Aga 8 Part 1
Aga 8 Part 1
8
Part 1
Thermodynamic Properties of
Natural Gas and Related Gases
DETAIL and GROSS Equations of State
Third Edition
April 2017
(A revision of AGA Report No. 8, 2nd edition, 1994)
Prepared by
Transmission Measurement Committee
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Thermodynamic Properties of
Natural Gas and Related Gases
DETAIL and GROSS Equations of State
Prepared by
Third Edition
April 2017
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
The American Gas Association’s (AGA) Operations and Engineering Section provides a forum for industry
experts to bring their collective knowledge together to improve the state of the art in the areas of operating,
engineering and technological aspects of producing, gathering, transporting, storing, distributing, measuring and
utilizing natural gas.
Through its publications, of which this is one, AGA provides for the exchange of information within the natural
gas industry and scientific, trade and governmental organizations. Many AGA publications are prepared or
sponsored by an AGA Operations and Engineering Section technical committee. While AGA may administer the
process, neither AGA nor the technical committee independently tests, evaluates or verifies the accuracy of any
information or the soundness of any judgments contained therein.
AGA disclaims liability for any personal injury, property or other damages of any nature whatsoever, whether
special, indirect, consequential or compensatory, directly or indirectly resulting from the publication, use of or
reliance on AGA publications. AGA makes no guaranty or warranty as to the accuracy and completeness of any
information published therein. The information contained therein is provided on an “as is” basis and AGA makes
no representations or warranties including any expressed or implied warranty of merchantability or fitness for a
particular purpose. Nothing contained in this document should be viewed as an endorsement or disapproval of any
particular manufacturer or product.
In issuing and making this document available, AGA is not undertaking to render professional or other services
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
for or on behalf of any person or entity. Nor is AGA undertaking to perform any duty owed by any person or
entity to someone else. Anyone using this document should rely on his or her own independent judgment or, as
appropriate, seek the advice of a competent professional in determining the exercise of reasonable care in any
given circumstances.
AGA has no power, nor does it undertake, to police or enforce compliance with the contents of this document.
Nor does AGA list, certify, test or inspect products, designs or installations for compliance with this document.
Any certification or other statement of compliance is solely the responsibility of the certifier or maker of the
statement.
AGA does not take any position with respect to the validity of any patent rights asserted in connection with any
items that are mentioned in or are the subject of AGA publications, and AGA disclaims liability for the
infringement of any patent resulting from the use of or reliance on its publications. Users of these publications are
expressly advised that determination of the validity of any such patent rights, and the risk of infringement of such
rights, is entirely their own responsibility.
Users of this publication should consult applicable federal, state and local laws and regulations. AGA does not,
through its publications, intend to urge action that is not in compliance with applicable laws, and its publications
may not be construed as doing so. Information concerning safety risks, proper installation or use, performance or
fitness or suitability for any purpose with respect to particular products or materials should be obtained from the
manufacturer or supplier of the material used.
Changes to this document may become necessary from time to time. If changes are believed appropriate by any
person or entity, such suggested changes should be communicated to AGA in writing and sent to: Operations &
Engineering Section, American Gas Association, 400 North Capitol Street, NW, 4th Floor, Washington, DC
20001, U.S.A. or E-mail to: [email protected]. Suggested changes must include: contact information,
including name, address and any corporate affiliation; full name of the document; suggested revisions to the
text of the document; the rationale for the suggested revisions; and permission to use the suggested revisions
in an amended publication of the document.
AGA Report No. 8, Part 1, is designated as the third edition of the revised 1994 second edition of AGA
Report No. 8. As in the second edition, the third edition provides the technical information necessary to compute
thermodynamic properties including compressibility factors and densities of natural gas and related gases for
states in the gas phase only. Additionally, equations for calculating speeds of sound and other thermodynamic
properties are given. Historical information on this document is given in Section 1.2, Background. Analyses of
the calculation of uncertainties are provided for various gas temperatures, pressures, and compositions. It is based
on research managed and sponsored by the then Gas Research Institute (GRI) in cooperation with AGA and the
Groupe Europeen de Recherches Gazieres (GERG). Because the equations of state contained within International
Standards Organization documents ISO 12213: Natural Gas – Calculation of Compression Factor, Part 2:
Calculation Using Molar-Composition Analysis, 2006 edition, and ISO 20765: Natural Gas – Calculation of
Thermodynamic Properties, Part 1: Gas Phase Properties for Transmission and Distribution Applications, 2005
edition, are based on the same equations in this revision, calculated properties should have the same values.
This revised edition, now designated as AGA Report No. 8, Part 1, uses the same DETAIL and GROSS
equations of state as in the 1994 edition of AGA Report No. 8. However, the temperature, pressure, and gas
composition limits have been modified in this edition; and therefore, it will be necessary to ensure that the
temperature, pressure, and gas composition fall within the new limits for the desired uncertainty. The users are
advised to verify the applicability/acceptability of the program for the new limits based on the 1994 edition of
AGA 8.
The documentation of programs for calculating properties from the methods described in this document
is available as supplementary material in Appendix C. Examples are available in Fortran, VB, and C++ code.
The supplementary material also contains a Microsoft Excel spreadsheet for property calculations. This can be
used to determine if, for a particular temperature, pressure, and composition, the property values calculated from
the equations in Part 1 are within the desired uncertainty (by comparing with those in Part 2) even though one or
more of these inputs may be outside the ranges given in Part 1. A file containing calculated points at different
compositions (not necessarily related to typical natural gas) is included, which can be used to verify that programs
or equipment have been implemented or upgraded correctly to produce values that are in agreement with the
equations in this document.
The user is encouraged to compare property values obtained through the use of the software
provided in this edition with those obtained from the user’s existing software and determine if the
differences are within the uncertainty limits of their respective calculation methods. The Excel spreadsheet
provided in the supplementary material can be used to determine if property values are within the
acceptable uncertainty limits even though the state point may be outside the ranges given in this revision.
If the property values are not within the uncertainty limits, the user should consider implementing AGA
Report No. 8, Part 2, noted below.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
AGA Report No. 8, Part 2, published separately, is based on the GERG-2008 equation of state, and is
applicable for temperature, pressure, and composition ranges outside those in Part 1, with uncertainties in density,
compressibility factor, and speed of sound still within 0.1 %. Part 2 can also be used for operating conditions and
compositions that are applicable for Part 1, with differences in the calculated values within the uncertainty limits
of each equation, where uncertainties are given with a 95 % level of confidence (k=2). While adoption of the
equations in Part 2 is encouraged, the decisions to upgrade existing installations to use AGA Report No. 8,
Part 2, are left to the discretion of the parties involved.
Some material described in Part 1 also applies to Part 2, and vice versa, and is not repeated in both parts.
