Designation: D2892 – 11a

Standard Test Method for

Distillation of Crude Petroleum (15-Theoretical Plate
Column)1
This standard is issued under the fixed designation D2892; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.

1. Scope* 1.4.5 Annex A5—Test Method for Determination of the

1.1 This test method covers the procedure for the distillation Temperature Response Time,
of stabilized crude petroleum (see Note 1) to a final cut 1.4.6 Annex A6—Practice for the Calibration of Sensors,
temperature of 400°C Atmospheric Equivalent Temperature 1.4.7 Annex A7—Test Method for the Verification of Reflux
(AET). This test method employs a fractionating column Dividing Valves,
having an efficiency of 14 to 18 theoretical plates operated at a 1.4.8 Annex A8—Practice for Conversion of Observed
reflux ratio of 5:1. Performance criteria for the necessary Vapor Temperature to Atmospheric Equivalent Temperature
equipment is specified. Some typical examples of acceptable (AET),
apparatus are presented in schematic form. This test method 1.4.9 Appendix X1—Test Method for Dehydration of a
offers a compromise between efficiency and time in order to Sample of Wet Crude Oil, and
facilitate the comparison of distillation data between laborato- 1.4.10 Appendix X2—Practice for Performance Check.
ries. 1.5 The values stated in SI units are to be regarded as the
standard. The values given in parentheses are for information
NOTE 1—Defined as having a Reid vapor pressure less than 82.7 kPa only.
(12 psi).
1.6 This standard does not purport to address all of the
1.2 This test method details procedures for the production of safety concerns, if any, associated with its use. It is the
a liquefied gas, distillate fractions, and residuum of standard- responsibility of the user of this standard to establish appro-
ized quality on which analytical data can be obtained, and the priate safety and health practices and determine the applica-
determination of yields of the above fractions by both mass and bility of regulatory limitations prior to use. For specific
volume. From the preceding information, a graph of tempera- warning statements, see Section 10.
ture versus mass % distilled can be produced. This distillation
curve corresponds to a laboratory technique, which is defined 2. Referenced Documents
at 15/5 (15 theoretical plate column, 5:1 reflux ratio) or TBP 2.1 ASTM Standards:2
(true boiling point). D941 Test Method for Density and Relative Density (Spe-
1.3 This test method can also be applied to any petroleum cific Gravity) of Liquids by Lipkin Bicapillary Pycnom-
mixture except liquefied petroleum gases, very light naphthas, eter3
and fractions having initial boiling points above 400°C. D1217 Test Method for Density and Relative Density (Spe-
1.4 This test method contains the following annexes and cific Gravity) of Liquids by Bingham Pycnometer
appendixes: D1298 Test Method for Density, Relative Density (Specific
1.4.1 Annex A1—Test Method for the Determination of the Gravity), or API Gravity of Crude Petroleum and Liquid
Efficiency of a Distillation Column, Petroleum Products by Hydrometer Method
1.4.2 Annex A2—Test Method for the Determination of the D2887 Test Method for Boiling Range Distribution of
Dynamic Holdup of a Distillation Column, Petroleum Fractions by Gas Chromatography
1.4.3 Annex A3—Test Method for the Determination of the D3710 Test Method for Boiling Range Distribution of
Heat Loss in a Distillation Column (Static Conditions), Gasoline and Gasoline Fractions by Gas Chromatography
1.4.4 Annex A4—Test Method for the Verification of Tem- D4006 Test Method for Water in Crude Oil by Distillation
perature Sensor Location, D4052 Test Method for Density, Relative Density, and API

1 2
This test method is under the jurisdiction of ASTM Committee D02 on For referenced ASTM standards, visit the ASTM website, www.astm.org, or
Petroleum Products and Lubricants and is the direct responsibility of Subcommittee contact ASTM Customer Service at service@astm.org. For Annual Book of ASTM
D02.08 on Volatility. Standards volume information, refer to the standard’s Document Summary page on
Current edition approved Dec. 1, 2011. Published January 2012. Originally the ASTM website.
3
approved in 1970. Last previous edition approved in 2011 as D2892–11. DOI: Withdrawn. The last approved version of this historical standard is referenced
10.1520/D2892-11A. on www.astm.org.

*A Summary of Changes section appears at the end of this standard.

Copyright (c) ASTM International. 100 Barr Harbor Dr., P.O. box C700, West Conshohocken Pennsylvania United States

Copyright by ASTM Int'l (all rights reserved); 1

 D2892 – 11a
Gravity of Liquids by Digital Density Meter 3.1.9 pressure drop, n—the difference between the pressure
D4057 Practice for Manual Sampling of Petroleum and measured in the condenser and the pressure measured in the
Petroleum Products distillation flask.
D4177 Practice for Automatic Sampling of Petroleum and 3.1.9.1 Discussion—It is expressed in kilopascals (mm Hg)
Petroleum Products per metre of packed height for packed columns, or kilopascals
D5134 Test Method for Detailed Analysis of Petroleum (mm Hg) overall for real plate columns. It is higher for
Naphthas through n-Nonane by Capillary Gas Chromatog- aromatics than for paraffins, and for higher molecular weights
raphy than for lighter molecules, at a given boilup rate.
D6300 Practice for Determination of Precision and Bias 3.1.10 reflux ratio, R, n—in distillation, the ratio of the
Data for Use in Test Methods for Petroleum Products and condensate at the head of the column that is returned to the
Lubricants column (reflux) to that withdrawn as product.
D6729 Test Method for Determination of Individual Com- 3.1.11 static hold-up or wettage, n—the quantity of liquid
ponents in Spark Ignition Engine Fuels by 100 Metre retained in the column after draining at the end of a distillation.
Capillary High Resolution Gas Chromatography 3.1.11.1 Discussion—It is characteristic of the packing or
D6730 Test Method for Determination of Individual Com- the design of the plates, and depends on the composition of the
ponents in Spark Ignition Engine Fuels by 100−Metre material in the column at the final cut point and on the final
Capillary (with Precolumn) High-Resolution Gas Chroma- temperature.
tography 3.1.12 takeoff rate, n—in distillation, the volume of product
D6733 Test Method for Determination of Individual Com- withdrawn from the reflux divider over a specified period.
ponents in Spark Ignition Engine Fuels by 50-Metre 3.1.13 theoretical plate, n—the section of a column required
Capillary High Resolution Gas Chromatography to achieve thermodynamic equilibrium between a liquid and its
vapor.
3. Terminology 3.1.13.1 Discussion—The height equivalent to one theoreti-
3.1 Definitions: cal plate (HETP) for packed columns is expressed in millime-
3.1.1 adiabaticity, n—the condition in which there is no tres. In the case of real plate columns, the efficiency is
significant gain or loss of heat throughout the length of the expressed as the percentage of one theoretical plate that is
column. achieved on one real plate.
3.1.1.1 Discussion—When distilling a mixture of com-
pounds as is the case of crude petroleum, there will be a normal 4. Summary of Test Method
increase in reflux ratio down the column. In the case where 4.1 A weighed sample of 1 to 30 L of stabilized crude
heat losses occur in the column, the internal reflux is abnor- petroleum is distilled to a maximum temperature of 400°C
mally greater than the reflux in the head. The opposite is true AET in a fractionating column having an efficiency at total
when the column gains heat, as with an overheated mantle. reflux of at least 14, but not greater than 18, theoretical plates.
3.1.2 boilup rate, n—in distillation, the quantity of vapor 4.2 A reflux ratio of 5:1 is maintained at all operating
entering the column per unit of time. pressures, except that at the lowest operating pressures be-
3.1.3 debutanization of crude petroleum, n—the removal of tween 0.674 and 0.27 kPa (5 and 2 mm Hg), a reflux ratio of
the light hydrocarbons up to and including n-butane, and 2:1 is optional. In cooperative testing or in cases of dispute, the
retention of the heavier hydrocarbons. stages of low pressure, the reflux ratios, and the temperatures
3.1.3.1 Discussion—In practice, a crude petroleum is re- of cut points must be mutually agreed upon by the interested
garded as debutanized if the light hydrocarbon cut collected in parties prior to beginning the distillation.
the cold trap contains more than 95 % of the C2 to C4 4.3 Observations of temperature, pressure, and other vari-
hydrocarbons and less than 5 % of the C5 hydrocarbons ables are recorded at intervals and at the end of each cut or
initially present in the sample. fraction.
3.1.4 distillation pressure, n—the pressure measured as 4.4 The mass and density of each cut or fraction are
close as possible to the point where the vapor temperature is obtained. Distillation yields by mass are calculated from the
taken, normally at the top of the condenser. mass of all fractions, including liquefied gas cut and the
3.1.5 distillation temperature, n—the temperature of the residue. Distillation yields by volume of all fractions and the
saturated vapor measured in the head just above the fraction- residue at 15°C are calculated from mass and density.
ating column. 4.5 From these data the TBP curves in mass or volume %,
3.1.5.1 Discussion—It is also known as the head tempera- or both, versus AET are drawn.
ture or the vapor temperature.
3.1.6 dynamic hold-up, n—in column distillation, the quan- 5. Significance and Use
tity of liquid held up in the column under normal operating 5.1 This test method is one of a number of tests conducted
conditions. on a crude oil to determine its value. It provides an estimate of
3.1.7 flood point, n—in distillation, the point at which the the yields of fractions of various boiling ranges and is therefore
velocity of the upflowing vapors obstructs the down-coming valuable in technical discussions of a commercial nature.
reflux and the column suddenly fills with liquid. 5.2 This test method corresponds to the standard laboratory
3.1.8 internal reflux, n—in distillation, the liquid normally distillation efficiency referred to as 15/5. The fractions pro-
running down inside the column. duced can be analyzed as produced or combined to produce

Copyright by ASTM Int'l (all rights reserved); 2

 D2892 – 11a
samples for analytical studies, engineering, and product quality matic devices can be employed provided they meet the same
evaluations. The preparation and evaluation of such blends is requirements. A typical apparatus is illustrated in Fig. 1.
not part of this test method. 6.1.1 Distillation Flask—The distillation flask shall be of a
5.3 This test method can be used as an analytical tool for size that is at least 50 % larger than the volume of the charge.
examination of other petroleum mixtures with the exception of The size of the charge, between 1.0 and 30 L, is determined by
LPG, very light naphthas, and mixtures with initial boiling the holdup characteristics of the fractionating column, as
points above 400°C. shown in Table 1 and described in Annex A2. The distillation
flask shall have at least one sidearm.
6. Apparatus 6.1.1.1 The sidearm is used as a thermowell. It shall
6.1 Distillation at Atmospheric Pressure—All components terminate about 5 mm from the bottom of the flask to ensure its
must conform to the requirements specified as follows. Auto- immersion at the end of the distillation. When a second sidearm

FIG. 1 Apparatus

Copyright by ASTM Int'l (all rights reserved); 3

 D2892 – 11a
TABLE 1 Data for n-Heptane-Methylcyclohexane Test Mixture at 75 % of Maximum Boilup and 101.3 kPa (760 mm Hg)
PropakA,B,C,D,E HelipakF,G,H Perforated PlatesE,I,J Wire MeshE,K
Column diameter, mm 25 50 70 25 50 25 50 25 50
Packing size, mm 4 6 6 No. 2917 No. 2918 NAL NAL NAL NAL
Boilup, mL/h 3 cm2 650 670 675 300 350 640 660 810 1050
Dynamic holdup
% of packed volume 17 15.3 17.0 15 14.3 NAL NAL 8.0 10.0
mL/theoretical plate 3.2 16 39 1.6 8.7 2.8 12.3 2.0 12.9
Pressure drop
kPa/m 1.2 1.05 0.94 1.53 1.41 NAL NAL 0.97 0.75
mm Hg/m 9.0 7.9 7.1 11.5 10.6 NAL NAL 7.3 5.6
kPa/theoretical plate 0.045 0.056 0.06 0.03 0.045 0.15 0.16 0.05 0.05
mm Hg/theoretical plate 0.34 0.42 0.43 0.24 0.34 1.1 1.2 0.35 0.37
HETP, mm (% of real plates) 38 53 61 21 32 (60 %) (65 %) 48 66
For 15-plate Towers
Packed height, cm (plates) 57 80 91 31.5 48 (25) (23) 72 99
Packed volume, mL 280 1570 3460 155 917 NAL NAL 353 1940
Dynamic holdup, mL 47 240 590 23 131 42 184 28 194
Pressure drop
kPa 0.68 0.84 0.86 0.48 0.68 2.2 2.4 0.70 0.73
mm Hg 5.1 6.3 6.5 3.6 5.1 16.5 18.0 5.3 5.5
Charge volume, L
Min (4 % Holdup) 1.2 6.0 15 0.575 3.3 1.0 4.6 0.7 4.9
Max (1 % Holdup) 4.8 24.0 60 2.3 13.0 4.2 10.4 2.8 19.4
A
Cooke, G. M. and Jameson, B. G. Analytical Chemistry, Vol 27, 1955, p. 1798.
B
Struck, R. T. and Kinner, C. R. Industrial and Engineering Chemistry, Vol 42, 1950, p. 77.
C
Cannon, M. R. Industrial and Engineering Chemistry, Vol 41, No. 9, 1949, p. 1953.
D
Bulletin 23, Scientific Development Co. P.O. Box 795, State College, PA 16801.
E
Cooke, G. M. Analytical Chemistry, Vol 39, 1967, p. 286.
F
Bulletin of Podbielniak Div. of Reliance Glass Works, P.O. Box 825, Bensenville, IL 60106.
G
Feldman, J., et al, Industrial and Engineering Chemistry, Vol 45, January 1953, p. 214.
H
Helipak Performance Characteristics, Begemean, C. R. and Turkal, P. J. (Laboratory Report of Podbielniak Inc.), 1950.
I
Umholtz, C. L. and Van Winkle, M. Petroleum Refiner, Vol 34, 1955, p. 114 for NH:MCH. Pressure Drop Calculated from data obtained on o- and m-xylene binary.
J
Oldershaw, C. F. Industrial and Engineering Chemistry, Vol 13, 1941, p. 265.
K
Bragg, L. B. Industrial and Engineering Chemistry, Vol 49, 1957, p. 1062.
L
NA = not applicable.

