of samples.. The provisions of this section
apply lo all samples for vapor pressure deirrminjiions.
A3.6.2 .Sample in accordance wiih Practice D 4057.
e«cerH lhal waler displacement (1 1. 3.1. 8 of Practice
D4057) must nol be uved
A.lft.3 Sample i'lwiiami'r .V/;r--The si/r of Ihe
sample container from which Ihe vapor pressure sample
is taken is I ql (I L). Il will be 70 to 80 % filled with
ihe sample.
AJ.6.4 /Viviiur/iMK
A3 6.4(0 Deierminr vapor pressure as Ihr fust ivsi
run on a sample. Do nol withdraw more than one
sample from the sample container for this lest.
A. 1.6.4(7) Protect samples from eicessive beat prior
lo lesling.
A 1 6 KJ) Do nol mi samples in kjky conuinen.
Uivard them and obtain new samples.
-------
0-2 PropoMl P 176
A3.A.4«) Discard samples lhal have separated inio
(wo phases and obuin new samples (see Note A3.4).
A J.6.3 Samplr Handling Temperature—In all cases,
cool the sample container and contend lu J2 lo 34'l;
(0 lo I*C) before the container is opened. To ensure
sufficient lime luf reach this temperature, directly mea-
uire trie icmperiture of a similar liquid in a like con-
tainer placed in the cooling bath at Ihe same lime as
the sample.
AJ.7 rrrp«nlk» for Tnl
A3.7.1 Vetificalton of Sample Container Filling—
With the sample al a temperature of 32 lo 34'F (0 lo
I'C), lake Ihe container from the cooling bath, wipe
dry with an absorbent material, unseal il. and examine
its ullage. The sample content, ai determined by use of
a suitable gage, must be equal lo 70 lo 80 % of Hie
container capacity.
A).7.l(/> Discard Ihe sample' if its volume is less
than 70 % of Ihe container capacity.
AJ.7 !(.?) If the container is more than 80 % full,
pour out enough sample lo bring the container contents
within Ihe 70 lo 80 % range. Under no circumstance
may any sample poured out be relumed lo Ihe con-
tainer.
A.) 7.2 Ate Saturation nf Sample in Sample Con-
lamn
A) 7 1(1) With Ihe umple again at a temperature
of .12 lo 34T (U lu I'C) take the conlainei from the
cooling bath, wipe it dry wiih an absorbent material.
unseal il momentarily, taking care 10 prevent water
entry, meal il. and shake il vigorously. Return il lo the
bath for a minimum of 2 min.
A).7.2(J) Repeal AJ.7.2 (I) twice mort. Return the
sample to Ihe bath and keep il there until Ihe beginning
of the procedure (A3.8).
AJ.7.3 Preparation of Fuel Chamber—Observe the
apparatus preparation procedure of A.3.8.J. then store
the stoppered fuel chamber and Ihe sample transfer
connection in a refrigerator or ice-water bath for a
sufficient lime lo allow the chamber and Ihe connection
lo reach a temperature of 32 lo J4'F (0 lo I'C). If an
ice-water bath is used, keep Ihe chamber upright and
not immersed over Ihe lop of the coupling threads. The
transfer connection is inserted into a plastic bag lo keep
il completely dry during cooling.
A3.7.4 Preparation of Air Chamber—Observe Ihe
apparatus preparation procedure of A3.8.5. Connect
Ihe gage lo (he air chamber and close the lower opening
securely with a dry No. 6Vi rubber stopper. Make sure
the stopper is inserted far enough lo securely dose the
vent hole in Ihe air chamber connection. Immerse Ihe
air chamber lu al least I in. (2) mm) above its lop in
Ihe water bath maintained al 100 ± 0 2'F( J7.8 ± 0 I'C)
for not less than 20 min. Do not remove the air chamber
from Ihe Water bath until the fuel chamber has been
filled with the sample as described in A3.8.1.
AJJ ProrHwt
A3.8.1 Sample Transfer—With everything in read-
iness, remove Ihe chilled sample container from Ihe
bath, dry il with absorbent material, uncap il. and dry
and insert Ihe chilled transfer apparatus (see Fig. A3.1).
Quickly place the chilled fuel chamber, in an inverted
position, over Ihe sample delivery luhe of the transfer
apparatus. Invcil the mine system rapiilly MI lli.ii inr
fuel chamber is upright, with the end of the delivery
luhe lourhing Hie bottom of Ihe fuel chamber. Till inc
fuel chamber to overflowing. Withdraw I he deliveiy
lube from Ihe fuel chamber while allowing the sample
lo continue flowing up lo the moment of complete
withdrawal.
A3.8.I (I) Caution—Make provision for suitable
collection and disposal of Ihe overflowing fuel to avoid
Tire hazard.
A3.8.2 A\scnihly nf Apparatus—Immediately re-
move Ihc air chamber from the water bath and imme-
diately dry Ihc etlerior of Ihc chamber with absorbent
material giving particular carr to Ihe connection be-
tween the air chamber and the fuel chamber. Remove
Ihe stopper after drying and immediately couple the
two chambers. Not more than IU s shall be consumed
in coupling Ihe two chambers.