For example, Part 1 describes an algorithm for obtaining densities through an iterative procedure that can also be
used with the equations in Part 2 (and which is applied for this purpose to the GROSS, DETAIL, and GERG-
2008 equations of state in the programs in the supplementary material). Similarly, Part 2 outlines the method for
reporting calculated results and uncertainties from the equations in both parts, and also describes the experimental
The publication AGA Report No. 10: Speed of Sound in Natural Gas and Other Related Hydrocarbon
Gases has been discontinued with the release of this AGA Report No. 8, Part 1. The calculation of the speed of
sound no longer requires integration of the compressibility factor equation, as was done in AGA Report No. 10,
but is implemented through differentiation of the fundamental Helmholtz energy equation of state. This results
in faster calculations through direct computation of the required derivatives.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
AGA acknowledges the contributions of the above individuals and thanks them for their time and effort in
reviewing the document or suggesting changes to the programs in the supplementary material.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Part 1 presents information for the computation of thermodynamic properties (e.g., compressibility factor,
density, and speed of sound) of natural gas and related gases with the DETAIL and GROSS equations of state.
Uncertainty estimations for different compositions, pressures, and temperatures are given. The computations are
valid for single-phase gaseous states only. For mixed or liquid phases, refer to Part 2.
1.2 Background
Research in 1928 and 1929 under the direction of Mr. Howard S. Bean of the National Bureau of Standards
provided the natural gas industry with its initial compressibility factor data covering pressures up to 600 psia
(4 MPa). However, it was not until 1954 that extensive tables of natural gas supercompressibility factors were
published, based on tests supervised by Professor Samuel R. Beitler of Ohio State University. The natural gas
supercompressibility factor tables were extended and an equation of state was developed in 1956-1962 under the
direction of Mr. R. H. Zimmerman of Ohio State University. The results of this project, designated PAR Project
NX-19, appear in AGA’s Manual for Determination of Supercompressibility Factors for Natural Gas, published
in 1962.
The research leading to the 1985 and 1992 reports was initiated in 1981 under the sponsorship of GRI in
close liaison with the AGA Transmission Measurement Committee. This research, carried out under the direction
of Professor Kenneth E. Starling of the University of Oklahoma, was aimed at extending capabilities for accurate
computation of compressibility factors beyond the temperature, pressure, and composition ranges of PAR Project
NX-19. The results for pipeline-quality natural gas, which were completed in 1984, provided the basis for the 1985
report.
The initial 1981-1984 research used data ranging in pressures up to approximately 900 psia (6 MPa)
obtained from the literature and provided by GERG. However, the GERG data bank was extended considerably
over the period 1985-1990. The new data showed that the original equation of state, developed in the period 1981-
1984, needed to be improved. In addition, speed of sound data obtained under GRI sponsorship during 1985-1989
showed calculations for rich gases were not sufficiently accurate for critical flow applications.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Although the 1992 and 1994 AGA 8 publications of the equation of state are explicit in the compressibility
factor Z, the formulation was originally developed with an equation of state explicit in the Helmholtz energy a with
independent variables of molar density d and temperature T. The Helmholtz energy equation was not given in the
1992 and 1994 AGA 8 documents, where only equations explicit in pressure and compressibility factor were
reported. The Helmholtz energy is a fundamental thermodynamic property from which all other thermodynamic
properties can be calculated as derivatives with respect to density or temperature. For example, the expression for
calculating pressure is
a
P d 2 . (1-1)
d T
When starting from the pressure explicit equation, the calculation of other properties, including the speed
of sound, requires not only derivatives of the compressibility factor with respect to density and temperature, but
also integration of Z over density, which can be quite complicated. Equations are presented in this revision that
start with the Helmholtz energy form, and allow access to other thermodynamic properties such as heat capacities,
enthalpies, and entropies by differentiation of this fundamental property. Formulations explicit in the Helmholtz
energy have been used heavily over the last two decades to represent the properties of pure fluids and mixtures due
to the ease of calculating all other thermodynamic properties.
Because the calculation of dew points requires knowledge of the liquid phase as well as the gas phase, the
DETAIL equation of state presented in Part 1 is not able to determine the state where liquids will condense from a
natural gas stream. With the addition of accurate liquid calculations in the GERG-2008 equation, dew points, as
well as bubble points and saturation states in the retrograde region can now be evaluated for any temperature or
pressure input state. The equations required to do so are covered in Part 2.
This report provides recommended equations of state to compute accurate compressibility factors,
densities, and other thermodynamic properties for natural gas. The information in this report can be used for
calculations of compressibility factors and densities for pure methane, ethane, nitrogen, carbon dioxide, hydrogen,
and hydrogen sulfide, and for gas mixtures of up to twenty-one components. Compressibility factors,
supercompressibility factors, and density values can be applied directly in calculations of gas volume and gas flow
rate. These computations can also be used in other instances where the relationship between the temperature,
pressure, and volume of a gas is important.
Other thermodynamic properties that can be calculated with the information in this report include the heat
capacity, enthalpy, entropy, sonic velocity, critical flow factor (see Appendix E of ASME/ANSI MFC-7M-1987),
and component chemical potentials. Applications that use these properties include calculations for sonic nozzles,
compressors, heat exchangers, gas mixture reaction equilibrium, and gas mixture component fugacities (for use in
vapor-liquid equilibrium calculations).
Two equations of state are provided in Part 1. All information needed to efficiently implement the
appropriate equation of state as a function of the natural gas composition (i.e., a gas analysis) is contained in this
report, except the molar heating values, which can be obtained from AGA Report No. 5, Natural Gas Energy
Measurement. The two equations differ by their complexity and the range over which calculations can be made.
The equation of state with a wider range of conditions is referred to as the "DETAIL Equation of State." See
References [1] and [2]. The equation of state with a limited range of conditions, herein referred to as the "GROSS
Equation of State," uses second and third virial coefficients to compute the compressibility factor. See References
[3] and [4]. Although gas properties are best calculated through the use of the natural gas composition, they can
also be obtained with an aggregate or gross knowledge of the gas (given by heating value and/or relative density,
and diluent gas content information).
The DETAIL equation of state was developed to accurately describe the gas-phase pressure-temperature-
density behavior of natural gas mixtures. This equation of state reduces the uncertainties of compressibility factor
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
and density calculations for natural gas, which can contain mole percentages of hexanes plus heavier hydrocarbons
greater than 1 %. Equations of the density behavior of pure hydrogen sulfide and binary mixtures of hydrogen
sulfide with methane, ethane, nitrogen, and carbon dioxide were developed to reduce the calculation uncertainty
for natural gas containing hydrogen sulfide (sour gas). Second virial equations were developed for water and binary
mixtures of water with methane, ethane, nitrogen, and carbon dioxide to reduce the calculation uncertainty for
natural gas containing water vapor. Section 4 presents complete information for the DETAIL equation of state.
Temperature, °C
–130 –60 –4
40,000 280
Region 4 – 1.0 %
10,000 70
psia
2,500 17.2
Region 2 – 0.3 %
1,500 10.3
Region 1 – 0.1 %
See Table 1 Range A for composition ranges.
0 0
–200 –80 25
Temperature, °F
Figure 1. Overview of uncertainties for natural gas compressibility factors with the DETAIL equation of state
(not to scale). The upper temperature limit is the onset of decomposition of the components in the gas.