is present, it can be used for pressure drop detection with a 6.1.2.2 The heat density in the flask heaters is approxi-
nitrogen bleed or for mechanical stirring, or both. mately equal to 0.5 to 0.6 W/cm2. This requires the use of
6.1.1.2 If a magnetic stirrer is used with a spherical flask, nickel reinforced quartz fabric to ensure a reasonable service
the flask shall have a slightly flattened or concave area at the life.
bottom on which the magnetic stirrer can rotate without 6.1.2.3 Immersion heaters can be employed in a similar way
grinding the glass. In this case, termination of the thermowell and have the advantage of faster response, but they are more
shall be off center 40 6 5 mm to avoid the magnetic stirring fragile and require a specially designed flask to ensure that the
bar. Boiling chips can be used as an alternative to a stirrer. heating elements remain immersed at the end of the run. When
6.1.1.3 (Warning—While the advantage of visibility in used, their heat density should be approximately equal to 4
glass distillation flasks is desirable, flasks of glass may become W/cm2.
hazardous the larger the charge they contain. For this reason, 6.1.2.4 The upper half of the flask shall be covered with a
glass flasks of a volume greater than 10 L are not recom- mantle to avoid unnecessary heat losses from the upper surface
mended.) and shall have an electric heater supplying about 0.25 W/cm2
6.1.2 Heating System—Heating of the flask shall be pro- at full-rated voltage.
vided in such a way that full boilup can be maintained at a 6.1.3 Fractionating Column—The fractionating column
steady rate at all pressure levels. An electric heating mantle must contain either particulate packing or real plates similar to
covering the lower half of the flask and having one third of the those whose performance characteristics are summarized in
heat in an element located in the bottom central area and the Table 1 and meet the specifications stated in 6.1.3.1 through
remaining two thirds in the rest of the hemisphere is recom- 6.1.3.4. Table 2 lists current North American suppliers of
mended. While proportioning controllers are preferred, heat suitable packings.
input can be manually adjusted by use of a variable auto 6.1.3.1 The internal diameter shall be between 25 and 70
transformer on each circuit, the smaller heater being automati- mm.
cally controlled by an instrument sensing the pressure drop of 6.1.3.2 The efficiency shall be between 14 and 18 theoreti-
the column as registered in a differential pressure instrument or cal plates at total reflux when measured by the procedure
alternatively by direct measurement of distillation rate. described in Annex A1.
6.1.2.1 Minimum wattage required to provide full boilup of 6.1.3.3 The fractionating column shall be comprised of a
crude petroleum is approximately 0.125 W/mL of charge. integral glass column and reflux divider totally enclosed in a
Twice this amount is recommended for quick heat-up. highly reflective vacuum jacket having a permanent vacuum of

Copyright by ASTM Int'l (all rights reserved); 4

 D2892 – 11a
TABLE 2 North American Sources of Commercially Available with a calcium chloride drying tube between them to keep
Packing Materials moisture from collecting in the traps. When analysis of the gas
Name Size Source sample is required, the gas can be collected in an empty plastic
Propak 6 by 6 mm Scientific Development Co. balloon of suitable size either in place of the meter or following
P.O. Box 795
State College, PA 16801
it. The volume of its contents can be determined by calculation
Helipak 2.5 by 4 mm Reliance Glass Works Inc. from the rise in pressure after expanding the sample into an
P.O. Box 825 evacuated vessel of known volume.
Bensenville, IL 60106
Perforated plates 25 and 50 mm Reliance Glass Works Inc. 6.1.7 Fraction Collector—This part of the apparatus per-
P.O. Box 825 mits the collection of the distillate without interruption during
Bensenville, IL 60106
W.A. Sales Inc.
withdrawal of product from the receiver under atmospheric or
419 Harvester Ct. reduced pressure. It also permits removal of product from the
Wheeling, IL 60090 vacuum system, without disturbing conditions in the column.
Knitted wire mesh- Pegasus Industrial Specialties Ltd.
Goodloe multiknit P.O. Box 319 6.1.8 Product Receivers—The receivers shall be of suitable
Agincourt, Ontario MIS 3B9 Canada size for the quantity of crude petroleum being distilled. The
Packed Column Co.
970 New Durham Rd.
recommended capacity is from 100 to 500 mL. They shall be
Edison, NJ 08817 calibrated and graduated to permit reading to the nearest 1 %.
6.2 Distillation Under Reduced Pressure—In addition to the
apparatus listed in 6.1, the apparatus for distillation under
less than 0.1 mPa (;10−6 mm Hg). It shall be essentially reduced pressure shall include the following:
adiabatic when tested in accordance with Annex A3. 6.2.1 Vacuum Pump—The vacuum system shall be capable
NOTE 2—In the case of an adiabatic column when distilling a pure of maintaining smooth pressure operation at all pressure levels.
compound, the internal reflux is constant from top to bottom and is equal It shall have the capacity to draw down the pressure in the
to the reflux at the reflux divider. When distilling crude petroleum, the receiver(s) from atmospheric to 0.25 kPa (2 mm Hg) in less
fractionation occurring in the dynamic holdup will cause a temperature than 30 s so as to avoid disturbance of the system during
gradient to be established with attendant greater amount of internal reflux emptying of receivers under vacuum. Alternatively, a separate
at the bottom of the column.
pump can be employed for this purpose.
6.1.3.4 The column shall be enclosed in a heat insulating 6.2.2 Vacuum Gauge—The point of connection of the
system, such as a glass-fabric mantle, capable of maintaining vacuum gauge to the system shall be as close as practical to the
the temperature of the outer wall of the glass vacuum jacket reflux dividing head. The connecting tubing shall be of
equal to that of the internal vapor temperature. To verify this, sufficient diameter to ensure that no measurable pressure drop
the vacuum jacket shall have a temperature sensor, such as a occurs in the line. In no case shall the vacuum gauge
thermocouple, soldered to about 6 cm2 of thin copper or brass connection be near the vacuum pump.
sheet and fastened to the outer wall of the glass jacket at a level
6.2.2.1 All gauges shall be carefully protected from con-
just below the reflux divider.
densable vapors, especially water vapor, by a cold trap main-
NOTE 3—For certain types of columns there is no significant difference tained at the temperature of dry ice.
in yields and fraction qualities between an uncompensated and a heat-
6.2.3 Pressure Regulator—The regulator shall maintain the
compensated column. In such a case, by mutual agreement between
parties concerned, the application of a heated insulating system can be pressure in the system essentially constant at all operating
omitted. pressures. Automatic regulation can be achieved by a device
that regulates the demand on the vacuum source. A satisfactory
6.1.3.5 The adjustable reflux divider shall be located about
device is a solenoid valve positioned between the vacuum
one column diameter above the top of the packing or topmost
source and a surge tank of at least 10-L capacity. Alternatively,
plate. It must be capable of dividing the condensate with an
a manual bleed valve can be maintained by a trained operator
accuracy of better than 90 % between the column and the
with a minimum of attention.
takeoff line over a range of rates from 25 to 95 % of the
maximum boilup rate of the column when determined in 6.3 Sensing and Recording Apparatus:
accordance with Annex A7. 6.3.1 Temperature Sensors—Only temperature measure-
6.1.4 Condenser—The condenser shall have sufficient ca- ment systems meeting the requirements of 6.3.1.1 and 6.3.1.2
pacity to condense essentially all the C4 and C5 vapors from the shall be used.
crude at the specified rate, using a coolant temperature 6.3.1.1 The vapor temperature sensor can be a platinum
of −20°C. resistance thermometer, a Type J thermocouple with the
6.1.5 Cold Traps—Two efficient traps of adequate capacity junction head fused to the lower tip of the thermowell, or any
cooled by dry ice and alcohol mixture shall be connected in other device that meets the requirements in this paragraph and
series to the vent line of the condenser when light hydrocar- 6.3.1.2. The tip of the sensor shall be located above the top of
bons are present, as at the beginning of the distillation. For the packing or the topmost glass plate and in close proximity to
vacuum distillation, a Dewar-style trap also cooled by dry ice the reflux divider but not in contact with the liquid reflux. The
is used to protect the vacuum gauge from vapors. location of the vapor temperature sensor shall be proved by the
6.1.6 Gas Collector—If uncondensed gas is to be measured, test method described in Annex A4. The sensor shall have a
a gas meter can be connected to the outlet of the cold trap but cooling time of not more than 175 s, as described in Annex A5.

Copyright by ASTM Int'l (all rights reserved); 5

 D2892 – 11a

FIG. 2 Approximate Pressure Drop-Fractionators Using Propak

6.3.1.2 The vapor temperature measuring device shall have 100-13.3 760 to 100 0.13 1.0
an accuracy of 0.5°C or better and be measured with a 13.3-1.33 99 to 10 0.013 0.1
1.33-0.266 9 to 2 0.006 0.06
resolution of 0.1°C or better. The liquid temperature measuring
device shall have an accuracy of 1.0°C or better and be 6.3.2.2 Noncertified gauges shall be calibrated from a non-
measured with a resolution of 0.5°C or better. Temperatures are tilting McLeod gauge or a secondary electronic standard
recorded either manually or automatically. traceable to a primary standard. A basic calibration procedure
6.3.1.3 Temperature sensors shall be calibrated as described is described in Annex A6. Recalibrate when either the sensor or
in Annex A6. Alternatively certified sensors may be used, the instrument is repaired or serviced. Verification of the
provided the calibration of the sensor and its associated calibration of the electronic pressure sensors is to be made on
recording instrument can be traced back to a primary tempera- a regular basis. A frequency of at least once a month is
ture standard. Temperature sensors are calibrated over the full recommended. Verification of the calibration of the sensors can
range of temperature (from 0 to 400°C) at the time of first use be accomplished using the procedures described in Annex A6
of the sensor in combination with its associated instrument. or against a certified reference system.
Recalibrate when either the sensor or the instrument is repaired 6.3.3 Boilup Rate—The boilup rate is normally controlled
or serviced. Verification of the calibration of the temperature by sensing the pressure drop in the column. The pressure drop
sensors is to be made on a regular basis. For vapor temperature during operation is measured by means of a manometer or
sensors, verification at least once a month is recommended and pressure transducer connected between the flask and the
for liquid temperature sensors once every six months. Verifi- condenser. Prevention of condensation in the connecting tube
cation of the calibration of the sensors can be accomplished can be accomplished by injecting a very small flow of nitrogen
potentiometrically by the use of standard precision resistance
(8 cm3/s) between the pressure drop sensor manometer and the
or by distilling a pure compound with accurately known
flask (see Fig. 1) or by placing a small water-cooled condenser
boiling point.
between the flask and the pressure drop sensor. Alternatively,
6.3.2 Vacuum Gauge—A nontilting McLeod gauge or a
the boilup rate can be controlled from the measurement of take
mercury manometer are primary standards and can be used
off rate.
without calibration when properly used and maintained. A
mercury manometer, however, will only be of satisfactory
7. Verification of Apparatus Performance
accuracy down to a pressure of about 1 kPa and then only when
read with a good cathetometer (an instrument based on a 7.1 Test Method D2892 provides a standard framework for
telescope mounted on a vernier scale to determine levels very the laboratory distillation of crude oils in order to produce cuts
accurately). Alternatively, a tensimeter or certified electronic of defined quality (for further testing) and the concurrent
sensors may be used, provided the calibration of the sensor and production of a boiling point curve. As the quantity require-
its associated recording instrument can be traced back to a ments and cut points might be widely different between
primary pressure standard. Sensors of the diaphragm type have companies and application areas, this test method does not
been found satisfactory. Vacuum gauges based on hot wires, standardize on equipment design but on equipment perfor-
radiation, or electrical conductivity detectors are not recom- mance.
mended. 7.2 The nature of the test method (the use of large sample
6.3.2.1 The gauge for measuring subatmospheric pressures quantities and very time consuming) and the nature of the
shall have an accuracy at least equal to that stated as follows: product being tested (highly volatile and unstable material),
Distillation Pressure Accuracy precludes the use of standard statistical control techniques.
kpa mm Hg kPa mm Hg Moreover, this test method does not produce a single result, nor