Not I- A.V2—When Ihe air chamber is removed fiiini
Ihe water l>alh. is dried, and the stopper is removrtl.
connect il lo Ihe fuel chamber without undue move-
ments through Ihc air. which could promote enchange
of room temperature air with Ihc IUOT (37.8V) air in
Ihe chamber.
A3.8.3 lntrtnliictit>n ttf Apparatus inltt llnth—I urn
Ihe assembled vapor pressure apparatus upside down
lo allow the sample in Ihe fuel chamber In run into ihe
air chamber. With Ihe apparatus still inveilrtl. shake il
vigorously eight limes in a direction parallel lo the
length of (lie a|>paralus. With (he gage end up. immerse
Ihe assembled apturalus in Ihe bath, maintained al 11*1
± 0.2T (37.8 t O.rC). in an inclined position so Ihil
Ihe connection of Ihe fuel and air chambers is below
Ihe waler level and may be carcfull, eiamincd for leaks.
If no leaks are observed, further immerse Ihe apparatus
lo at least I in. (2) mm) above the lop of Ihe air
chamber. Observe Ihe apparatus for leakage throughout
Ihe test. Discard Ihe test at any lime a leak is delected.
NOTI A3.3—Liquid leaks are more difficult lo detect
than vapor leaks, and because the coupling between ihe
chambers is normally in Ihe liquid section of Ihe ap-
paratus, give Ihe coupling particular attention.
NOTK A3.4—After the apparatus has been immened
in Ihe hath, check ihc remaining umple for phase
separation. If the sample is contained in a glass con-
tainer, this observation can be made prior lo sample
transfer (A3.8.1). If Ihe sample is contained in a non-
Iransparcm container, shake Ihe sample vigorously for
5 s arid then immediately pour a portion of Ihe remain-
ing sample into a clear glass container Immediately
after shaking this sample again for 3 s. observe the
sample for phase separation. If this, sample is not clear
and bright, and free of a second phase, discard Ihe lest
and Ihe sample.
A.V8.4 Arra.il/rf/itivi/ of Vapor Piruurr—After Ihf
assembled vapor pressure apparatus has been immersed
in Ihe hath for al least 3 min, lap Ihe pressure gage
lightly and observe Ihc reading. Withdraw the apparatus
from Ihe hath and repeal A3.8.3. Al intervals of not
Ins than 2 min. perform A3.8.3 until a total of not leu
than five shakings and gage readings have been made:
continue thereafter, if necessary, until Ihe last I*"1
consecutive gage readings are constant, indicating equi-
D-2 Propoaal P 176
lihrium attainment. These operations normally require
211 lo 3(1 min RCM\ the final gage pressure In Ihe nearest
0(15 |>si (0 2) kl'a) lor gages with intermediate gradu-
ations of O.I |ni (II.) kfa)ur less and lo Ihe nearest O.I
psi lor gages with graduations of 0 2 lo 0.3 psi (1.0 lo
2.3 kl'a). and record Ihe value as Ihe "uncorrecled vapor
pressure" of the sample. Without undue delay temove
the pressure gjge and. without attempting lo remove
any liquid which may be (rapped in the gage, check its
reading against lhal of Ihe manometer while both are
Subjected lo a common steady pressure which is no
nmrc than 0.2 psi (1.0 k Pa) different from Ihe recorded
"uncorrecled vapor pressure." If a difference is observed
between the gage and manometer readings. Ihc differ-
ence shall he added lo or subtracted from Ihe "uncor-
rct'ird vapor pressure" recorded for Ihe sample being
tested, and Ihe resulting value shall be recorded as Ihe
vapor pressure of Ihe sample.
Nun A3.3—Cooling the assembly prior lo discon-
necting Ihc gage will facilitate disassembly and reduce
ihe amount of hydrocarbon vapors released into Ihe
loom.
Nun A.I d—I'l-ri/icuiinn urn'i,ige
by directing a small jcl of air mm us Hourdnn tube for
al least 5 min. Kinse both chambers and the sample
transfer connection several limes with petroleum naph-
Iha. then several limes with acetone, then blow dry
using dried air. Stopper Ihe fuel chamber and place it
in Ihe refrigerator or ice-water bath for Ihe neit test.
NOIE A3.7—If the purging of Ihe air chamber is
done in a bath, be sure lo avoid small and.unnoliceable
films of floating sample by keeping the bottom and lop
openings of Ihe chamber closed as they pass through
the water surface.
A3.* Precautions
A 3.9 I Grots errors can be obtained in vapor pres-
sure measurements if Ihe prescribed procedure is not
followed carefully. The following list emphasi/cs Ihe
importance of strict adherence lo Ihe precautions given
in Ihe procedure.
A.3.9.!(/) Chfrkmg the Fressure (Sate—Check all
gages against a manometer after each lest in onler lo
ensure high precision of results (A) I 4) Read all gaR"
while ihe gages are in a vertical position and after
lapping them lightly.