Note: This figure is not intended for contractual use. Table 1 should be consulted for specific uncertainty
information.
Figure 1 gives a simple overview of typical uncertainties for the DETAIL equation of state, but should not
be used as a basis to estimate the uncertainties in calculated properties. Rather, Table 1 lists comprehensive ranges
for which the uncertainties are less than 0.1 % (details are given in Reference [5]). Three different sets of ranges
are defined below, each having its own range of validity for temperature, pressure, and component fractions.
Range A in Table 1 shows the limits of the mole percent of components in a natural gas mixture in which
the uncertainties in compressibility factors are less than 0.1 % for temperatures above 25 °F (–4 °C), pressures less
than 1500 psia (10.3 MPa), heating values from 630 to 1200 Btu/scf, and relative densities from 0.554 to 0.91.
However, a mixture with compositions near the upper limits of every component in this range, except methane and
nitrogen, will have an uncertainty higher than 0.1 %. Most typical natural gases do not fall under such a condition.
This was verified through the use of the 200 gas compositions available in the supplementary material. Deviations
between the DETAIL and GERG equations of state were compared for each gas composition – those that met the
0.1 % uncertainty limit fell within the ranges listed in Range A of Table 1, or were slightly outside this range at
the lower temperature limit with pressures slightly below the upper pressure limit. Several of the limits listed in
this range have been modified from those given in Reference [5] to give better results. The regions in Figure 1
with uncertainty limits of 0.3 %, 0.5 %, and 1.0 % are also based on the Range A compositions. However, it is
important to make use of a phase diagram to ensure that states are not in the 2-phase or in the critical region. The
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
DETAIL equation of state should not be used for states at pressures exceeding the dew or bubble point (including
2-phase states). This applies for Ranges B and C as well.
Range B in Table 1 shows the maximum allowable mole percent composition for temperatures above
25 °F (–4 °C), but with pressures less than 300 psia (2 MPa), heating values from 680 to 1500 Btu/scf, and relative
densities from 0.47 to 0.91. Range B allows higher upper range values of the compositions for operations below
300 psia.
Range C in Table 1 has a composition range that is much smaller than the other ranges, but allows
temperatures down to 17 °F (–8 °C) and pressures up to 3000 psia (21 MPa).
In this AGA 8, Part 1, document the upper pressure limit for 0.1 % uncertainty has been changed from
1750 psia to 1500 psia (12 MPa to 10.3 MPa), and the lower temperature limit has been changed from 17 °F
(–8 °C) to 25 °F (–4 °C) based on work presented in Reference [5]. However, uncertainties in the DETAIL equation
of state are often below 0.1 % far outside the ranges listed in Table 1, especially at temperatures above 40 °F. For
continued use of the DETAIL equation of state outside the ranges listed in Table 1, comparisons can be made
between this and the GERG-2008 equation of state. When the difference is less than 0.1 %, the DETAIL equation
of state can be used with a stated uncertainty of 0.1 %. The Excel application given in the supplementary material
will calculate the difference between the DETAIL and GERG-2008 equations for any gaseous state. Other tools
can also be used (e.g., Reference [6]) to ascertain the differences in the calculations between the two equations.
When exceeding the limits of Table 1 but claiming the 0.1 % uncertainty, users should specify the method or
software used to arrive at this claim.
The GROSS equation of state was developed for dry, sweet natural gas to calculate compressibility factors.
The limits for this method are listed in Table 2. For some gases at the upper limits of the allowed compositions,
with temperatures below 80 °F (27 °C) or at pressures above 800 psia (5.5 MPa), the uncertainty of the GROSS
equation of state can exceed 0.1 %. The maximum composition limits were therefore significantly reduced from
the 1994 edition of AGA 8 so as to meet the 0.1 % uncertainty level. When exceeding the bounds listed in
Table 2, the procedure given at the end of Section 1.3.1 for validating the uncertainty of a particular state should
be followed.
Section 5 and Appendix A present information for this equation of state. Optimally, natural gas property
calculations from the GROSS equation require the gas composition, i.e., the mole fractions of the components in
the mixture. When this information is not available, methods given in Appendix A can be used as a substitute to
obtain the molar ideal gross heating value HCH of the mixture of hydrocarbon components present in the natural
gas along with the compositions of nitrogen and carbon dioxide in the mixture.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Table 2 – Ranges for Temperature, Pressure, Heating Value, and Relative Density with
Uncertainties Less than 0.1 % in Compressibility Factors of Natural Gas Calculated with the
GROSS Equation of State
Quantity Range 1 Range 2
Temperature 25 °F to 143 °F (–4 °C to 62 °C) 17 °F to 143 °F (–8 °C to 62 °C)
Maximum pressure 1500 psia (10.3 MPa) 600 psia (4.1 MPa)
Relative density 0.554 to 0.63 0.554 to 0.89
Gross heating value† 930 to 1040 Btu/scf 665 to 1100 Btu/scf
34.7 to 38.7 MJ/m3 24.8 to 41 MJ/m3
Upper composition limits (mole percent)
Nitrogen 7 20
Carbon dioxide 3 25
Ethane 2 8
Propane 0.5 4
Total butanes 0.3 0.5
Total pentanes 0.2 0.3
Hexanes plus 0.04 0.08
Hydrogen‡ 0.2 2
Oxygen‡ 0.2 0.5
Carbon monoxide‡ 0.4 1
Water‡ 0.03 0.2
Hydrogen sulfide‡ 0.08 0.2
Helium‡ 0.05 0.3
Argon‡ 0.1 0.2
†Values are based on a methane lower composition limit of 60 mole percent. Reference conditions in U.S. customary units
are (60 °F, 14.73 psia; density at 60 °F, 14.73 psia) and in SI units are (15 °C, 0.101325 MPa; density at 0 °C,
0.101325 MPa).
‡
The nonhydrocarbon species listed in this table (except nitrogen and CO2) are considered here as impurities for use in the
GROSS equation of state, and should be added to the equivalent hydrocarbon mole fraction as explained in Section 5.2. The
SGERG-88 equation contains additional parameters for hydrogen and carbon monoxide, and Reference [3], ISO 12213:3, or
GERG TM 5 (1991) should be consulted when the mixture contains these fluids.
The GRI and GERG compressibility factor reference databases and comparisons with the GERG-2008
equation of state have verified the expected uncertainties given in Table 1 and within Region 1 and parts of Regions
2, 3, and 4 of Figure 1. The boundaries in Table 1 can often be exceeded at lower temperatures or higher pressures
than those given. The last paragraph of Section 1.3.1 explains the procedure that can be used to determine if a
particular gas or operating condition still resides within the 0.1 % uncertainty limit.
In general, the expected uncertainty in density of the GROSS equation of state is 0.1 % for natural gas
having compositions and operating conditions identified in Table 2. The GRI and GERG compressibility factor
reference databases and comparisons with the GERG-2008 equation of state have verified the expected
uncertainties given in Table 2. The equation was not designed for and should not be used outside of these limits,
unless comparisons as described in the last paragraph of Section 1.3.1 have been made.