Copyright by ASTM Int'l (all rights reserved); 6

 D2892 – 11a
is the series of results (the boiling point curve) derived under cluding both correctness of AET and column efficiency. It is the
rigidly defined conditions (see 4.2). responsibility of the laboratory to provide for sufficient quality
7.3 Equipment performance in the context of Test Method controls to guarantee conformance to the test method.
D2892 consists of two elements; the efficiency of the column,
defining cut quality, and the correctness of the cut point (AET), 8. Sampling
defining the boiling point curve. 8.1 Obtain a sample for distillation in accordance with
7.4 The correctness of the AET is mainly, but not only, instructions given in Practice D4057 or D4177. The sample
dependent on the accuracy of the (vapor) temperature and must be received in a sealed container and show no evidence of
(operating) pressure sensors (Annex A6). Other factors affect- leakage.
ing the accuracy and precision of the boiling point curve are: 8.2 Cool the sample to between 0 and 5°C by placing it in
7.4.1 The location of the temperature and pressure sensor a refrigerator for several hours (preferably overnight) before
(Annex A4). opening.
7.4.2 The dynamic response of the sensors (Annex A5). 8.3 If the sample appears waxy or too viscous, raise the
7.4.3 The correct operation of the reflux divider (Annex temperature to 5°C above its pour point.
A7). 8.4 Agitate the sample by whatever means are appropriate
7.4.4 The heat loss from the column (Annex A3). to its size to ensure that it is well-mixed.
7.4.5 The efficiency of the column (Annex A1). 8.5 Determine the water content of the sample by Test
7.4.6 These factors are basically covered through the appro- Method D4006 or any other suitable method. If the water
priate annexes. However, it should be realized that this takes content exceeds 0.3 % volume, the sample shall be dehydrated
only care of individual components and does not cover the prior to fractional distillation. A suitable practice for dehydra-
combined effect of small deviations. Moreover, the aforemen- tion of wet crude oil samples is described in Appendix X1.
tioned tests are all done under more or less static conditions, NOTE 4—Attempts to distill wet crude oil samples in glass columns
not necessarily representative for the behavior of the system might result in breakage of the glassware, which poses a potential fire
under actual dynamic conditions. hazard. Moreover, the presence of water will effect the accuracy of
7.5 Cut quality is mainly defined by the efficiency of the distillation yield in the naphtha region. These effects are more pronounced
column (Annex A1), but is also affected by: for heavy crude oils, containing low amounts of hydrocarbons boiling
below 100°C, than for light crudes where there is usually sufficient
7.5.1 The correct operation of the reflux divider (reflux hydrocarbon vapor generated to form an azeotrope and drive the water
ratio) (Annex A7). vapors through the column without problems.
7.5.2 The heat loss from the column, that is, internal reflux
(Annex A3). 9. Preparation of Apparatus
7.5.3 The dynamic hold up of the column (Annex A2). 9.1 Clean and dry the distillation column and all the
7.5.4 Column efficiency is covered in this test method ancillary glass apparatus before the distillation begins.
through Table 1 and Annex A1. However, Table 1 only 9.2 Ensure that the system is leak-free and all heaters,
provides an assumption on efficiency and is not a guarantee. control devices, and instruments are on and in working order.
Annex A1 only provides a check under static conditions, A clock or other timing device should be ready for use.
infinite reflux ratio, rather low actual temperatures and a binary
component system. Hence, although there is some safeguard on 10. Procedure
standard performance, through conformance to Annex A1, 10.1 Charging:
Annex A2, Annex A3, and Annex A7, again it remains 10.1.1 The charge size shall be such that the dynamic hold
questionable whether this is truly representative for columns up as determined in accordance with Annex A2 is between 1
under actual operating conditions. and 4 % of the charge when operating at 75 % of maximum
7.6 Theoretically, an overall performance check, like the boilup (see Table 1). Chill the flask to a temperature not lower
one described in Appendix X2, is capable of verifying the then 0°C.
performance of a column and the correctness of the AET under 10.1.2 Insert the stirring device or place some pieces of
actual operating conditions. Appendix X2, in principle, mea- glass or porcelain into the flask to control bumping.
sures the combined effect of all factors affecting the results of 10.1.3 Determine the density of the sample by Test Method
Test Method D2892. D941, D1217, or D1298.
7.6.1 The minimum tray number as defined in Appendix X2 10.1.4 Calculate to within 65 % the mass of crude petro-
is a measure of overall cut quality, and the difference between leum corresponding to the desired volume of the charge. Weigh
nominal cut point (AET) and effective cut point (ECP as to the nearest 1 % this quantity of sample into the flask.
defined in Appendix X2) provides a measure for the correct- 10.1.5 Attach the flask to the column and connect the
ness of the AET. However, insufficient data are available right pressure drop measuring device. Install the heating system,
now to define the allowable tolerances in a rigid statistical way. stirrer, and support device. (Warning—Poisonous H2S gas is
Moreover, the test method described is very labor intensive and frequently evolved from crude oil and precautions must be
precludes its use on a regular, short time interval basis and, taken either to absorb the gas that passes through the cold trap
therefore, its use as a mandatory statistical control technique. or to vent it to a safe place.)
7.6.2 Appendix X2, therefore, provides only recommended 10.2 Debutanization:
guidelines for statistical control on column performance, in- 10.2.1 For necessary apparatus refer to 6.1.5 and 6.1.6.

Copyright by ASTM Int'l (all rights reserved); 7

 D2892 – 11a
10.2.2 Begin circulation of refrigerant at a temperature no
higher than −20°C in the condenser, distillate cooler, and
receiver, if so equipped.
10.2.3 Record the barometric pressure at the beginning and
periodically throughout the distillation.
10.2.4 Apply heat to the flask at such a rate that vapors
reach the top of the column between 20 and 50 min after
startup. Adjust heat input so as to achieve a pressure drop of
less than 0.13 kPa/m (1.0 mm Hg/m) in packed columns or less
than 0.065 kPa (0.5 mm Hg) in real plate columns. Program
automated equipment in accordance with the preceding direc-
tions. Turn on the stirring device if used.
10.2.5 Allow the column to operate at total reflux until the
vapor temperature reaches equilibrium but not longer than 15
min after the first drop of condensate appears in the reflux
divider.
10.2.6 Record the vapor temperature as the initial vapor
temperature.
10.2.7 Stop the circulation of the refrigerant and observe the
vapor temperature. When the vapor temperature reaches 15°C,
start the circulation of refrigerant again.
FIG. 3 Expected Takeoff Rates at 5:1 Reflux Ratio for
10.2.8 If the vapor temperature drops below 15°C, continue Fractionators Using Propak
refluxing for at least 15 min. Repeat 10.2.7. If the vapor
temperature remains at 15°C or rises, continue with the
atmospheric distillation. (Warning—The following three steps
evaluation (see Annex A1) and is measured at the bottom of the column.
should not be done until after the first naphtha cut has been The maximum boilup of the n-heptane-methylcyclohexane test mixture is
removed to ensure that all the light gases have been recovered.) that which the column can handle under stable conditions without
10.2.9 Remove and weigh the dry ice traps containing light flooding. In routine adiabatic operation, the boilup rate can be estimated
hydrocarbon liquid after carefully wiping them dry. roughly from the takeoff rate multiplied by the reflux ratio plus one.
10.2.10 Sample the contents of the first dry ice trap using a 10.3.3 Commence takeoff at a reflux ratio of 5:1 and total
10 to 50 mL pressure vessel evacuated to no lower than 26.6 cycle time of not over 30 s nor less than 18 s.
kPa (200 mm Hg). Keep all containers at the temperature of
dry ice to ensure no loss of volatiles. The first trap next to the NOTE 6—The vapor reaching the top of the column is totally condensed
condenser should contain all of the sample. If condensate is and the resulting liquid is divided into two parts. One part L (reflux), is
returned to the column and the other part, D (distillate), is withdrawn as
found in the second trap, sample both traps or combine the product. The reflux ratio (R = L/D), can vary from zero at total takeoff
contents before sampling. (L = 0) to infinity at total reflux (D = 0).
10.2.11 Submit the trap sample and gas balloon, if used, for
analysis by a suitable gas chromatographic test method to be 10.3.4 Take off distillate in separate and consecutive frac-
reported on a fixed-gas free basis. Test Methods D6729, tions of suitable size. The recommended size of fraction is that
D6730, and D6733, equipped with liquid or gas sampling corresponding to 5 or 10°C in vapor temperature. Collect
valves, or both, for sample introduction equipment have been fractions boiling below 65°C in receivers cooled to 0°C or
used successfully for this analysis. below. When the vapor temperature reaches 65°C, refrigerant
10.3 Distillation at Atmospheric Pressure: in the condenser and related coolers can be discontinued and
water at ambient temperature substituted.
10.3.1 Maintain a temperature below −20°C in the lines of
the distillate cooler and receiver as well as in the condenser. 10.3.5 At the end of each fraction and at each cut point,
Turn on the column mantle heat controller and maintain the record the following observations:
column jacket temperature 0 to 5°C below the vapor tempera- 10.3.5.1 Time in hours and minutes,
ture. 10.3.5.2 Volume in millilitres,
10.3.2 Regulate the heat input as necessary to establish and 10.3.5.3 Vapor temperature in °C to the nearest 0.5°C,
maintain a boilup rate approximately 75 % of maximum. Fig. 10.3.5.4 Temperature of the boiling liquid in °C to the
3 can be used as a guide for Propak. Rates for other sizes can nearest 1°C,
be estimated by multiplying the boilup rate in Table 1 by the 10.3.5.5 Atmospheric pressure in kPa (mm Hg), and
cross-sectional area of the column and dividing by the sum of 10.3.5.6 Pressure drop in the column in kPa (mm Hg).
the reflux ratio + 1. 10.3.6 If signs of flooding are observed, reduce the heating
NOTE 5—Boilup rate is expressed in millilitres of liquid per hour for a
rate while continuing takeoff until steady conditions are
given column or in millilitres per hour per square centimetre of cross- restored. If a cut point is encountered during this period, stop
sectional area for comparative purposes. In the latter case, it refers to the the distillation, cool the charge, and recombine the off-
test mixture of n-heptane and methylcyclohexane in the efficiency condition cuts. Restart the distillation with a period at total

Copyright by ASTM Int'l (all rights reserved); 8

 D2892 – 11a
reflux, not to exceed 15 min, to restore operating conditions 10.4.7.2 Volume in millilitres observed at ambient tempera-
before continuing takeoff. Do not make a cut within 5°C of ture,
startup. 10.4.7.3 Vapor temperature in °C to the nearest 0.5°C with
10.3.7 Continue taking cuts until the desired maximum correction, if any,
vapor temperature is reached or until the charge shows signs of 10.4.7.4 Temperature of the boiling liquid in °C to the
cracking. Pronounced cracking is evidenced by a fog appearing nearest 1°C,
in the flask and later at the reflux divider. Do not allow the 10.4.7.5 Pressure drop in the column in kPa (mm Hg),
vapor temperature to exceed 210°C nor the temperature of the 10.4.7.6 Operating pressure measured at the top of the
boiling liquid to exceed 310°C. column in kPa (mm Hg) absolute with correction, if any, and
10.3.8 Shut off the reflux valve and the heating system. 10.4.7.7 AET using the equations given in Annex A8.
Allow the contents to cool to such a temperature that the 10.4.8 Continue taking cuts until the desired maximum
distillation can be commenced at 13.3 kPa (100 mm Hg) point is reached or until the charge shows signs of cracking.
without flooding. This temperature can be estimated by adding Pronounced cracking is evidenced by the evolution of gases as
the DT between the liquid and vapor temperatures found for the indicated by rising pressure as well as a fog appearing in the
column during atmospheric operation to the expected initial flask (see Note 7). Do not allow the temperature of the boiling
vapor temperature at the reduced pressure, or by subtracting liquid to exceed 310°C. (Warning—Automatic vacuum con-
the DT from the last recorded liquid temperature. trollers could mask a slight rise in pressure due to cracking.
NOTE 7—Cooling of the liquid in the flask can be accelerated by Vigilance is required to avoid this.)
blowing a gentle stream of compressed air onto the flask after its heating 10.4.9 Shut off the reflux valve and the heating system.
mantle has been removed. Avoid strong jets of cold air. Alternately, turn Allow the contents to cool to such a temperature that the
on coolant in the quench coil of the flask, if used. distillation can be commenced at a lower pressure without
10.3.9 Weigh all fractions and determine their densities. boiling. This temperature can be estimated by adding the DT
10.3.10 Submit the first distillate fraction for analysis by gas between the liquid and vapor temperatures found for the
chromatography. column during operation to the expected initial vapor tempera-
10.4 Distillation at 13.3 kPa (100 mm Hg): ture at the lower pressure, or by subtracting the DT from the
10.4.1 If further cuts at higher temperatures are required, last recorded liquid temperature.
distillation can be continued at reduced pressures, subject to 10.4.10 Weigh all fractions and determine their densities at
the maximum temperature that the boiling liquid will stand 15°C.
without significant cracking. This is about 310°C in most cases. 10.5 Distillation at Lower Pressures:
Notable exceptions are crude oils containing heat-sensitive 10.5.1 If the final cut point has not been reached, distillation
sulfur compounds. In any case, do not make a cut within 5°C can be continued at a lower pressure subject to the same
of the temperature at startup because the column will not be at limitation as before (see 10.4.1). Only one pressure level
equilibrium. between 13.3 kPa (100 mm Hg) and 0.266 kPa (2 mm Hg) is
10.4.2 Connect a vacuum pumping and control system to permitted. Where the maximum cut point is 400°C AET, the
the apparatus as shown in Fig. 1. minimum pressure is recommended.
10.4.3 Start the vacuum pump and adjust the pressure 10.5.2 Adjust the pressure to the desired level. If the liquid
downward gradually to the value of 13.3 kPa (100 mm Hg) or boils before the pressure is reached, increase the pressure and
set the pressure regulator at this value. The temperature of the cool further until the desired pressure can be achieved without
liquid in the flask must be below that at which it will boil at boiling. Follow the procedure in 10.4.4.
13.3 kPa (100 mm Hg). If the liquid boils before this pressure 10.5.3 Circulate cooling water in the condenser and liquid
is reached, increase the pressure and cool further until the cooler either at ambient temperature or warmed to a tempera-
desired pressure can be achieved without boiling. ture that will ensure that wax does not crystallize in the
10.4.4 Apply heat to the boiler and reestablish reflux at any condenser or takeoff lines. Alternatively, leave the cooling coils
moderate rate in the reflux divider for about 15 min to reheat full of water but vented and not circulating, or else circulate a
the column to operating temperature. Momentarily stop heat stream of air instead of water as a coolant.
input and raise the pressure with N2 for 1 min to drop the 10.5.4 Continue vacuum operation as in 10.4.5 through
holdup into the distillation flask. 10.4.8. During this operation, a reflux ratio of 2:1 is allowed if
10.4.5 Reapply heat to the distillation flask and adjust the mutually agreed upon in advance and noted in the report.
rate of heating to maintain a constant pressure drop equivalent Correct observed and corrected vapor temperatures to AET
to the boilup rate of approximately 75 % of the maximum rate using the equations given in Annex A8.
for this pressure and begin takeoff without delay. The approxi- 10.5.5 Check periodically that the condensate drips nor-
mate pressure drops required for this purpose are indicated in mally in the condenser and that the distillate flows smoothly
Fig. 3. Maintain a column insulation temperature 0 to 5°C into the takeoff line. If crystallization is observed, allow the
below the vapor temperature throughout the operation. coolant in the condenser to warm as in 10.5.3.
10.4.6 Remove separately, cuts of suitable size as in 10.3.4. 10.5.6 When the final cut point has been reached, or when
10.4.7 At the end of each distillate fraction and at each cut limits of boiling liquid temperature and column pressure
point, record the following observations: prevent further distillation, turn off the reflux valve and heating
10.4.7.1 Time in hours and minutes, system and allow to cool with the vacuum still applied.