A3.9.1(7) Shake Ihe c6nlainer vigorously lo ensure
equilibrium of Ihe sample with Ihe air in Ihe conlainei
(A3.72)
A39.l(.() Cherking for Isak.i—Check Ihe 1 prior a
lus before and during each test for both liquid ami
vapor kaks (Al.1.6 of Test Method D 323 and Note
3).
A3.9.1(1) Sampling—Because initial sampling ami-
Ihe handling of samples will greatly affect Ihe final
results, employ the utmost precaution and Ihe mosi
meticulous care lo avoid losses through evaporation
and even slight changes in composition (A3 6.3 ami
A3.8.1). In no case shall any part of Ihe apparatus nxll
be used as Ihe sample container previous to actually
conducting the lest.
A3.9.1(5) Purging the Apparatus—Thoroughly
purge the pressure gage, Ihe fuel chamber and Ihe air
chamber lo be sure they are free of residual sample
(Ihis is most conveniently done at the end of Ihe
previous lest. See A3.8.5.) Il is important to remove all
waler from Ihe apparatus before cooling Ihe gasoline
chambers and healing Ihe air chamber. In high-humid-
ity conditions be alert for and avoid condensation on
Ihe transfer connection and interior walls of Ihe ipp*-
ralus
A.3.9. l(rt) Coupling the Apparatus—Carefully ob-
serve Ihe requirements of A3.8.2.
A3.9.1(7) Shaking the Apparalut—Shake the appa-
ratus "vigorously" as directed in A3.8.3 in order 1.1
ensure equilibrium.
A.t.lO Rrpml
A3. III.I Ki'i»irii"f Keiulit—Report lo the nearest
0.115 psi (0 25 kPa) or O.I psi (0.5 kPa) the gate result
observed in A3.8.4. after correcting for any difference
between the gage and manomeler. as the "vapor pres-
sure" in pounds-force per square inch (or kilopascals) I
without reference lo temperature. '
1060
-------
O-2PropOMlP 178
D-2 Propotal P 176
AJ.II Prrcbioa a«4 BUi
A J.I 1.1 fr,rifiim—The precision of lhi$ proposed
ten method (us foi been determined, bul is under
Chi fled Sample
fansfer Connection
Sludy.
A3.11.2 Dim—There being no criteria 'for measur.
ing bias in these test-product combinations, no state
menl of bias can be made.
Chilled Gotolint Chombv
FIG AJ I
la)
Contanrr
fo Transfer
of Sam;*
Sl.,urk4 Skcir
(b)
™ Ck»un>
I* Smlr
(c)
Gtsolnr Chamber
»Vtd Ovw l«jjd
Orlivrry 1ub«
IVwilion of Syclrm for
5am(4f Transfer
olkW •( Tnuihnlm S*«>k> to Cuollw CkMbrf ham Oa«*-1 >•« rMlilont
A4. PROPOSED TEST METHOD FOR VAPOR PRESSURE OF SPARK-IGNITION ENGINE
FUEL USING THE AUTOMATIC APPARATUS
A4.I Scop*
A4.I.I This proposed lesl method cuvci> * deier-
minalion of vapor pressure of gasoline and gasoline-
oiygenalc Mends.
A4.I.2 The values suied in inch-pounds uniu are
the standard.
A4.2 S»«urr of Method
A4.2.1 The cold sample cup of the automatic vapor
pressure instrument is Tilled with the chilled, air-salu-
rated fuel sample and connected to the instrument. The
instrument operation is juried and the vapor pressure
is automatically determined.
A4J Affuntm
A4.3.1 Vapor Pressure Instrument''
A4.3.2 Mercury Manometer, as prescribed in Test
Method 0323. AI 6.
A4.3.3 Cooler, appropriate apparatus consisting of
a cooled air or water bath, and preferably a refrigerator
lo cool and maintain samples, sample cups, and sample
transfer connection* at 32 lo 34'F (0 lo I'C) Do not
use solid carbon dioiide (dry ice) for this purpose.
A4.3.4 Sample Transfer Connection, as illustrated
in Fig. I of Test Method O 323.
A4.4 RnfeMi aarf Materials
A4.4.I C'yrlopealane.V.Vl psi( 68.3k Pa) vapor pres-
sure. (Dugef—Eilremcly flammable and harmful if
inhaled. See Anne> A 1.2).
A4.4.2 Mrthyi C>*l,ipeniane. 4.50 psi (31.0 kPa)
vapor pressure (Dmngrr—Euremely flammable and
harmful if inhaled. See Annei A 1.2).
A4.4.3 n-Penlane, 15.57 psi (107..1 kPa) vapor pres-
sure. (Danger—fiilremely flammable and harmful if
inhaled See Anno AI 2).
A4.S Safely Prmullons
A4.S.I Recogni/e the potential for fire and r t* ohuinfll
from Southwell Rewarch Inililulc An et|ui»aknl may
he the TIIM test lun on a umple: do mil u\c more than
our \-iniplc Iroin the siimple container fin this lesl.
A4 h.4 (^) I'mlccl samples'from eneessive heal prior
to testing.