1.5 Recommendations
Two different equations for the compressibility factor, the DETAIL equation of state and the GROSS
equation of state, are provided in Part 1. The choice of a particular equation of state depends on the natural gas
composition and operating conditions, and the uncertainty expectation. See Part 2 for liquid phase and vapor-
liquid equilibrium phases, or for applications outside the uncertainty ranges given in Figure 1 and Tables 1 and 2.
The GERG-2008 equation of state given in Part 2 is the most accurate in the calculation of compressibility
factors and densities, but also has the most complex form. The DETAIL equation of state is more accurate than
the GROSS equation. The GROSS equation of state is only recommended when simplicity is needed for
calculations of natural gas compressibility factors and densities, provided the natural gas compositions and
operating conditions are within the ranges given in Table 2.
Use of the GERG-2008 equation of state may be considered due to its higher accuracy and for maintaining
consistency for operating conditions that may lie either within or outside the ranges in Table 1. Switching between
the GERG-2008 and DETAIL equations based on the operating state will lead to discontinuities in calculated
values at the point of overlap. For conditions that never exceed the ranges in Table 1, the DETAIL equation of
state may be the most practical for certain applications based on the operator’s needs.
7
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
This AGA 8, Part 1, puts forth a new approach to define the gas phase, dense phase, and liquid phase of a
mixture. Figure 2 shows the phase boundary of a natural gas mixture with about 83 % methane. The heaviest
hydrocarbons are heptane with a mole percent composition of 0.027 %, octane with 0.017 %, and nonane with
0.0009 %, where these three components play the largest role in the upper temperature and pressure limits of the
saturation boundaries.
Although a fluid transitions smoothly from vapor to liquid without discontinuities when the path is such
that it does not cross the phase boundary, defined regions are useful to give a general indication of the fluid’s state.
Most definitions of the phase regions use pressure and temperature parameters for bounding the gas, liquid, and
dense regions, resulting in different parsing techniques with varying degrees of complexity depending on a group’s
point of view. Most of these bounding techniques result in some inappropriately labeled areas, e.g., more gas-like
than dense-like or vice-versa, due to the extent to which a certain bounding box extends into a temperature or
pressure space.
As shown in Figure 2, the three phase regions are now defined in terms of the critical density only. (The
critical point is the state where the co-existing liquid and vapor phases have the same density and composition.)
The gas phase is the region with densities less than 50 % of the critical density. Liquids are defined as states with
densities greater than 125 % of the critical density. The dense phase of a fluid is any single-phase state between
these two areas, i.e., a state with a density greater than 50 % and less than 125 % of the critical density. Other
points or curves are also shown in Figure 2, including the cricondentherm (maximum 2-phase temperature), the
cricondenbar (maximum 2-phase pressure), the bubble point curve, the dew point curve, and the retrograde curve.
In the dense phase region near the critical point for any pure fluid or mixture, the properties change rapidly
as the state approaches the critical point. For example, for a pure fluid the isobaric heat capacity increases to
infinity and the speed of sound approaches zero. For temperatures near the critical point, the area between 50 %
and 125 % of the critical density represents states where property changes become more significant.
There are various approaches that can be used to determine the critical density of a mixture, such as the
tools available in Reference [6] that use the GERG-2008 equation of state (given in Part 2) to locate the critical
point. The use of cubic equations of state (PR, SRK, etc.), though not as accurate as the GERG-2008, will
determine a state point that is in the general vicinity of the true critical point of the mixture when volume translation
is applied. The simplest but least accurate approach is to use a mole fraction average of the critical volumes of the
pure fluids (the critical density would then be the reciprocal of the critical volume). The approach taken depends
on the needs of a particular application.
8
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
2.2 Nomenclature
The symbols used in this publication are generally specific to thermodynamics, although those specific to
flow measurement have been used as a convenience to the reader. General symbols that do not refer to a specific
equation of state are given below. Symbols that are specific in the equations of state are given in Sections 4 and 5.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
9
This section contains general equations involving the compressibility factor, molar density, and mass
density of natural gas and other hydrocarbon gases.
PV
Z , (2-1)
nRT
where T is the absolute temperature of the gas, P is the absolute static pressure of the gas, V is the gas volume, n is
the number of moles, and R is the molar gas constant. Both the DETAIL and GROSS equations of state express
the compressibility factor Z in terms of the molar density d (in moles per unit volume),
n
d . (2-2)
V
The gas mixture molar mass (molecular weight) M is calculated from the gas composition with the relation
N
M xi M i , (2-3)
i 1
where Mi is the molar mass of component i, N is the number of components in the gas mixture, and xi is the mole
fraction of component i in the gas mixture. The summation in this equation is over all components in the gas
mixture. The mass density (mass per unit volume) is related to the molar density d by the relation
Md . (2-4)
Combining Equations (2-2) and (2-4) in Equation (2-1) results in the following equations for the molar density d
and the mass density in terms of the compressibility factor:
P
d (2-5)
ZRT
P
M (2-6)
ZRT
Z b Tb , Pb
Fpv2 , (2-7)
Z T , P
where Fpv is the supercompressibility factor expressed as the square root of the ratio of a base compressibility factor
to the operating compressibility factor. The operating compressibility factor Z(T, P) is calculated at the measured
operating pressure and temperature. The base compressibility factor Zb is defined by the base conditions specified
in the custody transfer agreement. Base conditions may be specified by regulation, contract, local conditions, or
organizational needs. In the United States, for inter-state custody transfer of natural gas, they are considered to be
60 °F and 14.73 psia.
10
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Except for the mass and mole conversions (where exact conversions are given), the values given in
Table 3 are identical to those in API Chapter 15, Guidelines for Use of the International System of Units (SI) in the
Petroleum and Allied Industries, and ASTM-SI10, American National Standard for Metric Practice (which
rounded the conversion factors to seven significant digits). Most calculations in this report are not accurate to
more than seven significant digits; thus, the accuracy of computed compressibility factors is not improved by the
use of more significant digits. For example, a calculation precision of five significant digits is approximately 1 part
in 100,000 (or 10 ppm) for the compressibility factor. Appendix B presents information showing that the expected
uncertainties of computed natural gas compressibility factors are 0.048 % (one standard deviation) for both the
DETAIL equation of state and the GROSS equation of state. The uncertainty of 0.048 % corresponds to 48 parts
in 100,000 (or 480 ppm), so the calculation precision of 1 part in 100,000 is an order of magnitude better than the
uncertainty in the computed compressibility factor. The computed compressibility factors in the region of validity
for each equation of state generally agree with the best available compressibility factor data for natural gas with
differences approaching the experimental uncertainty of 0.1 %. Thus, for compressibility factors less than one,
calculated values are at best accurate to about one part in 10,000 (or 100 ppm), or to about four significant digits.
When more than four significant digits are quoted in this document, the purpose is generally for computer program
verification only.
used here to conform with the 1994 edition of AGA 8 so that calculated property values match
those from the original equations.