Copyright by ASTM Int'l (all rights reserved); 9

 D2892 – 11a
10.5.7 When the temperature of the residue in the flask has 10.5.9 Weigh all fractions and the residue in the flask and
fallen below 230°C, shut off the vacuum pump. Vent the determine their densities at 15°C by Test Method D4052 or by
fractionating unit with nitrogen or other inert gas. Do not use another suitable method. Convert the density to 15°C, if
air. (Warning—Air is suspected of initiating explosions in necessary.
fractionating units that are vented while too hot, such as at the
end of a run.) NOTE 8—Heavier flasks, such as those for 50 and 70-mm diameter
10.5.8 Stop circulation of coolant in the condenser and columns, are not normally removed for weighing. In these cases the
residue can be discharged at a temperature not over 200°C into a tared
ancillary equipment. Disconnect the flask. Recover the static
container for weighing. Nitrogen pressure of approximately 6.7 kPa (50
holdup of the column (wettage) by distilling a small quantity of mm Hg) will be sufficient for this. Wettage in these cases will include that
solvent such as toluene in a separate flask to wash the column, of the column and the flask together.
condenser, and takeoff system. Evaporate the solvent from the
collected residue at 10°C above the boiling point of the solvent, 11. Calculation
using a small purge of nitrogen. For distillations not involving
disagreement, or by mutual consent, the holdup can be esti- 11.1 Calculate the mass % of each distillate fraction and the
mated using a graph similar to Fig. 4. The density of the holdup residue to the 0.1 mass %, using Eq 1.
is estimated by extrapolation of the density line for the mass % 5 100~m/M! (1)
preceding cuts. The static holdup can be treated as a separate
small cut or blended into the bottoms before inspections are where:
made. The latter must be done if other analyses besides density m = mass of fraction or residue, g, and
M = mass of dry crude oil charged, g.
are to be performed on the residue.

FIG. 4 Approximate Static Holdup for Average Crude Oil Using 4 mm Propak in a 25-mm ID 3 570-mm Column

Copyright by ASTM Int'l (all rights reserved); 10

 D2892 – 11a
11.1.1 The first fraction is the gas fraction collected in the 12.1.5 The volume and mass % of each fraction to the
balloon. If this fraction is less than 0.1 mass %, it can be nearest 0.1 %,
ignored. 12.1.6 The cumulative volume and mass percentages, and
11.1.2 The second fraction (or first, if no gas is collected) is 12.1.7 The mass of water, if any.
the condensate in the dry ice trap. With density at 15°C 12.2 The gas, debutanized naphtha, and succeeding frac-
calculated from the gas chromatographic data on a fixed gas tions are listed in order of ascending boiling point with residue
free basis, its volume can be computed. recorded last.
11.1.3 The holdup is treated either as a separate cut or added 12.3 The observations made in 10.3.5, 10.4.7, and 10.5.4 are
to the residue fraction, in accordance with agreement. The included as a second sheet, which is normally attached to the
amount of holdup is determined by actual recovery by solvent summary sheet.
washing, as directed in 10.5.8, or estimated from Fig. 4. 12.4 Make plots of the temperature in degrees Celsius AET
11.2 Calculate the percent loss to the nearest 0.1 mass %, as the ordinate (y-axis) versus the percentage mass and volume
using Eq 2. distilled as the abscissa (x-axis). These are the final TBP
distillation curves.
Loss 5 100 2 ~ (100~m/M!! (2)
The weight loss as calculated above must not be greater than 13. Precision and Bias 4
0.4 %, otherwise the distillation must be discarded. Losses less 13.1 The precision of this test method, as determined by the
than this should be allocated two thirds to the trap cut and one statistical examination of interlaboratory test results obtained
third to the first naphtha cut. Where there is no trap cut, the from eight laboratories on five crude oil samples ranging in
acceptable losses are to be normalized among all cuts. gravity from 20.0° to 41.0° API. The sample set contained
11.3 Calculate the volume of the sample of crude oil in crudes with high and low sulfur content, high and low
millilitres at 15°C, using Eq 3. naphthenic acid content, and asphaltic types. The results of the
V 5 ~M/D! (3) statistical examination follow, and the statistical evaluation and
validation can be found in the research report.
where: 13.2 Repeatability—The difference between successive re-
D = density of charge at 15°C, g/mL, sults obtained by the same operator with the same apparatus
M = mass of dry charge, g, and under constant operating conditions on identical test material
V = volume of charge, mL. would, in the long run, in the normal and correct operation of
11.4 Calculate the volume of each fraction and of the this test method, exceed the values indicated as follows in one
residue in millilitres at 15°C, using Eq 4. case in twenty.
v 5 m/d (4) Mass % Degrees of
Freedom
where: Atmospheric pressure 0.6 19
d = density of the fraction or residue at 15°C, g/mL, 13.33 kPa Vacuum pressure 0.9 19
1.33 kPa Vacuum pressure 0.9 19
m = mass of fraction or residue corrected for loss, g, and
v = volume of fraction, mL. 13.3 Reproducibility—The difference between two single
11.5 Calculate the volume % of each distillate fraction to the and independent results obtained by different operators work-
nearest 0.1 volume %, using Eq 5. ing in different laboratories on identical test material would, in
vol % 5 100~v/V! (5)
the normal and correct operation of this test method, exceed the
values indicated as follows in one case in twenty.
11.6 Calculate the volume % gain or loss to the nearest 0.1 Mass % Degrees of
volume %, using Eq 6. Freedom
Atmospheric pressure 1.3 16
Loss 5 100 2 ~ (100~v/V!! (6) 13.33 kPa Vacuum pressure 1.5 25
1.33 kPa Vacuum pressure 2.0 15
Usually, the above expression is negative due to volume
expansion. Normalize any apparent expansion or contraction 13.4 Bias:
among fractions boiling below 150°C in proportion to their 13.4.1 Absolute Bias—Since there is no accepted reference
yields. material suitable for determining the bias for the procedure in
Test Method D2892 in determining distillation properties of
NOTE 9—In view of the foregoing rules for establishing yields, the ratio
crude petroleum, bias cannot be determined.
of mass to volume is not precise enough to be used to calculate the density
of any distillate fractions or residue. 13.4.2 Relative Bias—TBP is defined under the conditions
of this test method (see 1.2).
12. Report NOTE 10—The degrees of freedom associated with the repeatability/
12.1 A summary sheet for the run must include: reproducibility estimate from this round robin study ranged from 15 to 25.
Since the minimum requirement of 30 (in accordance with Practice
12.1.1 The mass of the dry sample charged, g,
D6300) is not met, users are cautioned that the actual repeatability/
12.1.2 The density of the sample at 15°C, g/mL,
12.1.3 The volume of the sample at 15°C, mL,
12.1.4 The gain or loss in mass and volume to the nearest 4
Supporting data have been filed at ASTM International Headquarters and may
0.1 %, be obtained by requesting Research Report RR:D02-1705.

Copyright by ASTM Int'l (all rights reserved); 11

 D2892 – 11a
reproducibility may be significantly different than these estimates. Given 14. Keywords
the fact that it is very difficult to get laboratories to participate, coupled
with the logistical difficulties of shipping precision study samples in the 14.1 boiling point distillation; crude oil distillation; distilla-
quantities required, the members of Subcommittee D02.08 have decided tion; fractional distillation; TBP curves
to publish all precision data, despite the lower degrees of freedom.

ANNEXES

(Mandatory Information)

A1. TEST METHOD FOR THE DETERMINATION OF THE EFFICIENCY OF A DISTILLATION COLUMN

A1.1 Scope described. Cleaning can be accomplished by washing with a

A1.1.1 This test method is for determining the efficiency of strong industrial detergent. Rinse thoroughly, dry, and reas-
a distillation column, under total reflux conditions using the semble.
test mixture n-heptane/methylcyclohexane at atmospheric A1.5.2 Distill a small quantity of pure n-heptane at a high
pressure. boilup rate for at least 5 min. Take off several small quantities
A1.1.2 The efficiency is not measured under vacuum con- through the overhead sampling system at intervals. Turn off the
ditions because there is no satisfactory test mixture that has a heat, remove the flask, and dry the column with air while still
constant relative volatility with pressure. hot. To keep the apparatus completely dry, connect a moisture
trap at the vent of the overhead condenser.
A1.2 Significance and Use
A1.6 Procedure
A1.2.1 The efficiency of the distillation column must be
between 14 and 18 theoretical plates to be used in Test Method A1.6.1 Introduce into the distillation flask a quantity of test
D2892 (see 6.1.3.2). mixture equal to the minimum volume permitted for the
A1.2.2 The performance of particulate packings is well column (see Table 1) but not more than two thirds of the
established in the literature. The data shown in Table 1 can be capacity of the flask. Add some pieces of glass or porcelain to
used in place of this test method. promote even boiling.
A1.6.2 Connect the flask to the pressure drop manometer as
A1.3 Apparatus shown in Fig. A1.1.
A1.3.1 An example of a suitable apparatus is shown in Fig. A1.6.3 Circulate water at ambient temperature in the con-
A1.1. It consists of the following: denser.
A1.3.1.1 Calibration Flask, of suitable size with a device A1.6.4 Apply heat to the flask until the test mixture boils,
for heating. An example of a suitable calibration flask is shown then increase the heat progressively up to the flood point. This
in Fig. A1.2. will be noted by visible slugs of liquid in the packing or on the
A1.3.1.2 Distillation Column and Condenser. plates or by liquid filling the neck of the condenser and a
A1.3.1.3 Manometer, or equivalent to measure the pressure sudden increase in pressure drop. Reduce heat to allow the
drop in the column. flooding to subside. Increase the heat gradually to just below
the flood point. Record boilup and DP measurements just
A1.4 Reagents and Materials below the flood point. This is considered the maximum rate.
A1.4.1 The test mixture is a 50/50 mixture by volume of NOTE A1.1—Under these conditions no vapor can reach the head and
n-heptane and methylcyclohexane with refractive indexes of: the heat to the distillation flask must be reduced to establish normal
n-heptane nD20 = 1.38764 operations again. The flood point is normally determined during the
methylcyclohexane nD20 = 1.42312 efficiency evaluation of a column using the n-heptane-methylcyclohexane
test mixture (see Annex A1).
A1.4.2 Chromatographic analysis of the components of the
test mixture must show less than 0.01 % contamination with A1.6.5 Adjust the heat input until the column is refluxing at
lighter compounds and greater than 99.75 % purity. They must a steady rate corresponding to about 200 mL/h times the
be transparent to ultraviolet light at 260 to 270 nm to ensure cross-sectional area of the column in cm2.
freedom from aromatics. A1.6.6 Maintain the column under total reflux for 1 h.
A1.4.3 If the components do not meet the above specifica- Record the pressure drop and determine the boilup rate by
tion, they can be purified either by redistillation in a 30 to 50 timing the filling of the calibrated bulb.
plate column, or by percolation through 200-mesh silica gel, A1.6.7 Take rapidly, and almost simultaneously, a sample of
discarding the first 10 % of the eluted liquid. Exercise care that the flask liquid and reflux in sufficient quantity for the
the gel does not become overloaded. determination of the refractive indexes at 20°C but not more
than 0.5 mL each. The duration of withdrawal of the reflux
A1.5 Preparation of Apparatus sample must not exceed 2 s.
A1.5.1 The distillation column and all the glassware must A1.6.8 Measure refractive indexes of reflux and flask
be clean and dry before proceeding with the test method samples.

Copyright by ASTM Int'l (all rights reserved); 12

 D2892 – 11a
A1.6.9 Repeat A1.6.7 and A1.6.8 at intervals from 12 to 30 ria have been determined by Adler et al.6 and relative volatility
min until refractive index measurements indicate that steady is taken as 1.075 as recommended by IUPAC.7
conditions of maximum efficiency have been attained. A1.7.4 The efficiency of the column is numerically equal to
A1.6.10 Check the ultraviolet absorption of the flask sample the difference between the plate numbers indicated on the
at 260 and 270 nm. If an absorption is detected, aromatics curve for top and bottom samples. Note that one plate is
contamination is present and the efficiency determination will subtracted for the contribution of the flask, or alternatively a
be in error. The apparatus must be cleaned and the test mixture sample of liquid from the bottom of the column rather than the
repercolated through fresh silica gel before starting over. flask can be taken to obtain the efficiency of the column alone.
A1.6.11 Repeat successively the series of operations A1.6.5 A1.7.5 For packed columns, draw curves of HETP in
to A1.6.10, without preflooding, at four or more other boilup millimetres, DP in kPa and kPa/m, and number of theoretical
rates approximately equally spaced between (200 mL/ plates as a function of boilup rate expressed in mL/h and
h) 3 cm2 and the maximum rate. One of these should be at a absolute rate in (mL/h) 3 cm2. For real plate columns, the
rate that is above 90 % of maximum. efficiency should be plotted as percent of a theoretical plate per
real plate and DP should be plotted both in kPa/real plate and
A1.7 Calculation in absolute units of kPa/theoretical plate using the same units
A1.7.1 For each pair of samples taken, determine the molar for boilup rate as above for packed columns. The efficiency of
composition of each sample by comparison of its refractive the column corresponds to the value of N determined from the
index with a curve of refractive index versus molar composi- curve at 75 % of its maximum boilup rate.
tion of n-heptane drawn from the following data: A1.7.6 Typical curves for the popular packings and plate
Molar fraction of n-heptane Refractive Index columns are shown in Figs. A1.4-A1.7. These columns are to
0.00 1.4231 be used in the unpreflooded condition. Data on other packings
0.10 1.4191
0.20 1.4151 are inconclusive or incomplete in the literature and are not
0.30 1.4113 included.
0.40 1.4076 A1.7.7 Draw a vertical line on the graph at 75 % of
0.50 1.4040
0.60 1.4005 maximum boilup and read the HETP or the plate efficiency
0.70 1.3971 corresponding to this rate.
0.80 1.3939 A1.7.8 Multiply the HETP in millimetres by 14 and by 18
0.90 1.3907
1.00 1.3876 for packed columns to obtain the range of permissible heights
of packing for use in this test method. If the efficiency of a
A1.7.2 Calculate the number of theoretical plates in the packed column is greater than 18 plates, as measured above,
column by means of the Fenske equation5 (Eq A1.1): the efficiency can be reduced by removing a suitable amount of
XD XO packing without detriment to performance.
log 1 2 X 2 log 1 2 X
N 5
D O
21 (A1.1)
A1.7.9 For real plate columns, divide 14 and 18 by plate
log a efficiency to obtain the number of actual plates (rounded to the
where: nearest integer) for an acceptable column.
N = number of theoretical plates in the column, A1.8 Precision and Bias
a = relative volatility of n-heptane to methylcyclo-
hexane, A1.8.1 No statement is made concerning either the preci-
XD = molar fraction of n-heptane in the reflux liquid, and sion or bias of Annex A1 for measuring the efficiency of a
XO = molar fraction of n-heptane in the flask liquid. distillation column because the result is used to determine
A1.7.3 Fig. A1.3 is a graphical solution to the equation for whether there is conformance to the criteria stated in Test
the n-heptane-methylcyclohexane binary. Vapor-liquid equilib- Method D2892.