Ain for Trtl
A4 7 I I'm/iiuiiiin i,j Sami'li' (imlainft filling—
With the sample at a lem|ieraliire 32 lo 34'F (0 lo I'C)
lake the container from the cooler, wipe dry with an
jtnorhcnl null-rial, unseal it. and eiamine ils ullage.
I he umple content, as determined by use of a suitable
g:i£C. shall be equal lo 71) lo HO 71 nf the container
ranauly.
A4.7 !(/) Divjnl the umplc if its volume is less
than 70 "/„ of the container capacity.
A'l 7 !(.') If the container is more than 80% full,
(Kiiu out enough uniple In bring I he innuincr ronlrnls
within the 70 to Kl) % range. Under no circumn
inverted position over the sample delivery lube of the
transfer apparatus Invert the entire system rapidly sn
that Ihe sample cup is upright, with the end of Ihe
delivery lube touching the bottom of Ihe cup. (The end '
of Ihe delivery lube is cut al an angle lo prevent
interference with the delivery flow.) I'lll the cup in
overflowing and withdraw the delivery lube from the
cup.
Nun A4.2—Caution—Make provisions for suitable
restraint and dis|iosal of any spilled gasoline lo avoid <
fire ha/ard.
A4.9.4 .ItirfriWr <•/ Jp/*ir
-------
0-2 Proposal P 176
1
0-2 Propo««l P 176
NOTE A4..1—Aflcr initialing the lot. check the re-
maining sample Tor phase separation. If the sample is
contained in a glass coniaincr. this observation can he
made prior to sample transfer (A9.9.J). If the sample is
contained in a •nnlransparenl container, shake the
sample vigorous^ for 5 s and then immediately pour a
portion of the rrVnaining sample inlo a clear-glass con-
tainer Immediately after shaking this sample again fur
5 s. observe the sample for phase separation. If this
sample is nol clear and bright, and free or « second
phase, discard the lest and the sample.
A4.9.6 Al the end of the instrument operation, read
and record the displayed DPM reading.
A4.9.7 Perform the Leak Test described in Section
V. Part E. of Ihe instrument manual." Discard any
result obtained where leakage is indicated.
A.IO Calrublio* of Result*
A4.IO.I Use Ihe OPM readings obtained in
A4.S.3 (/) for Ihe pure compounds versus their known
vapor pressures lo construct a graph relating DPM and
true vapor pressure.
A4.I0.2 Using Ihe graph, determine fhc true vapor
pressure corresponding lo the DH>1 reading of each of
the unknown samples.
A4.ll Report
A4.I I.I Report the DPM readings obtained for rath
of Ihe pure- compounds.
A4.11.2 Rrpoit Ihe DPM readings obtained for each
of Ihe unknown samples.
A4.I 1.3 Report Ihe indicated I rue vapor prcvsure of
each unknown as determined in A4.11.2.
*
A4.I2 Precision and Bias
A4.I2.I Precision—The precision of this proposed
test method has nol yel been determined.
A4.12.2 Bias—There being no criteria for measur-
ing bias in these test-product combinations, no stale-
menl of bias can be made.
AS. PROPOSED TEST METHOD FOR WATKR TOLERANCE (PI I ASK SEPARATION) OK
SPARK-IGNITION ENGINE FUEL
A5.I Scop*
A).1.1 This proposed Icsl method determines the
ability of gasoline-oxygenate blends lo retain water in
solution or in a stable suspension at the lowest temper-
ature to which they are likely lo be exposed in use.
A3.1.2 The values staled in SI units are Ihe standard.
A5.J Summary of Method
A5.2.1 The sample of fuel is cooled slowly to its
expected use temperature and is observed for phase
separation. The apparatus of Test Method D 2500 or a
dry ice/isopropyl alcohol bath may be used. The pro-
cedure of Test Method D 2500. in which the sample is
cooled rapidly lo Ihe lest temperature by immersion in
a balh that is maintained at a considerably lower tem-
perature, must not he used due lo the large temperature
gradient employed, and because phase separation in
gasoline-oxygenate blends has a relatively long bul un-
predictable induction period.
A5J Significance and Use
A5.3.1 Some oiygenale-containing fuels, gasoline-
alcohol blends in particular, have a very limited ability
lo retain water in solution or in stable suspension, and
if Ihe amount of water in the blend exceeds this limit.
Ihe fuel will separate inlo a lower oxygenale-rich
aqueous phase and an upper oxygenale-lean hydrocar-
bon phase. The most important factor governing Ihe
ability of a specific fuel to retain water without such
separation is ils temperature. This test method is in-
tended to determine Ihe maximum temperature al
which Ihe fuel will separate. The I Oth percenlile 6-h
minimum temperatures or 10'C (50"F). whichever is
lower, for Ihe lime of year and geographic area of the
United Slates in which Ihe fuel may be used are tabu-
lated in Table 4 of this proposed specification. These
temperatures represent Ihe maximum temperatures
above which the fuel blend must nol separate inlo two
distinct phases
A5.3.2 Nole lh.il in this lest actual separation of the
sample inlo two distinct phases separated by a single
common boundary (which may consist of a layer of
emulsion) is Ihe criterion for failure. Formation of a
haze without such separation inlo distinct phases is not
cause for rejection.