‡Corresponds to the IT value of 1 cal = 4.1868 J.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
11
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Wi Association parameters
Mixture parameters (Table 6 for ij parameters)
B*nij Binary characterization parameters
C*n Parameters that are functions of composition
Eij Second virial coefficient energy binary interaction parameters
F Mixture high temperature parameters
G Orientation parameters
Gij Binary interaction parameters for orientation
K Size parameters
Kij Binary interaction parameters for size
Q Quadrupole parameters
U Mixture energy parameters
Uij Binary interaction parameters for conformal energy
Properties
a Molar Helmholtz energy (J/mol)
a0 Molar ideal-gas Helmholtz energy (J/mol)
ar Molar residual Helmholtz energy (J/mol)
B Second virial density coefficient (dm3/mol)
cp0 Molar ideal-gas isobaric heat capacity [J/(mol∙K)]
cp Molar isobaric heat capacity [J/(mol∙K)]
cv Molar isochoric heat capacity [J/(mol∙K)]
d Molar density (mol/dm3)
D Reduced density of gas
g Molar Gibbs energy (J/mol)
h Molar enthalpy (J/mol)
Isentropic exponent
N Number of components in the mixture
P Absolute pressure (MPa)
R Gas constant [J/(mol∙K)]
s Molar entropy [J/(mol∙K)]
T Absolute temperature (K)
u Molar energy (J/mol)
w Speed of sound (m/s)
xi Mole fraction of component i in the mixture (mol)
Z Compressibility factor
12
In the DETAIL equation of state, a natural gas is characterized by its composition, that is, the mole
fractions or mole percentages of the components in the natural gas. The equations, constants, and parameters
needed to calculate thermodynamic properties for natural gas mixtures with the DETAIL equation of state are
given here. Procedures to apply the DETAIL equation of state and the available programs are provided in the
supplementary material accompanying this document. An uncertainty discussion is given in Appendix B.
The equation of state used in this method is a hybrid formulation based on the work of Starling et al.,
Reference [1]. It combines features of the virial equation of state (a power series in density) for low-density
conditions and exponential functions for applications at high-density conditions (extended Benedict-Webb-Rubin
equation of state). This formulation provides high accuracy, a broad temperature-pressure-composition application
range, and derived thermodynamic property capabilities. A detailed description of the performance of the equation
is available in Reference [2].
The equation of state for the compressibility factor Z for the DETAIL method is given by the equation
18 58
Z 1 Bd D Cn*T un Cn*T un bn cn kn D kn Dbn exp cn D kn , (4-1)
n13 n13
where d is the molar density of the gas, B is the second virial coefficient, D is the reduced density, Cn* are parameters
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
that are functions of composition, and un, bn, cn, and kn are coefficients and parameters given in Table 4. The units
for temperature and molar density are kelvins and moles per cubic decimeter. The reduced density D is related to
the molar density d by the equation
D K 3d , (4-2)
where K is the mixture size parameter. Converting the expression for the compressibility factor of the gas to
pressure results in the equation
18 58
P dRT 1 Bd D Cn*T un Cn*T un bn cn kn D kn Dbn exp cn D kn . (4-3)
n13 n13
In the computation of Z or P through the use of the DETAIL equation of state, the composition of the gas,
the absolute temperature T, and the molar density d are required. When the temperature, pressure, and composition
of a gas are known, the only unknown quantity in this equation is the molar density d. The density can be
determined with appropriate iterative procedures, as explained in Section 6.
Before computations can be performed with the equations above, the values of B, Cn*, and K must be
calculated from the composition and temperature of the gas. The mixture size parameter K for a mixture of N
components is
2
N 5
N 1 N
K xi Ki 2 2 xi x j Kij5 1 Ki K j 2 ,
5
5
(4-4)
i 1 i 1 j i 1
where the Kij are binary interaction parameters for size (Table 6), Ki and Kj are size parameters for components i
and j (Table 5), and xi and xj are mole fractions of components i and j in the gas mixture.
13
18
B anT u n B* , with (4-5)
n 1
K K
N 1
2 xi x j Eij Ei E j 2
N N
B* xi2 Ei n K i3 Bnii
* u 1 un 3
*
i j
2
Bnij (4-6)
i 1 i 1 j i 1
*
Bnij
Gij Gi G j / 2 1 g n Q Q
gn
i j 1 qn
qn
F F i j 1 fn
fn
S S 1 sn n WiW j 1 w
s wn
(4-7)
i j n
The coefficients Cn* (n=13 to 58) in Equations (4-1) and (4-3) are
Cn* an G 1 g n
gn
Q 2
1 qn
qn
F 1 f n f U u n n
. (4-8)
The parameters in these equations are defined in the nomenclature. The constants an, fn, gn, qn, sn, un, and wn are
given in Table 4. The mixture parameters F, G, Q, and U are calculated with the following equations, where in the
double sums i ranges from 1 to N–1, and each value of j ranges from i+1 to N:
2
N 5
N 1 N
U xi Ei 2 2 xi x j U ij5 1 Ei E j 2
5
5
(4-9)
i 1 i 1 j i 1
N 1
G xi Gi xi x j Gij 1Gi G j
N N
(4-10)
i 1 i 1 j i 1
N
Q xi Qi (4-11)
i 1
N
F xi2 Fi (4-12)
i 1
Values of Ei, Fi, Gi, Ki, Qi, Si, and Wi are given in Table 5. The binary interaction parameters Eij, Gij, Kij, and Uij
are given in Table 6. All values of the binary interaction parameters Eij, Gij, Kij, and Uij are one except for the
values in Table 6.
14
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
The Helmholtz energy a of a mixture is the sum of two contributions (see References [7,8]), one accounting
for that of the ideal gas a0 and the other for that of the real fluid ar,
a a0 ar . (4-13)
a0 d N
ln xi nio,1 nio, 2 / T nio,3 1 ln T
ln xi (4-14)
RT d0
o
o
i 1 ni , k ln sinh i , k / T ni , k ln cosh i , k / T
o o
k 4, 6 k 5, 7
The coefficients are given in Table 7, and the derivation of this equation is given further below. The value of the
molar density d is given in mol/dm3, and that of T is given in kelvins. The real gas contribution is
ar
18 58
Bd D Cn*T un Cn*T un D bn exp cn D kn , (4-15)
RT n13 n13
where a is the molar Helmholtz energy and the superscript r indicates a residual property, or the contribution from
the interaction of molecules in a non-ideal gas. All other symbols were defined under Equation (4-1). This equation
results from the integration of Equation (1-1) with the pressure taken from Equation (4-3), and is the original
formulation used in the development of AGA 8.