5 6
A more convenient form of the equation from Fenske, M. R., Industrial and Adler, et al., American Institute of Chemical Engineers , Vol 12, 1966, p. 629.
7
Engineering Chemistry, Vol 24, 1932, p. 482. Seig, Chemie-Ingenieur-Tecknik, Vol 22, 1950, p. 322.

Copyright by ASTM Int'l (all rights reserved); 13

 D2892 – 11a

FIG. A1.1 Determination of Efficiency

FIG. A1.2 Boiling Rate Timer

Copyright by ASTM Int'l (all rights reserved); 14

 D2892 – 11a

FIG. A1.3 Refractive Index

FIG. A1.4 4-mm Propak in 25-mm Inside Diameter Towers

Copyright by ASTM Int'l (all rights reserved); 15

 D2892 – 11a

FIG. A1.5 4-mm Propak in 50-mm Inside Diameter Towers

FIG. A1.6 6-mm Propak in 50-mm Towers

Copyright by ASTM Int'l (all rights reserved); 16

 D2892 – 11a

NOTE—(1) Analytical Chemistry, ANCHA, Vol 39, 1967, p. 286. (2) Exxon unpublished data.
FIG. A1.7 Perforated Plates 50-mm Inside Diameter

A2. TEST METHOD FOR THE DETERMINATION OF THE DYNAMIC HOLDUP OF A DISTILLATION COLUMN

A2.1 Scope A2.4 Apparatus

A2.1.1 This test method is for determining the dynamic A2.4.1 The apparatus is identical to that described in A1.3,
holdup of a distillation column using a test mixture of stearic Fig. A1.1.
acid in n-heptane.
A2.5 Reagents and Materials
NOTE A2.1—Dynamic hold-up is expressed as a percentage of the
packed volume for packed columns so that the data can be compared. For A2.5.1 The test mixture consists of 20 mass % stearic acid
real plate columns, it is expressed in millilitres per plate. The data can only in n-heptane. The n-heptane shall have a refractive index at
be compared with others of the same diameter because of different tray 20°C of 1.3878 6 0.0002. The stearic acid shall be greater than
spacing. Data for packed columns cannot be compared with those of real 95 % pure and have a melting point of 68 to 70°C.
plate columns except in absolute units of millilitres per theoretical plate
(see Table 1). Dynamic hold-up increases with increasing distillation rate
up to the flood point and varies from one kind of fractionator to another. A2.6 Preparation of Apparatus
A2.6.1 See A1.5.
A2.2 Summary of Test Method
A2.2.1 A test mixture, composed of stearic acid in A2.7 Procedure
n-heptane, is distilled under total reflux conditions. From the A2.7.1 Measure the concentration of stearic acid in the test
difference in concentration of stearic acid in the initial mixture mixture by a convenient means. For example, titrate with 0.1 N
and in the mixture during refluxing, the dynamic holdup of the NaOH solution to a pH of 9.0 potentiometrically. Record the
column is calculated. results as mass % stearic acid (PO).
A2.7.2 Introduce pieces of glass or porcelain into the flask
A2.3 Significance and Use or use a good stirrer to promote even boiling.
A2.3.1 The amount of sample charged to a particular A2.7.3 Add 1 L of the test mixture for a 25-mm inside
distillation column must be of such a size that the dynamic diameter column or 4 L for a 50-mm column to the flask.
holdup of that column is between 2 and 4 % of that charge size Weigh the flask to the nearest 1 g.
at 75 % boilup rate (see 9.1.1). A2.7.4 Attach the flask to the distillation column and to the
A2.3.2 The performance of the particle packings is well DP measuring system.
enough defined in the literature that the data in Table 1 can be A2.7.5 Circulate water at ambient temperature through the
used instead of this test method. condenser.

Copyright by ASTM Int'l (all rights reserved); 17

 D2892 – 11a
A2.7.6 Apply heat to the flask and bring the test mixture to d = density of n-heptane, 0.688 g/mL at 15°C.
a boil. Adjust the boilup rate to approximately 200 mL/h times A2.8.2 Calculate the dynamic holdup per theoretical plate at
the cross-sectional area of the column in centimetres squared each rate of boilup at which the determination was made, using
measured by timing the filling of the calibrated bulb. When the Eq A2.2:
desired rate has been established, hold for 30 min noting the
H
pressure drop. h 5N (A2.2)
A2.7.7 Sample the reflux and the flask liquids, taking
samples of not more than the amount necessary for a determi- where:
nation, and immediately observe and record the boilup rate and N = efficiency in theoretical plates of the column under
DP. total reflux at this boilup rate (see Annex A1).
A2.7.8 Analyze the reflux and flask samples to determine A2.8.3 Convert the boilup measurement in litres per hour to
the concentration of stearic acid in mass % (P). rates per hour in (mL/h) 3 cm2. Convert also the DP measure-
A2.7.9 Repeat A2.7.7 and A2.7.8 at 15 min intervals until ments to kPa/m (mm Hg/m) and the holdup measurement to
the concentration of stearic acid in the flask sample is steady. millilitres per theoretical plate.
A2.7.10 Raise the boilup rate by an additional 200 (mL/ A2.8.4 Plot all data as ordinates versus boilup rate in litres
h) 3 cm2 and repeat A2.7.7-A2.7.9. per hour and (mL/h) 3 cm2 as abscissa. Draw smooth curves
A2.7.11 Continue making measurements as above at incre- through the points for holdup in millilitres and in millilitres per
ments of about 200 (mL/h) 3 cm2 until near the flood point. At theoretical plate. The boilup versus DP measurements should
least four sets of measurements should be obtained including be compared with those made in the efficiency measurements if
one near the maximum operable rate. available, to ensure that they are in reasonable agreement.
A2.8 Calculation A2.8.5 Draw a vertical line at 75 % of maximum boilup
A2.8.1 Calculate the dynamic holdup of the column for rate. The dynamic holdup for the column will be that read from
each observation, using Eq A2.1: the intersection of the holdup curve and the line for 75 % of
maximum boilup rate. The charge size, ranging from a holdup
P 2 PO M ratio from 2 to 4 % will thus be from 50 to 25 times the above
H 5 P 3 d (A2.1)
figure.
where:
H = dynamic holdup in column, mL at 15°C, A2.9 Precision and Bias
M = mass of test mixture in the flask, g, A2.9.1 No statement is made concerning either the preci-
PO = stearic acid in the test mixture initially, mass %, sion or bias of Annex A2 for measuring dynamic holdup
P = stearic acid in the test mixture after distillation, mass because the result is used to determine whether there is
%, and conformance to the criteria stated in Test Method D2892.

A3. TEST METHOD FOR THE DETERMINATION OF THE HEAT LOSS IN A DISTILLATION COLUMN
(STATIC CONDITIONS)

A3.1 Scope A3.3.2 The test should be performed on all new glass
A3.1.1 This test method is for determining the heat loss of vacuum jackets before use, and checked at least once per year
a distillation column under static conditions when a tempera- thereafter.
ture differential exists between the inner and outer walls of a A3.3.3 The heat loss as determined by this test method must
distillation column. be less than 30 % for the column to be acceptable for use in
Test Method D2892.
A3.2 Summary of Test Method
A3.2.1 The outer wall of the column vacuum jacket is A3.4 Apparatus
maintained at an elevated constant temperature. The tempera-
ture increase inside the column, as recorded by the sensor in A3.4.1 The column is enclosed in its heat compensating
the reflux divider, is a measure of the heat gained and thus heat mantle with the thermocouple in place on the column wall.
lost by the column. A3.4.2 A twin pen chart recorder to monitor the column
outer wall and heat sensor temperatures and an automatic
A3.3 Significance and Use
proportioning controller for the heat input are recommended.
A3.3.1 It is important to have an effective silvered glass
vacuum jacket surrounding the column and reflux divider. This A3.5 Preparation of Apparatus
reduces the effects of ambient air temperature near the distil-
lation apparatus and promotes easier control at the maximum A3.5.1 The heat sensor location, response time, and calibra-
distillation temperatures. The use of a heat compensating tion must be checked as specified in Annex A4 to Annex A6.
mantle further reduces losses by reducing the temperature A3.5.2 To reduce chimney effects inside the column during
gradient between inside of the column and the ambient air. the test, the top of the condenser must be closed.

Copyright by ASTM Int'l (all rights reserved); 18

 D2892 – 11a
A3.6 Procedure B2A
Q 5 C 2 A 3 100 (A3.1)
A3.6.1 Record the time and ambient temperature.
A3.6.2 Apply heat to the outer wall of the column vacuum where:
jacket. Adjust the heat progressively until the temperature A = ambient temperature, °C,
sensor located on the column wall records a temperature 100°C B = temperature inside column, °C, and
above ambient. This condition provides a suitable temperature C = temperature of outer wall, °C.
difference to measure without placing unnecessary thermal or, when the differential temperature is 100°C, use Eq A3.2:
strain on the glassware. Attain the 100°C differential tempera-
ture within 30 min and maintain the column at this temperature Q5B2A (A3.2)
for 1 h.
A3.6.3 Record the time, the temperature of column outer A3.8 Precision and Bias
wall, and the temperature inside the column. A3.8.1 No statement is made concerning either the preci-
A3.7 Calculation sion or bias of Annex A3 for measuring heat loss because the
A3.7.1 Calculate the heat gained and thus lost by the result is used to determine whether there is conformance to the
column using Eq A3.1: stated criteria (see A3.3.3).

A4. TEST METHOD FOR THE VERIFICATION OF TEMPERATURE SENSOR LOCATION

A4.1 Scope A4.4.1.4 If the observed boiling point is not 253.5 6 0.5°C,
A4.1.1 This test method is for determining whether the the location of the sensor is suspect and must be corrected
temperature sensor is in the proper position for optimum before use in this test method.
performance. A4.4.2 Vacuum Distillation:
A4.2 Summary of Test Method A4.4.2.1 Assemble the apparatus for vacuum distillation.
Charge 0.5 to 1 L of pure, dry n-hexadecane containing less
A4.2.1 The vapor temperature of a pure compound mea-
than 0.1 % light contaminants as determined by gas chroma-
sured by the sensor and its recording instrument is compared to
tography. Calibrate the temperature and vacuum instruments as
the accepted boiling point for the compound. The test is
prescribed in Annex A6.
conducted both at atmospheric pressure and under vacuum of
0.133 kPa (1 mm Hg). A4.4.2.2 Reduce the pressure in the system to 0.133 kPa (1
mm Hg). Circulate air in the condenser to avoid crystallization.
A4.3 Significance and Use
A4.4.2.3 Apply heat and establish equilibrium at a boilup
A4.3.1 A poorly positioned sensor can give temperatures rate between 25 and 50 (mL/h) 3 cm2. This corresponds to a
that are in error due to inadequate heat supply from the vapors. DP of about 0.133 kPa/m (1 mm Hg/m) or less for particle
It is especially important under vacuum when heat content of packing or 0.08 kPa/plate (0.06 mm Hg/plate) for real plate
the vapor is at a minimum. columns to 50-mm diameter. Hold at these conditions for 15
A4.3.2 The procedure is normally performed once only for min.
the approval of a design and need not be repeated thereafter.
A4.4.2.4 Remove 2 % overhead at 20:1 reflux ratio and then
A4.4 Procedure again hold at total reflux for 5 min. Record the boiling point to
A4.4.1 Atmospheric Distillation: the nearest 0.5°C and the operating pressure to the nearest 0.05
A4.4.1.1 Assemble the apparatus for atmospheric distilla- kPa (0.3 mm Hg), using instruments calibrated as described in
tion. Use cooling water at ambient temperature. Charge 0.5 to Annex A6.
1 L of pure, dry n-tetradecane containing less than 0.1 % light A4.4.2.5 If the observed boiling point is not steady with
contaminants as determined by gas chromatography. Calibrate 0.2°C at 0.133 kPa (1 mm Hg), remove 2 % cuts at 5:1 reflux
the temperature sensor as prescribed in Annex A6. ratio and then hold at total reflux for 15 min, repeating the
A4.4.1.2 Apply heat and establish equilibrium at a boilup above until total reflux conditions produce a steady tempera-
rate of about 400 (mL/h) 3 cm2. This corresponds to a DP of ture. Not more than three trials should be needed.
about 0.4 kPa/m (3 mm Hg/m) for particle packing or 0.09 A4.4.2.6 If the observed boiling point is not 105.2 6 0.5°C
kPa/plate (0.07 mm Hg/plate) for real plate columns. Hold at at 0.133 kPa (1 mm Hg), the location of the sensor is suspect
these conditions for 15 min. If the vapor temperature drops and must be corrected before use in this test method.
more than 0.2°C, remove 2 % overhead at 5:1 reflux ratio and
then again hold at total reflux for 15 min. Continue taking 2 %
A4.5 Precision and Bias
cuts as above until the vapor temperature remains steady within
0.2°C for 15 min at total reflux. A4.5.1 No statement is made concerning either the preci-
A4.4.1.3 Record the boiling point to the nearest 0.5°C and sion or bias of Annex A4 for checking the location of the
the atmospheric pressure to the nearest 0.1 kPa (1 mm Hg), temperature sensor because the result is used to determine
using instruments calibrated in accordance with Annex A6. whether there is conformance to the criteria stated in A4.4.2.6.