AS.4 Apparatus
AS.4.1 Sample Conlainfr—This may be as specified
in 4.1, of Test Method 1)2500, bul any glass coniaincr
of about 100-mL capacity capable of accommodating
a thermometer may be used.
A5.4.2 TherniHineten mccling Ihe requirements of
Specification 1. I, wilh a range appropriate lo the lesl
require mcnl.
AS.4.3 ('ink. lo fit sample container borcil cenli:illy
for Ihc lest thermometer.
AS.4.4 Citiiling ttmli may he of similar dimensions
lo those specified in Tesl Mcfhod O2500. 5.7 and
provided wilh n jacket, disk, and gasket as specified by
5.4 through 5.6 of Test Method D2500, filled wilh an
equal-volume mixture of water and "permanent" anti-
freeze, and provided wilh refrigeration coils capable of
reducing ils temperature lo -4()"C (-40T). Alterna-
tively, any balh of adequate site may he used.
AS.S Procedure '
A5.5.1 Rinse oul Ihe sample coniaincr wilh some of
Ihe fuel lo be tested. Drain.
A5.5.2 Pour about 40 mL of Ihc fuel inlo Ihe sample
container. The precise amount is nol critical, hul il
must be enough lo submerge Ihe thermometer bulb
adequately, without being so much as lo require an
excessive amount of cooling lime.
A5.5.3 Seal the sample container. Locale a ther-
mometer of Ihe appropriate temperature range wilh the
bulb approximately al Ihe center of Ihe fuel sample.
A5.5.4 Cool ihc sample by intermittent immersion
jn or circulation of Ihc coolant. The fuel is either stirred
conlinously or vigorously shaken. Starting al a lemner-
jiure about I6"C (30°F) above Ihe lesl temperature,
cool Ihe sample al a maximum rate of 2*C (4*F) per
ininiilc lo ihc temperature given in Table 4 for Ihe
water tolerance required. If phase separation is observed
prior lo reaching Ihe lesl temperature. |>er AS.3.2, Ihe
"cooling" It.i/c ixuni temperature is recorded. Then Ihe
sample is allowed lo warm while being shaken fre-
quently or stirred and a "warming" haze point temper-
ature is recorded. The "cooling" and "warming" tem-
peratures arc averaged lo determine Ihe actual hare
point. Ihc "cooling" and "warming" haze point tem-
peratures may be repealed for improved accuracy.
A5.6 Report
A5.A. I Report "pass" if no separation occurs al Ihe
specified temperature for Ihe water tolerance class re-
quired, otherwise note Ihe temperature for phase scpa-
ration (see A5.3.2) and report "fail."
AS.7 Precision and Ilia*
AS.7.1 I'rffisinn—The precision for this proposed
lesl melhod has nol been determined.
AS.7.2 Bias—There being no criteria for measuring
bias in these lesl-produci combinations, no statement
of bias can be made.
APPENDIXES
(Nonmandalory Information)
XI. PROPOSED TEST METHOD KOR VAPOR-LIQUID RATIO OK SPARK-IGNITION ENGINE
KIEL (MODIFIED TEST METHOD D 1533 PROCEDURE)
XI.I Scope
XI I.I This proposed lesl melhod covers a proce-
dure lor mcDSuimg Ihc volume of vapor formed al
atmospheric pressure from a given volume* of gasoline
or gasoline-oxygenate blend. The ratio ol these volumes
is expressed as the vj|xir liquid (V/L) ratio of Ihe fuel
al Ihe temperature of Ihe lest.
XI.1.2 Ihe values staled in inch-pounds units are
Ihc standard.
XI.2 Summary of Method
XI.2.1 A measured volume of liquid fuel al 32 lo
4UT (0 lo 4'C) is introduced through a rubber septum
inlo » mercury-filled burcl. The charged huret is placed
in a Icmperaiure-conlrolled waler balh. The volume of
vapor in equilibrium wilh liquid fuel is measured al Ihe
desired temperature or temperatures and Ihe specified
pressure, usually 760 mm Hg. The vapor-liquid ratio
(>'/;.) is then calculated.
XI.2.2 If it is desired to know Ihe temperature cor-
responding lo a given I'//.. Ihe vapor-liquid ratio is
determined al several temperatures and Ihe selected
pressure. The results are plotted and Ihe temperature
read al Ihc given Vfl..
X1.3 Significance end Use
XI.3.1 The tendency of a fuel lo vaporize in auto-
mobile fuel systems is indicated by Ihe VfL ratio of
that fuel at conditions approximating those in critical
pans of Ihe fuel systems.
X 1.3.2 Tesl Melhod D 2533 is nol applicable lo
fuels containing alcohol, ethers, or other compounds
soluble in glycerine. This proposed test melhod substi-
tutes mercury for glycerine as Ihe confining fluid.
X 1.3.3 A relationship between vapor-liquid ralio of
gasoline-oxygenate blends and vehicle performance has
nol been dcfer,mined.