The equation for the ideal-gas Helmholtz energy [given in Equation (4-14)] was obtained through the use
of thermodynamic relationships. The contribution to the ideal gas can be expressed as a summation involving the
ideal-gas enthalpy h0 and entropy s0,
a 0 h0 RT Ts0 . (4-16)
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
When an expression for the ideal-gas heat capacity cp0 is available, the ideal-gas enthalpy can be derived from the
relation
h0
c 0p , (4-17)
T p
resulting in
T
h 0 h00 c 0p dT , (4-18)
T0
where the enthalpy at T0 is h00. (The value of T0 is not necessarily set to zero, but can be any arbitrary value,
resulting in the integration constant h00.) Likewise, the ideal-gas entropy can be derived from a fundamental
relationship (see Reference [7]), resulting in
c 0p Td N
R xi ln xi ,
T
s s
0 0
0 dT R ln (4-19)
T0 T T0 d 0 i 1
where d0 is the ideal-gas density at T0 and P0 (d0 = P0/T0R), and T0 and P0 are arbitrary reference conditions. The
molar gas constant R is 8.31451 J/(mol∙K) (as given in the 1994 revision of AGA Report No. 8). The last part of
15
The reference state of zero enthalpy and zero entropy is here adopted at T0 = 298.15 K and P0 = 0.101325 MPa for
the ideal unmixed gas [i.e., no entropy of mixing (the last term in this equation is set to zero)]. Expressions for the
ideal-gas heat capacity cp0 were taken from Jaeschke and Schley (see Reference [9]), given as
c op io,k / T
2
io,k / T
2
N
xi nio,3 nio, k i,k o
.
R i 1
k 4, 6
sinh i , k / T
o
n o
k 5, 7 cosh i , k / T
(4-21)
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
The coefficients and parameters are given in Table 7. The equation for the ideal-gas isobaric heat capacity is taken
from [9] and given as a function of the temperature T. The values of the parameters k (for k = 4 to 7) are in
o
o o o o o
agreement with the original parameters D , F , H , and J in reference [9], and the values of nk (for k = 3 to 7) are
o o o o o
identical to the coefficients B , C , E , G , and I . Integrating Equation (4-21) and combining the outcome with
Equation (4-20) results in the ideal-gas Helmholtz energy given in Equation (4-14).
P d a r
Z 1
RT d T
(4-22)
dRT
a 2 a
r
P d dRT d
2
(4-23)
d T d T
P a r 2 a
2 r
RT 2d
d d 2
(4-24)
d T d T T
2P a r 2a r 2 a
3 r
2 2 4d 2 d 3 (4-25)
d T d T d T d T
P 2 a
2 r
dR d (4-26)
T d dT
The equations for calculating entropy s, energy u, enthalpy h, Gibbs energy g, isochoric heat capacity cv,
isobaric heat capacity cp, speed of sound w, Joule-Thomson coefficient , and the isentropic exponent are given
16
a a 0 a r
s (4-27)
T d T d T d
u a Ts (4-28)
P
hu (4-29)
d
P
g a (4-30)
d
u s 2a 2 a 0 2 a r
cv T T 2 T 2 2
(4-31)
T d T d T d T d T d
2
2 a r
1 R d
T P P dT
2
c p cv 2
cv T (4-32)
d T d d T a r 2 a
2 r
RT 2d
d d 2
d T T
1 T P P 1
1 (4-34)
cpd d T d d T
w2 M
(4-35)
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
ZRT
The following nomenclature is used to simplify the appearance of the derivative equations:
n Cn*T u n
(4-36)
n Cn*T u Db exp cn Dk
n n n
(4-37)
n cn kn Dk n
(4-38)
n bn n (4-39)
i , k io,k / T (4-40)
17
0
N
ln xi nio,1 nio, 2 / T nio,3 1 ln T
ln xi
a d
RT 0
d
ni
o
, k ln sinh i , k ni
o
, k ln cosh i , k
(4-41)
k 4,6
i 1
k 5, 7
i i ,1 i ,3
ln x no no 1 1 ln T
1 a d N
ln xi ni , k ln sinh i , k i , k / tanhi , k
0
o
(4-42)
R T d0 i 1 k 4,6
k
o
ni , k ln cosh i , k i , k tanhi , k
5, 7
T 2a0 2
xi nio,3 1 nio, k i , k / sinh i , k nio, k i , k / coshi , k
N
2
(4-43)
R T 2
i 1 k 4, 6 k 5, 7
ar 18 58
Bd D n n (4-44)
RT n 13 n 13
d a r 18 58
Bd D n n n (4-45)
RT d n 13 n 13
d 2 2a r 58
n n 1 knn n (4-46)
RT d 2 n 13
d 3 3a r 58
n 2 n n 1 knn knn 1 kn 2 n n (4-47)
RT d 3 n 13
1 a r B 18 58
Bd Td D un 1 n un 1 n (4-48)
R T T n 13 n 13
T 2a r B 2 B
2 18 58
2Td T d D u u 1 un un 1 n (4-49)
R T 2 T T 2
n n n
n 13 n 13
d 2a r B 18 58
Bd Td D un 1 n un 1bn n n (4-50)
R dT T
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
n 13 n 13
B 18
T un anT un B* (4-51)
T n1
2 B 18
T2 un un 1 anT u n B* (4-52)
T 2
n 1
18
45 −0.005399808 5 0 0 1.0 0 0 0 0 0
46 −0.2432567 5 1 2 9.0 0 0 0 0 0
47 0.04987016 5 1 2 3.0 0 1 0 0 0
48 0.003733797 5 1 4 8.0 0 0 0 0 0
49 1.874951 5 1 4 23.0 0 1 0 0 0
50 0.002168144 6 0 0 1.5 0 0 0 0 0
51 −0.6587164 6 1 2 5.0 1 0 0 0 0
52 0.000205518 7 0 0 −0.5 0 1 0 0 0
53 0.009776195 7 1 2 4.0 0 0 0 0 0
54 −0.02048708 8 1 1 7.0 1 0 0 0 0
55 0.01557322 8 1 2 3.0 0 0 0 0 0
56 0.006862415 8 1 2 0.0 1 0 0 0 0
57 −0.001226752 9 1 2 1.0 0 0 0 0 0
58 0.002850908 9 1 2 0.0 0 1 0 0 0
19
AGA 8 to maintain consistency. For other properties not related to the DETAIL or GROSS equations of state, the molar masses in AGA 5
or GPA 2145 should be used.
‡
The nomenclature "(K)" in the energy parameter column refers to the kelvin temperature unit of the values in the column, not to the variable
used for the size parameter.
20
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
7 15 n-Butane Hydrogen 1.3
10 19 n-Hexane Hydrogen sulfide 1.008692 1.028973 0.96813
11 19 n-Heptane Hydrogen sulfide 1.010126 1.033754 0.96287
12 19 n-Octane Hydrogen sulfide 1.011501 1.038338 0.957828
13 19 n-Nonane Hydrogen sulfide 1.012821 1.042735 0.952441
14 19 n-Decane Hydrogen sulfide 1.014089 1.046966 0.948338
15 17 Hydrogen Carbon monoxide 1.1
†Values for the j,i pair are equal to the i,j pair, that is, E =E , U =U , K =K , and G =G . Values of 1.0 should be used for all binary
ji ij ji ij ji ij ji ij
interaction parameters except for the entries in this table.