Copyright by ASTM Int'l (all rights reserved); 19

 D2892 – 11a

A5. TEST METHOD FOR DETERMINATION OF TEMPERATURE RESPONSE TIME

A5.1 Scope connect the sensor to a strip chart recorder of suitable range
A5.1.1 This test method is for the determination of tempera- allowing interpolation to 0.1°C. Set the chart speed at 30 cm/h
ture response time based upon the rate of cooling of the sensor for readability.
under prescribed conditions. A5.3.3 Insert the sensor into a hole in the center of one side
of a closed cardboard box about 30 cm on a side. Hold the
A5.2 Significance and Use sensor in place by a friction fit on the joint. Allow the sensor to
reach equilibrium temperature. Record the temperature when it
A5.2.1 This test method is performed to ensure that the
becomes stable.
sensor is able to respond to changes in temperature fast enough
A5.3.4 Remove the sensor and insert it into the heated
that no error due to lag is introduced in a rapidly rising
thermowell in the beaker of water. After the sensor has reached
temperature curve.
a temperature of 80°C, remove it and immediately insert it into
A5.2.2 The importance of this test method is greatest under
the hole in the box. Note with a stopwatch, or record on the
vacuum conditions when the heat content of the vapors is
strip chart, the time interval while the sensor cools from 30°C
minimal.
above to 5°C above the temperature recorded in A5.3.3.
A5.3.5 A time interval in excess of 175 s is unacceptable.
A5.3 Procedure
A5.3.1 Arrange a 1-L beaker of water on a hot plate with a A5.4 Precision and Bias
glass thermowell supported vertically in the water. Maintain A5.4.1 No statement is made concerning either the preci-
the temperature of the water at 90 6 5°C. sion or bias of Annex A5 for measuring temperature response
A5.3.2 Correct the sensor to an instrument, preferably with because the result is used to determine whether there is
a digital readout, with readability to 0.1°C. Alternatively, conformance to the criteria stated in A5.3.5.

A6. PRACTICE FOR CALIBRATION OF SENSORS

A6.1 Principle
A6.1.1 This practice deals with the basic calibration of
temperature sensors and vacuum sensors and their associated
recording instruments.
A6.1.2 The temperature sensor and its associated instrument
are calibrated by observing and recording the temperatures of
the melting point and boiling point of pure compounds or
eutectic mixtures.
A6.1.3 The vacuum sensor and its associated instrument are
calibrated against a McLeod gauge.

A6.2 Temperature Sensors

A6.2.1 Apparatus—A suitable apparatus is shown in Fig.
A6.1. For the freezing point of water, a Dewar flask filled with
crushed ice and water can be substituted. For the boiling point
of water, use an equilibrium still or ebulliometer, a tensimeter,
or other apparatus for measuring vapor-liquid equilibrium.
A6.2.2 Procedure:
A6.2.2.1 Ensure that approximately 0.5 mL of silicone oil or
other inert liquid is in the bottom of the thermowell and insert
one or more thermocouples or other sensors connected to their
respective measuring instruments.
A6.2.2.2 Heat the melting point bath to a temperature 10°C
above the melting point of the metal inside and hold this
temperature at least 5 min to ensure that all of the metal is
melted.
A6.2.2.3 Discontinue heat input to the melting point bath
and observe and record the cooling curve. When the curve
FIG. A6.1 Melting Point Bath for Temperature Standards
exhibits a plateau of constant temperature for longer than 1
min, the temperature of the recorded plateau is accepted as the

Copyright by ASTM Int'l (all rights reserved); 20

 D2892 – 11a
calibration temperature. If the freezing plateau is too short, it nometer is the primary reference standard. Fig. A6.3 shows a
can be prolonged by employing some heat during the cooling convenient design in which any internal air above the mercury
cycle. Alternatively, the melt bath may have become contami- is trapped in the fine tip at the top where it can be seen at a
nated or excessively oxidized. In this case, replace the metal. glance. For a suitably evacuated gauge, there is no air in the tip
A6.2.2.4 Record the calibration temperature at each of the visible to the unaided eye when vented. The glass shall be
following points to the nearest 0.1°C: chemically clean inside.
Material Temperature, °C
Ice melting point 0.0
A6.3.1.3 If air can be seen in the tip, clamp the gauge in a
Water boiling point 100.0 horizontal position with the vacuum connection facing upward
Tin: Lead: Cadmium (50:32:18) melting point 145.0 and the top of the gauge lower than the bottom so as to expose
Sn melting point 231.9
Pb melting point 327.4
the hole in the bottom of the central tube. Apply a vacuum of
less than 0.0133 kPa (0.1 mm Hg) and heat the end of the
A6.2.2.5 Set up a correction table by listing the correction to gauge containing the mercury with an infrared heat lamp or a
be added algebraically to the observed temperature to give the hot air gun until the mercury comes to its boiling point.
true temperature at each calibration point. A graphical plot of
Continue heating slowly until the mercury in the central tube
the above corrections connected by a smooth curve may be
has partly distilled out into the outer tube. Condensed mercury
helpful in routine use.
on the wall will be evidence of this. Return the manometer to
A6.3 Vacuum Sensors the vertical position and slowly release the vacuum. Verify that
A6.3.1 Apparatus: there is no air visible at the tip of the central tube when fully
A6.3.1.1 Assemble a vacuum manifold such as that shown vented. (Warning—Mercury vapor is poisonous. Harmful or
in Fig. A6.2. It shall be capable of maintaining steady pressures fatal if inhaled or ingested.)
within 1% at all desired levels. A6.3.1.4 For pressures below 13.3 kPa, the nontilting
A6.3.1.2 For pressures between atmospheric down to 13.3 McLeod gauge is the primary standard and shall be carefully
kPa (100 mm Hg), an evacuated mercury barometer or ma- maintained.

FIG. A6.2 Calibration of Vacuum Gauges

Copyright by ASTM Int'l (all rights reserved); 21

 D2892 – 11a
pressure below 10 Pa (0.075 mm Hg). Thereafter, carefully
protect the reference gauge from exposure to moisture such as
that from atmospheric air. The use of two reference McLeod
gauges of different pressure ranges is recommended as a
precaution. If they agree at the test pressure, it is an indication
that the system is free of moisture and other condensables.
NOTE A6.1—The general principles of construction of McLeod gauges
are well-established. The dimensions and tolerances of such a gauge are
beyond the scope of this test method.
A6.3.1.6 Alternatively, certified secondary gauges, elec-
tronic or otherwise, can be used, provided the output can be
traced back to a primary standard. Secondary gauges shall be
recertified at a regular basis, but at least once a year.
A6.3.2 Procedure:
A6.3.2.1 Set up the test manifold such as that shown in Fig.
A6.2. Ensure that the test manifold is leak-free and can be
maintained at a steady pressure at the required level. A suitable
leak test is to pump down to a pressure below 0.1 kPa and
isolate the pump. Observe the pressure inside the unit for at
least 1 min. If the pressure rises no more than 0.01 kPa in that
period, the apparatus is considered acceptable.
A6.3.2.2 Connect the reference (primary) vacuum gauge(s)
and the gauge(s) to be calibrated to the manifold. The gauges
shall have such a range that the desired calibration pressure
falls between 10 and 90 % of the scale. Insert a dry ice trap
between the manifold and the vacuum pump. Adjust the
pressure to the required level for the test and run a final leak
test as above.
A6.3.2.3 After steady conditions have been maintained for
at least 3 min, readings are made of all gauges and compared
FIG. A6.3 Vacuum Manometer
with the reference gauge.
A6.3.2.4 Repeat the above procedure at the other required
A6.3.1.5 Choose a McLeod gauge with a range such that the pressure levels.
desired calibration pressure falls between 10 and 90 % of the A6.3.2.5 Make up a chart of corrections to be added at each
scale. Before refilling with clean mercury, heat the empty pressure level for each gauge tested. This can be used for
reference McLeod gauge at 250°C for at least 30 min at a interpolation when necessary.

A7. TEST METHOD FOR VERIFICATION OF REFLUX DIVIDING VALVES

A7.1 Scope A7.4 Apparatus

A7.1.1 This test method is for determining whether a liquid A7.4.1 The apparatus consists of the column and reflux
reflux dividing valve produces the prescribed reflux ratio. divider with its controller and the condenser if necessary for
convenient entry of liquid to the divider.
A7.2 Summary of Test Method A7.4.2 A means must be provided for introducing the test
liquid at steady rates over a range that includes 10 to 90 % of
A7.2.1 A hydrocarbon distillate of medium density is intro-
the maximum capacity of the column under test. This apparatus
duced to the reflux dividing valve while operating in the
can be a pumping system, but a simple gravity flow system can
normal way. The reflux ratio is determined by the ratio of the
be used if its reservoir is of a capacity that will ensure
two streams so obtained. The test is conducted over a range of
relatively uniform rates for each test. A supply of 400-mL
rates normally encountered in use to ensure that performance is
beakers tared to the nearest gram and a stopwatch are also
acceptable at all levels.
required.
A7.3 Significance and Use A7.5 Reagents and Materials
A7.3.1 This test method is intended to ensure that the valve A7.5.1 The test liquid can be any hydrocarbon fraction in
in operation actually divides the reflux in the desired ratio. the kerosine to diesel oil range (from 150 to 300°C).

Copyright by ASTM Int'l (all rights reserved); 22

 D2892 – 11a
A7.6 Preparation of Apparatus A7.7.9 Repeat sequence A7.7.1-A7.7.7 for a valve opera-
A7.6.1 Assemble the column in the normal way with the tion in which the closed portion of the cycle is equal to two
valve in place. Attach the condenser in its normal position if times the open period (2:1 ratio).
this will facilitate the introduction of the test liquid to the A7.7.10 When the actual reflux ratio differs by more than
divider. Connect the control device to the liquid dividing valve. 10 % from the desired ratio, the valve is not acceptable and
must be corrected. Fig. A7.1 illustrates graphically the test
A7.6.2 Set an untared beaker of suitable size under the results of a typical valve of acceptable performance.
bottom of the column and another under the takeoff point.
A7.6.3 Mount the liquid flow system so that the necessary A7.8 Precision and Bias
range of flow rates can be provided. A7.8.1 No statement is made concerning either the preci-
sion or bias of Annex A7 for checking the location of the
A7.7 Procedure temperature sensor because the result is used to determine
whether there is conformance to the criteria stated in A7.7.10.
A7.7.1 Start the valve and set the control device in such a
way that the valve is open for not less than 4 s and not more
than 6 s, and closed for a period that is five times as long (5:1
ratio).
A7.7.2 Commence introduction of the test liquid to the
valve at a rate equal to about 10 % of the maximum boilup rate
for the column under test.
A7.7.3 After 2 min have elapsed, simultaneously replace the
beakers under the column and under the takeoff point with
tared beakers.
A7.7.4 A precise 5 min intervals, simultaneously replace
those beakers with another pair of tared beakers. After three
sets of tared beakers have been collected, return the untared
beakers and shut off the flow of liquid.
A7.7.5 Weigh each of the tared beakers and obtain the mass
of liquid in each to the nearest gram.
A7.7.6 Calculate the ratio of the mass of test liquid recov-
ered in the beaker at the bottom of the column to that at the top
of the column in each of the three 5-min tests.
A7.7.7 If the ratios so obtained are not consistent to within
65 %, repeat A7.7.1 through A7.7.6.
A7.7.8 Repeat the sequence A7.7.1-A7.7.7 at flow rates
equal to about 30, 60, and 90 % of the maximum boilup rates FIG. A7.1 Graphical Illustration of Actual Reflux Ratio of a Typical
for the column under test. Reflux Divider

A8. PRACTICE FOR CONVERSION OF OBSERVED VAPOR TEMPERATURE TO ATMOSPHERIC EQUIVALENT

TEMPERATURE (AET)

A8.1 Scope A8.3 Calculation

A8.1.1 This practice is for conversion of the actual distilla- A8.3.1 Convert observed vapor temperature to atmospheric
tion temperature obtained at sub-ambient pressure to AET equivalent temperature using Eq A8.1:
corresponding to the equivalent boiling point at atmospheric
748.1A
pressure, 101.3 kPa (760 mm Hg), by means of equations AET 5 2 273.1
@1/~T 1 273.1!# 1 0.3861A 2 0.00051606
derived by Maxwell and Bonnell.8 (A8.1)
A8.2 Significance and Use where:
A8.2.1 Final data on atmospheric equivalent temperatures AET = atmospheric equivalent temperature, °C, and
are to be obtained by computation. T = observed vapor temperature, °C.
A8.3.1.1 Calculate A using Eq A8.2 or Eq A8.3:
8
Maxwell and Bonnell, Industrial Engineering Chemistry , Vol 49, 1957, p. 5.143222 2 0.972546 log10 P
A5 2579.329 2 95.76 log10 P (A8.2)
1187.

Copyright by ASTM Int'l (all rights reserved); 23

 D2892 – 11a

where: where:
P = operating pressure, kPa, (operating pressure $0.266 B = mean average boiling point, °C, and
kPa), or D = relative density at 15.6/15.6C.
5.994295 2 0.972546 log10 P A8.3.3.1 By custom, either the mid vapor temperature of the
A 5 2663.129 2 95.76 log P (A8.3) fraction or the midpoint of a gas chromatographic distillation
10
of the fraction can be used for the mean average boiling point.
where: In either case the method must be specified.
P = operating pressure, mm Hg (operating pressure $2 mm A8.3.3.2 An estimate of the K-factor can be made using Fig.
Hg). A8.1.
A8.3.2 The equations are correct only for fractions that have A8.3.4 Calculate the correction to be applied to the AET
a Watson K-factor of 12.0 6 0.2. The K-factor shall be using Eq A8.5:
assumed to be 12 and any effect of K-factor ignored unless
there is mutual agreement to the contrary. F S DG Pa
t 5 21.4[K 2 12] log10 P
o
(A8.5)

A8.3.3 If correction is required, calculate the K-factor using

where:
Eq A8.4: t = correction, °C
3
= 1.8~B 1 273.1! Pa = atmospheric pressure, kPa (mm Hg), and
K5 D (A8.4) Po = observed pressure, kPa (mm Hg).
A8.3.4.1 An estimate of the correction can be made using
Fig. A8.2.