XI.4 Definition
XI.4.1 vapor-liquid ralio of a fuel, al any specific
temperature and pressure, is Ihe ratio, al thai temper-
ature and pressure, of Ihe volume of vapor in equilib-
rium wilh liquid lo Ihe volume of sample charged, as a
liquid, al 32'F(0'C).
Noir XI.I — fhis ralio differs from the ahsolulr
vapor-liquid ralio because corrections are nol m.ide for:
(/) liquid sample expansion wilh increasing tempera-
ture. (!) decrease in liquid sample volume by vapori-
zation, (0 dissolved air in Ihe liquid sample, and (/I
deviation from Ihe perfect gas law.
XI.5 Appinius
X1.5.1 y/L Uiiri-i," constructed of borosilicalc glau
in accordance wilh dimensions shown in Fig. X 1. 1.1 he
short bottom arm is closed wilh a rubber serum Nil lie
stopper, a U.S. Army Medical Corps type.
X1.5.2 rrrsMirr Cimlriil Equipment:
XI.5.2(7) A 250-mL leveling bulb containing mer-
cury, attached In Ihe \'/L hurel by rubber tubing at
shown in Fig. XI.2. The lop of Ihc leveling bulb must
be filled with a drying lube containing mercury vapor
absnibcnl" packed between balls of glass wool. This
drying lube is used lo minimize Ihe escape of mercury
vapor.
XI.5.2(2) Means for measuring Ihe difference in
liquid level between the I'/l. burel and Ihe leveling
bulb. A calhelomeler or similar optical leveling device
is suitable. A millimetre scale can be used to provide a
rough estimate.
X I.5.2(J) Baromi-lrr. accurate lo 0.5 mm llg.
XI.5.2(4) A mercury manomeler with I mm divi-
sions, required only for measurements al pressure
appreciably above or below Ihe prevailing almosphciii
pressure. The manomeler shall be connected wilh rub
" The V/l. bum No. 247I8-M2 and rilihrinon tiuppci Nn
247HI-CS. manufanuml by Continental Olan Blowing. Co .
26J6 S. Hill Si.. Un Anitln. CA 90007. have been foun.1
laliilanory for lhi> purpoic.
" Rnhorto available from J. T. P-ikrr Co.. Phillipst-ufj. NI
008AS has hern found saliiTaciorjr (CM (hit purrhrw.
1064
1065
-------
Attachment E
ENGINEERING OPERATIONS DIVISION
Evaluation of Several Methods to Measure Volatility of Motor Fuels
Vapor Pressure Correlation Data Sheet
Laboratory Number: ' Laboratory
Date Received: Address
Date of Analyses:
Instructions:
Each of the enclosed sample containers have been filled to 70 to 80$
capacity as required by ASTM P 176. After you have cooled the samples to 32
to 34°F, continue sample preparations according to sections A 3.7.2 or A 4.7.2
(depending on which equipment is used).
The table below is for the results from each of eight samples. The test
operator should fill out the sample's vapor pressure (in psig) in the labeled
rows for each method/equipment used.
If you have any comments or suggestions for assessment of gasoline vapor
pressure, please list them on the back of this form.
Results of Vapor Pressure Analyses (psig)
Sample Number
Method/ Equipment 1 2 3 ' '4 5 6 ' '7 8
P 176/gauges & bath:
Herzog Instrument:
SwRI Instrument:
Gas Chromatograph
Note: The results above are to be expressed in psig.
Analyst: Telephone; Date:
When finished with testing please telephone the above data to Dr. Marcus
Haubenstricker at (313) 668-4378 or (313) 668-4209 and mail this form to:
United States Environmental Protection Agency
Engineering Operations Division
2565 Plymouth Road
Ann Arbor, Michigan 48105
Attention: Dr. Marcus Haubenstricker
-------
Attachment F
ENGINEERING OPERATIONS DIVISION
Evaluation of Several Methods to Measure Volatility of Motor Fuels
Reported Manual Tank and Gauges Results
Replicate :
Lab. -
No.