21
4 5 6 7
o o o o
22
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Coefficients
Bmix Second virial coefficient of the mixture (dm3/mol)
Cmix Third virial coefficient of the mixture (dm6/mol2)
B0, B1, B2 Coefficients in the second virial equation for the CH-CH interaction
C0, C1, C2 Coefficients in the third virial equation for the CH-CH interaction
b0, b1, b2, c0, c1, c2 Constants in Table 8
bi0, bi1, bi2, ci0, ci1, ci2 Constants in Table 9
Properties
d Molar density (mol/dm3)
HCH Molar ideal gross heating value of the equivalent hydrocarbon (kJ/mol) at 25 °C
HN oi Molar ideal gross heating value of component i (kJ/mol) at 25 °C
N Number of components in the mixture
P Absolute pressure (MPa)
R Gas constant [J/(mol∙K)]
T Absolute temperature (K)
xCH Mole fraction of the equivalent hydrocarbon
xN2 Mole fraction of nitrogen
xCO2 Mole fraction of carbon dioxide
xi, xj, xk Mole fractions of components i, j, and k in the mixture
Z Compressibility factor
The GROSS equation of state approximates a natural gas mixture by treating it as a mixture of three
components: an equivalent hydrocarbon component (i.e., a pseudo-hydrocarbon component), nitrogen, and carbon
dioxide. The equivalent hydrocarbon CH is used to collectively represent all the hydrocarbons found in the gas
mixture. Nitrogen and carbon dioxide are the diluent components.
The GROSS equation of state allows for the characterization of a natural gas by a number of different
methods. These methods include knowledge of the heating value, relative density, or mole fractions of the
components in the mixture. Two methods were available in the 1994 edition of AGA 8, labeled as Method 1 and
Method 2. Two additional methods have been added in this document to allow direct calculations from the GROSS
equation of state without the need for the iterative procedures required by Methods 1 and 2. These new methods
are labeled as Method 0 and Method 3. A description of each method and the required inputs are given below.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
23
Method 1
Inputs: Volumetric heating value of the mixture, relative density (specific gravity), and the mole
fraction of carbon dioxide.
Method 1 requires an iterative procedure that determines the molar heating value of the equivalent
hydrocarbon HCH and the mole fraction of nitrogen from the inputs to this method. The iterative
method is given in Appendix A. Once HCH and xN2 have been determined, pressures and
compressibility factors are calculated as done in Method 0. Reference conditions must be specified
for the relative density and volumetric heating value.
Method 2
Inputs: Relative density and mole fractions of nitrogen and carbon dioxide.
Method 2 requires an iterative procedure that determines the molar heating value of the equivalent
hydrocarbon HCH. The iterative method is given in Appendix A. Once HCH has been determined,
pressures and compressibility factors are calculated as done in Method 0. Reference conditions
must be specified for the relative density.
Method 3
Inputs: Molar mass of the mixture and mole fractions of nitrogen and carbon dioxide.
Method 3 is similar to Method 0 except that the molar mass of the mixture is used as an input, and
the molar heating value of the equivalent hydrocarbon HCH is determined from the molar mass
through Equation (A-16), as explained in Appendix A. No iterative procedure is required to set
up the method, no reference state is needed, and all subsequent calculations are done as in
Method 0. The constants G1 and G2 in Equation (A-16) were determined in Reference [3] through
a fit of heating values as a function of molar mass, and properties are slightly less accurate than
those obtained with Method 0 through the use of a gas analysis. This method is useful for GROSS
equation of state applications where network bandwidth restrictions might prohibit the
transmission of the full compositional data to a gas flow computer in the field.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
In Methods 1 and 2, the user may still find that the compositional analysis is available at some point
throughout the measurement process (see introduction to Section A.3). Method 0 (or 3) should then be used rather
than the methods given in Appendix A.
The GROSS equation of state calculates compressibility factors for natural gas that contain component
concentrations in the ranges given in Table 2. However, it should only be used for real natural gas, i.e., with heavy
hydrocarbon tail, and not for synthetic gas such as LNG (in which case the uncertainty estimate should be re-
assessed). Equations, constants, and parameters needed to calculate compressibility factors for natural gas mixtures
with the GROSS equation of state are presented in this section. An uncertainty discussion is given in Appendix B.
The GROSS equation of state is a virial type model. It is based on the work of Schouten et al. (1990) (see
Reference [3]). It is also known as the SGERG model (see Reference [4]), and should not be confused with the
GERG-2008 equation of state described in Part 2. The SGERG model was developed with SI units. These units
24
A virial equation of state is a polynomial expansion in density, where the second virial coefficient B
determines the departure of Z from unity at low densities, the third virial coefficient C contributes more to the
departure of Z from an ideal gas state as density increases, and so on. Each density term is preceded by these virial
coefficients, which are functions of temperature and composition. The application of the virial equation of state
via the GROSS model truncates the virial equation for the compressibility factor after the third virial coefficient
term C. This truncation provides high accuracy calculations for natural gas pipeline transmission and distribution
conditions only at pressures up to the limit in Table 2.
The GROSS model expresses the compressibility factor in terms of the molar density d, the mixture second
virial coefficient Bmix, and the mixture third virial coefficient Cmix as given in the following equations:
3 3
Bmix Bij xi x j (5-2)
i 1 j 1
3 3 3
Cmix Cijk xi x j xk (5-3)
i 1 j 1 k 1
where Bij is the component interaction second virial coefficient for the i,j pair, Cijk is the component interaction
third virial coefficient for i,j,k, and xi, xj, and xk are the mole fractions of the gas components. The number of
components in the gas mixture is set to 3: the equivalent hydrocarbon, nitrogen, and carbon dioxide. The mole
fraction of the equivalent hydrocarbon is
The Bij and Cijk terms in Equations (5-2) and (5-3) are the interaction virial coefficient terms. They are
temperature dependent functions. Expanding Equations (5-2) and (5-3) for Bmix and Cmix identifies all the terms
needed to implement the SGERG model, given in the following equations:
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
3CCHCH N 2 xCH
2
xN 2 3CCHCHCO2 xCH
2
xCO2
3CCH N 2 N 2 xCH xN2 2 3CCHCO2 CO2 xCH xCO
2
2
(5-6)
3C N 2 N 2 CO2 xN2 2 xCO2 3C N 2 CO2 CO2 xN 2 xCO
2
2
The following sections provide the equations needed to compute the second and third virial interaction terms for
Equations (5-5) and (5-6).
25
where values for b0, b1, and b2 are given in Table 8, and T is the temperature in kelvins. Similarly, the Cijk values
for the terms involving only nitrogen and carbon dioxide are expressed in dm6/mol2, and are given by
Cijk c0 c1T c2T 2 , (5-8)
where values for c0, c1, and c2 are also given in Table 8.