FIG. A8.1 Watson Characterization Factor of Petroleum Fractions

Copyright by ASTM Int'l (all rights reserved); 24

 D2892 – 11a

FIG. A8.2 Boiling Point Corrections for K-Factor

APPENDIXES

(Nonmandatory Information)

X1. PRACTICE FOR DEHYDRATION OF A SAMPLE OF WET CRUDE OIL

X1.1 Scope X1.4 Apparatus

X1.1.1 This practice is for dehydrating a sample of wet X1.4.1 The dehydration of a sample of wet crude oil
crude oil prior to fractional distillation. requires apparatus, such as that shown in Fig. 1, composed of:
X1.2 Summary of Practice X1.4.1.1 Distillation Flask, with two side arms. In place of
the differential pressure manometer in the second sidearm, a
X1.2.1 A sufficient quantity of the sample is distilled in a
low efficiency column under atmospheric pressure at zero capillary is fitted for the passage of nitrogen into the liquid.
reflux ratio (total takeoff) to 130°C, the water decanted, and dry When a sample is suspected of containing emulsified water or
components recombined. significant amounts of clay or sediment, or both, additional risk
to glass apparatus is involved. In this case, remove gross water
X1.3 Significance and Use and sediment and use a metal flask for dewatering.
X1.3.1 Dehydration is important in order to achieve accu- X1.4.1.2 Distillation Column, shall be of the type described
rate yields in the light naphtha region. in 6.1.3.

Copyright by ASTM Int'l (all rights reserved); 25

 D2892 – 11a
X1.4.1.3 The rest of the apparatus is identical to that X1.6.5 To separate the water from each distillate fraction,
described in 6.1. cool to − 5°C and decant the hydrocarbon liquid. Weigh the
water.
X1.5 Preparation of Apparatus X1.6.6 Remove the condenser and rinse it with alcohol or
X1.5.1 Clean the distillation column and all the glassware acetone to remove adhering drops of water. Dry with air and
before starting the test. replace it.
X1.6.7 Recombine the cooled decanted fractions with the
X1.6 Procedure distillation residue observing the usual precautions against
X1.6.1 Cool the charge to a temperature not lower than 0°C. losses. Do not recombine the trap fraction.
Decant any bulk water that may be present. Weigh by differ- X1.6.8 Record the quantity of dry oil recovered. If reblend-
ence to the nearest gram, into a chilled distillation flask ing of the dried fractions was done in the original flask, this
containing some pieces of glass or porcelain, a given volume of flask can be used for subsequent distillation.
wet crude oil.
X1.6.2 Attach the flask to the column and pass a slow X1.7 Calculation
stream of nitrogen (8 cm3/s) through the capillary. Vent the X1.7.1 Calculate the mass % of water using Eq X1.1.
condenser through two traps in series maintained at the
W 5 100~A/B! (X1.1)
temperature of dry ice. Circulate coolant at a temperature
of − 20°C in the condensers. where:
X1.6.3 Apply heat to the flask, regulating it to attain a A = mass of water recovered, g,
moderate reflux rate as specified in 10.2.4 to 10.2.8 for B = mass of charge, g,
debutanization. Remove distillate slowly at total takeoff (reflux W = mass % of water, and
ratio = 0) until a vapor temperature of 130°C is reached. 100 = percentage constant.
Collect the fraction distilling below 65°C in a receiver cooled
to − 20°C or lower. X1.8 Precision and Bias
X1.6.4 Shut off the reflux valve and the heating system. X1.8.1 No statement is made concerning either the preci-
Cool the flask and contents to ambient temperature. Maintain sion or bias of Appendix X1 for mass % water because the test
the traps at the temperature of dry ice. Weigh the distillate method is used primarily for sample preparation for Test
fractions obtained. Method D2892.

X2. PRACTICE FOR PERFORMANCE CHECK

X2.1 Scope scribed and is recommended. For convenience sake, a simple

X2.1.1 This practice covers a procedure for calculating graphical solution based on the same method is also included.
column performance from GC boiling point distributions on X2.1.4 Overall column performance is assessed in terms of
fractions and residues, obtained by distilling an average (30 to column efficiency (minimum tray number) and in terms of the
40 API-gravity) crude oil under actual Test Method D2892 differential between the nominal cut point (AET) and the
distillation conditions. calculated effective cut point (ECP). Criteria are given for
NOTE X2.1—There are no theoretical reasons to limit the API-gravity acceptance of column performance. Possible corrective ac-
range from 30 to 40. However, the use of a crude oil in the quoted range tion(s), if required, are also indicated.
will, more or less, ensure that sufficient product is available to assess
performance, both at the upper and lower end of the temperature scale. X2.2 Significance and Use
The use of heavier crudes may not yield sufficient quantities at the low
end, while the reverse is true for lighter crudes. X2.2.1 Good agreement in yield of fractions can be
X2.1.2 The assessment of column performance can be made achieved under a variety of conditions because efficiency has
at any cut point where samples of two adjacent fractions or a no measurable effect on fraction yields except for very low
residue can be analyzed by gas chromatography. Either fresh or efficiencies and at the beginning and end of a distillation.
stored samples can be analyzed, as long as they have been However, fractions produced at high efficiency will have a
protected from loss by evaporation. Recommendations are narrower boiling range and hence some different properties
given for the number and spacing of cut points to be analyzed than the same fraction made at low efficiency. To arrive at a
for performance. standard level of efficiency and standard fraction quality, an
X2.1.3 A precise mathematical method9 for the calculation overall check is recommended to assess the actual performance
of distillation efficiency of multi-component mixtures is de- of the system. If the results of this test (Appendix X2) are
unacceptable, then the basic performance determinations as
9
Butler and Pasternak, The Canadian Journal of Chemical Engineering, Vol outlined in Annex A1-Annex A5 inclusive and Annex A7
42, 1964, p. 47. should be considered to determine the cause. The results of the

Copyright by ASTM Int'l (all rights reserved); 26

 D2892 – 11a
overall performance check as described in this appendix provided that they are based on the same principles of boiling
provides clues for possible causes and corrective actions in point separation and calibration, that is, employ a methylsili-
such a case. cone column and retention times calibrated for boiling point by
paraffins with boiling points assigned as given in Test Method
X2.3 Summary of Practice D2887.
X2.3.1 The distribution of (pseudo) components between
NOTE X2.4—A high temperature SIMDIST method is still under
the overlapping tail and front of two adjacent fractions (or a development by ASTM D02.04.H and in its present draft is only
residue) is calculated from GC analysis. The linearized trans- applicable to petroleum distillates. However, the methodology is widely
form of the distribution coefficient is plotted against the boiling used in the industry to measure the boiling range distribution of residues
point of the (pseudo) component. By linear regression, the best as well. Hence, it is envisaged that at a later date residues will be included
possible straight line is obtained. The efficiency is calculated in the scope of this proposed test method.
from the slope of the line by means of a modified Fenske X2.4.2 Using the results of the GC analyses of two con-
equation. The ECP is defined as the temperature at which the secutive distillate fractions, formulate groups defined by
overlaps are equal on the basis of mass-percent of the charge, equally spaced temperature ranges (pseudo components). The
or in mathematical terms, the temperature at which the distri- groups shall be formed such that the temperature range of the
bution coefficient equals 0.5. This form of the efficiency first group and that of the last group bracket the overlapping
calculation is mathematically identical to the graphical proce- portions of the distillate fractions. For each group (pseudo
dure described by Butler and Pasternak. component) list the noncumulative yields (% M/M on whole
fraction, referred to as Yield o.f.) contained within that tem-
X2.4 Procedure
perature range in each fraction. The temperature range used
X2.4.1 Obtain samples of contiguous fractions or a residue should have a width not smaller than 5°C nor wider then 15°C.
for each level of pressure and analyze them according to an The GC slices of adjacent distillate cuts by Test Method D2892
appropriate boiling point distribution method. The fractions shall have the same start and end boiling point temperature. To
shall be wide enough so that the overlap by GC analysis does ensure sufficient accuracy, within the constraint of fraction
not extend beyond the fraction. Uniform 25°C wide fractions width, at least four non-outlying components should be iden-
are recommended, but wider fractions can be employed with- tifiable with a distribution factor $0.05 and #0.95 (see
out deterioration of the reliability of the assessment. The X2.5.1).
recommended number of fractions to be obtained and the X2.4.3 Calculate the yield of the (pseudo) components in %
spacing of the cut points is discussed in X2.6. The range of M/M on the original (Test Method D2892) charge mass:
application of the available boiling point distribution GC test
mass fraction, g
methods, together with the recommended application to cuts by Yield o.c. 5 Yield o. f. 3 mass charge, g (X2.1)
Test Method D2892 is listed in X2.4.1.1-X2.4.1.3. Although
adjacent fractions may be analyzed by different test methods, it X2.4.3.1 Assign a boiling point to each (pseudo) component
is highly recommended that adjacent fractions be analyzed by exactly halfway the start and end boiling point temperature of
the same test method. the (pseudo) component.
NOTE X2.2—Different GC boiling range test methods might show NOTE X2.5—If actual components as obtained by Test Method D5134
relative bias towards each other. If such a bias exists it will affect the slope are used in the calculations, assign a boiling point as obtained from
of the regression line and, therefore, could lead to misleading efficiency literature for that component.
results.
X2.4.1.1 Fractions–Cut Point # 150°C—Analyze by Test X2.5 Calculation
Method D3710. The final boiling point of the fraction >150°C X2.5.1 Calculate the distribution coefficient (D) for each
should not exceed 260°C. Alternatively Test Method D5134 (pseudo) component by means of the following equation:
can be used. In the latter case it can only be applied to cuts Y1i
#100°C. If Test Method D5134 is used, the distribution Di 5 (X2.2)
~Y1i 1 Y2i!
coefficients are calculated for well defined and separated
individual compounds. Any number of compounds can be where:
used, with a minimum of four. Aromatic compounds shall not Y1i = yield (% M/M on charge) of (pseudo) component i in
be used for efficiency calculations. Fraction 1, and
Y2i = yield (% M/M on charge) of (pseudo) component i in
NOTE X2.3—Due to azeotropic effects, aromatic compounds will not Fraction 2.
fall on the same straight line as aliphatic compounds with the same boiling
point.
X2.5.2 Calculate the coefficients of the following equation:
T 5 aX 1 b (X2.3)
X2.4.1.2 Fractions—Cut Point >100°C—Analyze by Test
Method D2887. from:
X2.4.1.3 Residues (FBP >538°C)—Presently there are no (Ti Xi 2 ~ ( Ti ( Xi/n!
standardized methods for the characterization of the boiling a5 (X2.4)
( Ti2 2 ~~ ( Ti!2/n!
point distribution of residues on the books. However, methods
for residue containing materials are under development (see and:
Note X2.3). For the time being, in-house methods can be used b 5 ( Xi/n 2 a~ ( Ti/n! (X2.5)

Copyright by ASTM Int'l (all rights reserved); 27

 D2892 – 11a

where: Nominal cut point (AET) = 300°C

n = number of (pseudo) components, and Actual cut temperature = 221.4°C (from Annex A8)
Ti = boiling point of (pseudo) component, °C. Yield % M/M on charge; Fraction 1 (<200°C) = 10.3
and: Yield % M/M on charge; Fraction 2 (>200°C) = 9.8