1
2
3
4
5
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
42
43
44
45
46
47
48
49
Results
1
9.06
8.4
8.73
9.3
9.00
9.00
8.6
8.60
8.90
10.00
9.13
8.40
8.60
8.1
8.5
8.0
8.8
8.99
10.0
9.0
10.1
8.5
8.8
8.3
9.74
8.72
8.73
2
11.82
11.1
11.40
11.9
11.75
11.50
11.0
11.00
11.60
12.00
11.55
11.20
11.35
11.0
11.7
10.8
11.7
11.79
12.7
11.8
11.2
11.1
11.7
11.5
11.71
11.51
11.51
-Sample Identification Numl
3
12.40
11.3
11.40
12.4
11.90
11.60
10.7
10.70
11.65
11.50
11.50
11.10
11.65
11.3
9.70
11.65
10.8
11.6
12.43
12.8
10.6
10.9
11.2
11.5
9.7
12.03
11.35
11.25
4
10.37
10.4
9.85
10.6
10.55
10.20
9.1
9.10
9.95
9.50
9.98
9.80
10.00
9.4
10.1
9.4
10.1
9.38
11.3
10.3
10.4
9.4
10.1
10
10.1
9.73
9.93
5
8.63
8.3
8.85
8.8
9.05
8.80
9.2
9.20
8.90
8.00
8.90
8.40
8.25
8.2
8.8
8.0
8.9
8.3
10.1
8.9
8.2
8.7
8.8
8.7
8.99
8.43
8.7
6
11.89
•
12.3
12.40
12.7
12.70
12.30
12.3
12.30
12.30
12.00
12.38
12.40
11.80
11.4
9.75
12.3
7.9
12.3
12.18
13.0
12.0
12.5
12.3
12.4
12.5
12.96
11.57
12.77
7
10.08
10
10.00
10.7
10.30
9.80
9.6
9.60
9.85
9.10
9.93
9.30
9.80
9.5
9.2
9.4
9.9
10.02
11.1
10.0
10.2
9.6
10.1
9.1
10.72
10.01
10.06
8
11.17
11.9
12.40
13.0
12.25
12.20
11.6
11.60
12.25
12.00
12.28
12.60
12.25
11.1
9.20
12.4
11.6
11.8
12.97
13.5
12.2
12.3
12.1
12.3
12.3
12.78
12.41
12.62
-------
Attachment G
ENGINEERING OPERATIONS DIVISION
Evaluation of Several Methods to Measure Volatility of Motor Fuels
Reported Herzog Results
Replicate Results
No.
1
2
3
4
5
6
7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
[25]
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
42
43
44
45
46
47
48
49
1
8.64
8.92
8.7
6.95
8.49
8.60
9.40
8.55
9.41
8.40
8.80
9.60
8.46
8.4
4.05
8.20
8.98
8.80
9.4
8.99
8.44
2
11.54
11.64
11.15
9.85
10.66
11.25
11.15
12.08
10.80
11.45
11.60
11.11
11.5
10.00
11.20
11.58
11.50
12.0
11.75
10.97
-Sample Identification Numl
3
11.49
11.64
11.8
9.85
11.16
11.15
12.10
11.25
11.9
11.00
11.60
11.90
10.98
11.4
7.35
11.20
11.55
12.00
11.76
9.68
4
9.99
10.02
9.7
8.00
9.4
9.15
10.70
9.80
10.45
9.40
9.90
10.10
9.46
9.4
6.60
9.20
10.02
10.00
8.5
10.15
9.68
5
8.73
8.92
8.6
8.40
8.49
8.40
10.10
8.45
9.3
8.40
8.95
8.60
8.51
8.8
5.45
8.20
8.80
8.90
9.0
8.29
6
12.44
12.66
12.1
11.70
11.53
12.15
13.00
12.05
12.87
12.50
12.20
12.80
11.98
9.20
11.40
12.07
12.70
12.86
11.30
7
10.02
10.00
9.55
8.10
9.51
9.40
10.70
9.70
10.62
9.40
10.00
9.90
9.46
9.9
5.55
9.70
9.87
10.10
9.3
10.12
9.71
8
12.04
12.62
12.05
11.60
12.22
12.10
12.90
12.00
12.88
11.80
12.20
12.50
11.99
7.15
11.80
12.32
12.80
12.61
12.26
-------
Attachment H
ENGINEERING OPERATIONS DIVISION
Evaluation of Several Methods to Measure Volatility of Motor Fuels
Reported SwRI Results
Replicate Results
T oK _ _
LidD . — —
No. 1
1
2
3
4
5
6
7 8.9
9
10 8.76
11
12
13
14
15
16 8.46
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
42
43
44 8.3
45
46
47
48
49
' 2
11.7
11.54
11.06
11.3
Sample Identification Number-
11.7 10.0 8.9 13.2 10.1 13.2
11.55 9.70 8.98 12.59 9.74 12.58
11.10 9.39 8.44 12.11 9.52 12.13
11.3 9.4 8.4 12.3 9.8 12.4
-------
Attachment I
ENGINEERING OPERATIONS DIVISION
Evaluation of Several Method* to Measure the Volatility of Motor Fuel
Comparison of Repeatability and Reproducibility
Measurement
Technique
Treatment of Data
Gauges £ Bath "All" data (Note 1)
in accordance with
ASTM P 176 Outliers removed (Note 2)
After add'l processing
This Study-
Afo. of Repeat- Reproducibility
Labs ability Actual Normalized
28 0.75 psi 2.48 psi 2.52 psi
26 0.68 1.33 1.02
ASTM's 1986 Study ASTM D 323
Repeat- Repro-
ability ducibility
0.82 psi 1.96 psi
0.77 1.42
0.71 1.3
Repeat- Repro-
ability ducibility
0.25 psi 0.55 psi
Herzog (dry)
"All" data (Note 3) 21 0.76 1.59 1.31
Outliers removed (Note 4) 21 0.48 1.57 1.01
0.55
0.43
1.44
1.16
SwRI Instrument All data
(there were no outliers)
0.28
1.32
0.42
0.31
1.32
The data from this study and ASTM's 1986 effort were processed in accordance with Section RR D-2-1007
of "Manual of Determining Precision Data for ASTM Methods on Petroleum Products and Lubricants".