S
Di
Xi 5 log10 1 2D
i
D X2.5.6.1 Linear regression of Xi versus Ti (X2.5.2):
T = − 14.77X + 300.1 thus: ECP = 300.1°C
X2.5.3 By definition, the ECP is the temperature where Di = X2.5.6.2 Taking as a reference component a component
0.5, hence Xi = 0 thus ECP = b. with a boiling point T = 280°C ( Tb = 280°C) and substituting
X2.5.4 The efficiency is calculated from: this in the derived equation with the corresponding X = 1.3605.
X 3 ~Tt 1 273! X2.5.6.3 For the actual efficiency Tt = 300°C. For the
N5 (X2.6) minimum efficiency (corrected for pressure) Tt = 221.4°C.
4.6 3 ~Tc 2 Tb!
Substituting these numbers in Eq X2.5 will yield an actual
where: efficiency Nactual = 8.4 and a minimum efficiency of Nminimum =
N = efficiency, 7.3.
X =
S
Dx
D
log10 1 2 D of reference component,
x
X2.5.7 Graphical Solution:
X2.5.7.1 A graphical solution to Eq X2.2 is shown in Fig.
X2.1. On Chart A percent to bottoms is plotted against the
Tt = average tower temperature, °C,
Tc = cut point temperature, °C, and boiling point (BP) of the (pseudo) component. The best
Tb = boiling point of reference component, °C. possible straight line is drawn through the points and extended
X2.5.4.1 As a reference compound any (pseudo) component down through Chart B.
can be chosen in the boiling range overlap of the two adjacent X2.5.7.2 From Chart A, the boiling point corresponding
fractions, except a component whose boiling point (Tb) is with the intersection of the line with 50 % to bottoms line is
equivalent to the ECP. Subsequently, the corresponding X is read. This is the ECP.
calculated from Eq X2.1. X2.5.7.3 From the intersection of the line on Chart B with
X2.5.4.2 For cuts obtained under atmospheric conditions the curved line for temperature denoting the ECP, the efficiency
and to obtain the actual tray number use the following Nactual can be read from the vertical axis. For atmospheric
temperatures: distillations the minimum efficiency, Nminimum, is equivalent to
Tc = ECP = AET, °C, and the actual efficiency, Nactual. For distillations at subatmospheric
Tt = actual (nominal) cut point, °C. pressure, the minimum efficiency can be calculated from the
X2.5.4.3 For cuts obtained under reduced pressure and for following equation:
obtaining the column efficiency (Nminimum) for comparative Tt 1 273
Nminimum 5 Nactual 5 T 1 273 (X2.7)
purposes, the actual column temperature shall be used. In this n
case the tower temperature Tt shall be considered equivalent to
where:
the actual vapor temperature at the nominal cutpoint (AET)
Tt = actual cut point, °C, and
under the pressure condition prevailing at the nominal cut Tn = nominal cut point, °C.
point.
X2.5.7.4 Examples—The same basic data as for the ex-
X2.5.5 Example of efficiency calculation for cuts obtained
amples in X2.5.5 and X2.5.6 are used. Percent to bottoms is
under atmospheric conditions.
just another way of expressing the distribution coefficient and
Excerpt of Test Method D2892 Analysis:
Operating pressure = 101.3 kPa
is related to the distribution coefficient by:
Percent to bottoms = (1 − Di) 3 100
Nominal cut point (AET) = 200°C (1) For convenience sake the relevant values for the
Actual cut point = 200°C
graphical solution are given in Table X2.3.
Yield % M/M on charge; Fraction 1 (<200°C) = 8.5 X2.5.8 Outliers—Because of inaccuracies in the GC data,
Yield % M/M on charge; Fraction 2 (>200°C) = 9.0
especially at the beginning and end of the (GC) distillation
X2.5.5.1 Linear regression of Xi versus Ti (X2.5.2): curve, it is not unusual to observe outlying points for distribu-
T = −17.32X + 199.9 thus ECP = 199.9°C tion coefficients <0.05 and >0.95, therefore, it is recommended
X2.5.5.2 Taking as a reference component a component to inspect the linear plot of T versus X for outliers. The same
with a boiling point T = 180°C (Tb = 180°C) and substituting holds for the graphical solution, but there outliers will be
this in the derived equation the corresponding X = 1.1503. readily observed. If outliers occur they should be removed,
Furthermore, Tt = 200°C. Substituting these numbers in Eq because of their very pronounced effect on the slope of the line
X2.5 will yield an efficiency of N = 5.9; or Nactual= Nminimum= and hence the value for N. Another cause for significant scatter
5.9. in data points can be incorrect sample size (too small) with
X2.5.6 Example of efficiency calculation for cuts obtained respect to column diameter or dynamic hold-up, or both. Check
under subatmospheric conditions. Table 1 or Annex A2, or both, for correct sample size.
Excerpt of Test Method D2892 analysis: Inaccuracies in the GC data can be especially pronounced in
Operating pressure = 13.3 kPa cuts with very narrow boiling ranges. This is especially evident
in lighter boiling cuts (<100 °C). Appropriate attention to skew

Copyright by ASTM Int'l (all rights reserved); 28

 D2892 – 11a

FIG. X2.1 Efficiency Calculation

TABLE X2.1 Relevant Values for Calculation Method of Efficiency

NOTE—Excerpts from GC analysis (X2.4.2), calculation of distribution coefficient (X2.5.1) and Xi (X2.5.2) where Yield o.f. is % M/M of the
component on the fraction (GC result from analysis of the fraction).
Fraction 1 Fraction 2
Yield Yield Yield Yield
Component BP o.f. o.c. o.f. o.c.
Ti°C Y1i Y2i Di Xi
160-170 165 19.75 1.68 0.21 0.02 0.989 1.9485
170–180 175 19.40 1.65 0.62 0.06 0.967 1.4706
180–190 185 17.95 1.53 2.03 0.18 0.893 0.9217
190–200 195 13.04 1.11 6.39 0.58 0.658 0.2850
200–210 205 6.96 0.59 12.52 1.13 0.344 −0.2798
210–220 215 2.05 0.17 16.67 1.50 0.104 −0.9350
220–230 225 0.77 0.07 18.45 1.66 0.038 −1.4043
230–240 235 0.18 0.02 18.72 1.68 0.009 −2.0419

TABLE X2.2 Relevant Values for Calculation Method of Efficiency

NOTE—Excerpts from GC analysis (X2.4.2), calculation of distribution coefficient (X2.5.1), and Xi (X2.5.2) where Yield o.f. is % M/M of the
component on the fraction (GC result from analysis of the fraction).
Fraction 1 Fraction 2
Yield Yield Yield Yield
Component BP o.f. o.c. o.f. o.c.
Ti°C Y1i Y2i Di Xi
260–270 265 16.43 1.69 0.08 0.01 0.995 2.3342
270–280 275 16.26 1.67 0.27 0.03 0.984 1.8014
280–290 285 15.00 1.55 1.58 0.15 0.909 0.9991
290–300 295 11.07 1.14 5.92 0.58 0.663 0.2934
300–310 305 5.23 0.54 11.85 1.16 0.317 −0.3336
310–320 315 1.50 0.15 15.77 1.55 0.091 −1.0001
320–330 325 0.30 0.03 16.98 1.66 0.018 −1.7312
330–340 335 0.08 0.01 17.27 1.69 0.005 −2.3126

(column overloading) and linearity of signal (detector over- methods mentioned in X2.1.4. Ensure that correct units are
loading) is advisable if unusual GC results are detected. If compared (that is, obtain Test Method D3710 results in mass
using Test Method D3710, note that results are typically percent).
calculated in liquid volume %, not mass % as in the other X2.5.9 Final Calculation:

Copyright by ASTM Int'l (all rights reserved); 29

 D2892 – 11a
TABLE X2.3 Relevant Values for Graphical Solution X2.6.2.1 Only three pressure levels are permitted in Test
NOTE—From Chart A for Example 1, an ECP of 200°C is read and for Method D2892. Distillation at 0.266 kPa provides an alterna-
Example 2 an ECP of 300°C. From Chart B for Example 1, the efficiency tive to distillation under 1.33 kPa, but both pressure levels shall
Nactual = 5.9 and for Example 2 Nactual= 8.4. For Example 1, atmospheric not be applied consecutively in the same distillation run. The
distillation, Nactual = Nminimum = 5.9. For Example 2, subatmospheric reflux ratio applied shall be 5:1 in all cases, including distilla-
distillation, Nminimum is derived from Eq X2.7, and amounts to 7.2. tion at 0.266 kPa.
Example 1 Example 2 X2.6.3 Performance Verification—Determine the ECP and
BP Di % to bottom BP Di % to bottom
the efficiency Nminimum at three appropriate pressure levels, but
165 0.989 1.0 265 0.995 0.5 only at the highest cut point usually attained at that pressure
175 0.967 3.0 275 0.984 1.6
185 0.893 10.7 285 0.909 9.1 level.
195 0.658 34.2 295 0.663 33.7 X2.6.4 Data Interpretation–Effıciency:
205 0.344 65.6 305 0.317 68.3
215 0.104 89.6 315 0.091 90.9
X2.6.4.1 Regression of the efficiency (Nminimum) obtained
225 0.038 96.2 325 0.018 98.2 against the AET should indicate a best fit for linear regression
235 0.009 99.1 335 0.005 99.5 and produce a straight line more or less parallel to the lines
given in Fig. X2.1. The regression line shall not exceed the
upper or lower limits shown in Fig. X2.1. Individual points
X2.5.9.1 Calculate the efficiency Nminimum and ECP (°C) at should be randomly spread along the regression line and be
each point. Calculate the difference between ECP and AET for within 0.7 theoretical plates from the regression line. If one
each cut point. point is outside these limits, rerun the appropriate GCs, redo
X2.5.9.2 Table X2.4 defines the standard efficiency perfor- the efficiency calculation, and check again. If this check fails or
mance of a Test Method D2892 TBP (15/5) distillation, more then one point fails this criterion, the performance of the
including the standard spread for individual data points. Fig. column shall be considered suspect. Consult X2.6.4.2-X2.6.4.4
X2.2 is a graphical presentation of the standard efficiency and for possible causes. Take corrective action and repeat the whole
the acceptable upper and lower limits of Test Method D2892 procedure, including the distillation (see Note X2.6).
column efficiency performance (14 to 18 theoretical plates). A X2.6.4.2 If the regression line is curved, yielding relatively
discussion on the acceptability of deviations, probable causes high efficiency (Nminimum) values at the lowest pressure level, it
and corrective actions is given in the next chapter. is an indication for unacceptable heat loss from the column.
Check Annex A3. The reverse is true for an overhead column
X2.6 Frequency of Test and Data Interpretation (see 6.1.3.4).
X2.6.1 It is recommended to carry out a full performance X2.6.4.3 If the regression line is not parallel and yielding a
check at the first time of commissioning or any other point of steeper slope, it is an indication of unacceptable heat loss in the
time when there are significant changes to the equipment. After column (check Annex A3). The reverse is true for an over-
that, a shortened evaluation may be satisfactory, if carried out heated column (see 6.1.3.4). Nonparallel regression lines can
at regular intervals, for example, once or twice a year, to verify also be an indication of the application of a non-approved and
that the equipment is still performing satisfactory. non suitable packing (see 6.1.3 and Table 1).
X2.6.2 Full Performance Check—Determine the ECP and X2.6.4.4 If the (parallel) regression line is located outside
the efficiency Nminimum for at least two, but if possibly three, the upper or lower Nminimum limits, or both, the following
cut points at each of the three most frequently used pressure causes might be applicable.
levels. One of the cut points should be the maximum cut point (1) Efficiency of the packing too high or too low. Check
(AET) usually obtained at the appropriate pressure level and Table 1 or Annex A1, or both. If appropriate, correct the
one of the cut points should be obtained at the lower end of the efficiency by adding or subtracting packing material from the
temperature range usually covered at the appropriate pressure column.
level. Table X2.5 gives an example of such a scheme. (2) Reflux ratio too high or too low. Check Annex A7.
(3) Incorrect distillation rate. Consult 10.3.2 and 10.4.5.
X2.6.4.5 For a (shortened) periodic verification of effi-
TABLE X2.4 Standard Efficiency for 15/5, Test Method D2892
Columns ciency, the three values for efficiency (Nminimum) obtained shall
be located inside the band for the regression line obtained from
NOTE—The data from Table X2.4 were estimated from the efficiency the full performance check and more or less parallel to that
results obtained as a spin off of a round robin carried out in 1978 on a
regression line. If one or both criteria are not met, it is an
crude oil with an API-gravity of 35. However, the lack of detailed
documentation of these results prevents rigid statistical treatment. indication that the performance of the column has changed. For
possible causes refer to X2.6.4.1-X2.6.4.4.
Standard Efficiency,
Cut Point, °C
Nminimum NOTE X2.6—It is recognized that the setting of acceptable tolerance
50 4.1 6 0.5 limits to trigger corrective action is the responsibility of the laboratory.
100 4.7 6 0.6 However, it is recommended that they do not exceed the limits as
150 5.3 6 0.6 indicated.
200 5.9 6 0.7
250 6.6 6 0.8 X2.6.5 Data Interpretation–Cut Point:
300 7.2 6 0.9
350 7.8 6 0.9 X2.6.5.1 When using an average crude oil (30 to 40 API)
distilled under TBP (15/5) conditions, it is expected that the

Copyright by ASTM Int'l (all rights reserved); 30

 D2892 – 11a

FIG. X2.2 Efficiency Nomograph

TABLE X2.5 Performance Check; Cut Point Scheme (2) If the difference between ECP and AET is significantly
Pressure, Range for Recommended Cut Point, °C larger at one or more reduced pressure level(s), it is an
kPa indication for incorrect pressure measurement. Check Annex
Low Middle High
101.3 60 − 70 130 − 140 190 − 200 A6.
13.3 230 −240 260 − 270 290 − 300 (3) Incorrect efficiency (too low). Refer to X2.6.4.
1.33 330 − 340 − 360 − 370 (4) Incorrect sample size (too small) with respect to column
0.266 330 − 340 360 − 370 390 − 400
diameter or dynamic hold-up, or both. Check Table 1 or Annex
A2, or both.
(5) Excessive heat loss. Evidence for the latter can also be
obtained by careful examination of the T versus X plot
difference between ECP and AET will not exceed 0.7 times the
(X2.5.8). In case of excessive heat loss the distribution will be
reproducibility of the test method. The reproducibility of the
asymmetrical, evidenced by significantly more outlying points
test method in degrees Celsius is tentatively estimated for
at the high end of the temperature scale than at the low end. An
average crude oils to be approximately 8°C at any pressure
overheated mantle (6.1.3.4) will lead to a too low ECP in
level. Therefore, deviations in excess of 6°C shall be consid-
comparison to the targeted AET.
ered suspect (see Note X2.7).
X2.6.5.2 The following causes for excess deviations of ECP NOTE X2.7—The setting of acceptable tolerance limits to trigger
corrective action is the responsibility of the laboratory. However, the limit
from AET can be identified.
given has been proven indicative of incorrect performance in many years
(1) If the ECP is consistently and significantly higher or of application of Test Method D2892.
lower than the AET, it is an indication of incorrect temperature NOTE X2.8—Too high efficiency will have no measurable effect on the
measurement, Check Annex A4-Annex A6. ECP-AET differential.

SUMMARY OF CHANGES

Subcommittee D02.08 has identified the location of selected changes to this standard since the last issue
(D2892–11) that may impact the use of this standard. (Approved Dec. 1, 2011.)

(1) Revised 3.1.7 to be consistent with one definition and

Form and Style.

Subcommittee D02.08 has identified the location of selected changes to this standard since the last issue
(D2892–10) that may impact the use of this standard. (Approved July 1, 2011.)

(1) Updated terms in Section 3. (2) Made some of the original discussions into notes.

Copyright by ASTM Int'l (all rights reserved); 31

 D2892 – 11a

Subcommittee D02.08 has identified the location of selected changes to this standard since the last issue
(D2892–05) that may impact the use of this standard. (Approved Oct. 1, 2010.)

(1) Revised Section 13 to reflect the actual precision state- the research report containing the data and statistical evaluation
ment from the D2892 precision study, including reference to and validation.

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
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 and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or service@astm.org (e-mail); or through the ASTM website
(www.astm.org). Permission rights to photocopy the standard may also be secured from the ASTM website (www.astm.org/
COPYRIGHT/).

Copyright by ASTM Int'l (all rights reserved); 32