Note 1:
After substitution for
single reported results
(lab 32, sample 2) and
substitution for missing
pairs (lab 32, pairs 145
and 447)
Note 2:
Lab
No.
20
32
38
35
41
46
Sample (s)
excluded
1 45
all
all
6
1 &5
243
Note 3:
After substitution for
single reported results
(lab 13, sample 2; lab 32,
sample 2; lab 47, sample 5)
and substitution for missing
pairs (lab 24, pair 648;
lab 32, pair 6s8)
Note 4:
Lab
No.
7
20
32
48
Sample (s)
excluded
1 US
1 &5
447
243
updated: 9/2/87 JTW
-------
Attachment J
ENGINEERING OPERATIONS DIVISION
Evaluation of Several Method* to Measure Volatility of Motor Fuel*
List of Statistics
Method
Level of Data Reduction
Number of Sample Pairs
Number of Labs
Degrees of Freedom Replicates
Degrees of Freedom by Laboratories
Degrees of Freedom of Laboratories
Missing Single Results
Missing or Substituted Replicate Pairs
Mean Squares of Laboratories
Mean Squares by Laboratories
Mean Squares Replicates
Correction Term
Sums of Squares Samples
Sums of Squares Laboratories (missing data)
Sums of Squares Laboratories (no missing data)
Sums of Squares Sample Pairs
Sums of Squares by Laboratories
Sums of Squares Sample Replicates
Random Error
Estimate of Variance between Laboratories
Estimate of Variance by Laboratories
Variance of Repeatability
Trial ratio f
Mean Square Coefficients
to
Kl
K2
Estimate of Variance in Laboratories for Reproducibi 1 ity
Estimate of Variance between Laboratories for Reproducibility
Estimate of Variance in Laboratories for Reproducibility
Variance for Reproducibility
Degrees of Freedom for Reproducibility
Precision
Repeatability
Reproducibility
P 176
Before Cut
4
28
107
78
27
3
2
0.0712
1.0829
2.6466
24562.4111
414.5952
71.4578
NA
570.5218
84.4688
7.6232
0.0356
0.1955
0.5058
0.0712
0.0658
0.4971
0.3738
0.1285
0.0354
0.4048
0.3402
0.7804
95.1757
0.75
2.48
P 176
After Cut
4
26
100
75
25
0
4
0.0579
0.0453
1.4311
23328.9228
380.3452
35.7780
NA
419.5190
3.3957
5.7922
0.0290
0.1732
0.0063
0.0579
1.2793
0.4979
0.3839
0.1184
0.0288
0.0174
0.1695
0.2157
40.0680
0.68
1.33
P 176
Normalized
4
26
99
74
25
1
4
0.0585
0.0279
0.7599
23239.5704
380.3897
18.9977
m
401.4499
2.0624
5.7922
0.0293
0.0915
0.0153
0.0585
2.0992
0.0498
0.3839
0.1184
0.0291
0.0107
0.0900
0.1298
50.4569
0.68
1.02
Herzog
Before Cut
4
20
74
53
19
4
2
0.0726
0.0910
1.8822
17597.8992
292.0917
35.7613
1R
332.6744
4.8215
5.3689
0.0363
0.2239
0.0092
0.0726
0.7975
0.4914
0.3853
0.1240
0.0357
0.0351
0.2333
0.3041
31.8117
0.76
1.59
Herzog
After Cut
4
20
76
55
19
2
2
0.0290
0.1994
1.6358
17614.2634
302.5768
31.0800
NA
344.6254
10.9686
2.2048
0.0145
0.1795
0.0852
0.0290
0.1455
0.4913
0.3824
0.1272
0.0143
0.0763
0.2080
0.2985
37.3541
0.48
1.57
Herzog
Normalized
4
20
76
55
19
2
2
0.0290
0.0740
0.6636
18211.0198
316.7506
12.6088
NA
333.4314
4.0720
2.2048
0.0145
0.0737
0.0225
0.0290
0.3918
0.4913
3824.0000
0.1272
0.0143
0.0283
0.0844
0.1270
41.1084
0.48
1.01
SwRI
Before Cut
4
4
16
9
3
0
0
0.0081
0.0402
0.7493
3581.7532
73.0824
NA
2.2478
75.6915
0.3614
0.1298
0.0041
0.0886
0.0160
0.0081
0.2019
0.5000
0.3750
0.1250
0.0041
0.0151
0.0937
0.1128
4.3107
0.28
1.32
SwRI
Normalized
4
4
16
9
3
0
0
0.0081
0.0283
0.0345
3833.3714
76.1060
NA
0.1035
76.4646
0.2551
0.1298
0.0041
0.0008
0.0101
0.0081
0.2861
0.5000
0.3750
0.1250
0.0041
0.0106
0.0043
0.0190
18.2432
0.28
0.42
NA, not applicable
-------
Attachment K: Total Vapor Pressure vs. Days Since Sampled
M
PJ
M
O
ft
(0
4J
O
E-i
92 1
90 -
88 -I
86
84 -
82 -
80 -
78
y = 84.420 R = 0.02
10
20
30
40
Days Since Sampled (Note: one outlier was not included)
------- |