Table 8 – Interaction Virial Coefficient Terms for Nitrogen and Carbon Dioxide
Bij Binary set b0 (dm3/mol) b1 [dm3/(mol∙K)] b2 [dm3/(mol∙K2)]
N2-N2 –0.1446 0.740910×10–3 –0.911950×10–6
N2-CO2 –0.339693 0.161176×10–2 –0.204429×10–5
CO2-CO2 –0.86834 0.403760×10–2 –0.516570×10–5
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
CO2-CO2-CO2 0.205130×10–2 0.348880×10–4 –0.837030×10–7
N
H CH xi HNio (25 C) / xCH , (5-9)
i1
where HN oi is the gross component molar heating value (for the ideal gas) in kJ/mol at 25 °C and is available in
AGA 5, Natural Gas Energy Measurement, GPA 2172, Calculation of Gross Heating Value, Relative Density,
Compressibility and Theoretical Hydrocarbon Liquid Content for Natural Gas Mixtures for Custody Transfer, or
ISO 6976, Natural Gas – Calculation of Calorific Values, Density, Relative Density and Wobbe Index from
Composition. When the gas composition is not known, the value of HCH must be determined by one of the methods
summarized in Appendix A.
The equations for the second and third interaction virial coefficients for the equivalent hydrocarbon are
BCH-CH B0 B1H CH B2 H CH
2
and (5-10)
CCH-CH-CH C0 C1H CH C2 H CH
2
, (5-11)
where B0, B1, B2, C0, C1, and C2 are temperature dependent functions defined as
Bi bi 0 bi1T bi 2T 2 , i 0,1,2 and (5-12)
The interaction second virial coefficient term for the equivalent hydrocarbon CH with nitrogen N 2 is
calculated as
BCH N2 0.72 1.875 10 5 (320 T ) 2 BCHCH BN2 N2 / 2 , (5-14)
where T is given in kelvins. For the equivalent hydrocarbon CH with carbon dioxide CO2, the relation is
The interaction third virial coefficient terms for nitrogen and carbon dioxide are calculated with the
following equations:
CCH N 2 CO2 1.10 CCHCHCHCN 2 N 2 N 2 CCO2 CO2 CO2 1
3
(5-20)
The equation for calculating the pressure with the GROSS equation of state is obtained by substituting
Equation (5-1) into Equation (2-5), resulting in
P dRT 1 Bmix d Cmix d 2 . (5-21)
The derivative of pressure with respect to density is useful in iterative routines to determine the density
with pressure and temperature inputs, as given by
P
RT 1 2 Bmix d 3Cmix d .
2
(5-22)
d T
Section 6 describes an iterative method to obtain the density in the equation of state.
27
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Flow rate calculations for gas metering applications typically require values of both the compressibility
factor Z at the flowing temperature and pressure, along with the compressibility factor at base conditions Zb. This
calculation requires first the determination of the density at flowing conditions. Procedures for these computations
for the DETAIL and GROSS equations of state are given below and in the supplementary material for this
document.
There are many types of iterative methods that could be implemented to determine density, which is an
independent variable (along with temperature) in the equation of state. Each has advantages (such as bounded
regions that guarantee a root) and disadvantages (such as low speed of calculation or non-bounded procedures).
The iterative process in the supplementary material uses a Newton’s method for rapid convergence. Because
pressures and densities can be very small values, the natural log of these values is used.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
The iterative method starts with the calculation of density from the ideal-gas law, given the input
temperature and pressure, and then converts it to the logarithm of the specific volume,
RT
vln ln d ln input , (6-1)
P
input
where vln = ln(v). With the calculated density, the pressure used in the iteration (labeled Pcalc here) is calculated
from either the DETAIL or GROSS equation of state, along with the first partial of pressure with respect to density.
From Newton’s method, a new density (or logarithm of the specific volume in this case) is calculated from the
difference between the calculated and known values of pressure and the partial derivative,
ln v
vln,new vln,old
ln Pcalc ln Pinput , (6-2)
ln P T
where
ln v
1
P v P P
. (6-3)
ln P T v P T d d T
The variable vln,old is replaced with the new calculated value vln,new, and the iteration continues until the difference
in the pressures is less than some tolerance value.
28
The two equations of state (DETAIL and GROSS) have been incorporated into efficient computer
programs to compute the compressibility factor Z, the molar density d, the mass density , and the
supercompressibility factor Fpv. For the DETAIL method, the other thermodynamic variables given in Section 5
are also calculated. The computer programs can be used for the following purposes:
The computer programs and their documentation are given in Appendix C and the supplementary material
for this document. Tables of computed compressibility factors are provided for the DETAIL and GROSS equations
of state in Appendix B and in the supplementary material. These tables can be used to verify computer programs.
The tables cover a wide range of gas types.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
29
In order to calculate the thermodynamic properties of a natural gas or similar mixture that contains
components that do not appear in Table 1, it is necessary to assign each component to one of the 21 major or minor
components for which the DETAIL equation of state was developed. Recommendations for appropriate
assignments are given in Table 10. Each recommendation is based on an assessment of which substance is likely
to give the lowest overall uncertainty for the complete set of thermodynamic properties. Because, no single
assignment is likely to be equally satisfactory for all properties, it is reasonable that the user may prefer an
alternative assignment for a particular application in which, for example, only a single property is needed.
Implementations that include assignments for trace components need to be carefully documented.
30
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
31
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
1 Starling, K.E., R.T Jacobsen, S.W. Beyerlein, C.W. Fitz, W.P. Clarke, E.W. Lemmon, Y.C. Chen, and
E. Rondon, GRI High Accuracy Natural Gas Equation of State for Gas Measurement Applications–1991
Revision of A.G.A. No. 8 Equation, Technical Reference Document, GRI-91/0184, 1991.
2 Savidge, J.L., S.W. Beyerlein, and E.W. Lemmon, Technical Reference Document for the 2nd Edition of
AGA Report Number 8, GRI-93/0181, 1993.
3 Schouten, J.A., J.P.J. Michels, and M. Jaeschke, Calculation of the Compressibility Factor of Natural Gases
Based on the Calorific Value and the Specific Gravity, Int. J. Thermophys., 11(1):145-156, 1990.
4 Jaeschke, M. and A.E. Humphreys, GERG Technical Monograph 5, Fortschritt-Berichte VDI Reihe 6, No.
266, 1991.
5 Lemmon, E.W., Better Defining the Uncertainties for the AGA 8 Equation, Catalog No. PR-381-12604-R01,
Pipeline Research Council International, Inc., 2013.
6 Lemmon, E.W., M.L. Huber, and M.O. McLinden, NIST Standard Reference Database 23: Reference Fluid
Thermodynamic and Transport Properties-REFPROP, Version 9.1, National Institute of Standards and
Technology, Standard Reference Data Program, Gaithersburg, 2013.
7 Lemmon, E.W., M.O. McLinden, and W. Wagner, Thermodynamic Properties of Propane. III. A Reference
Equation of State for Temperatures from the Melting Line to 650 K and Pressures up to 1000 MPa, J. Chem.
Eng. Data, 54:3141-3180, 2009.
8 Lemmon, E.W. and K.E. Starling, Speed of Sound and Related Thermodynamic Properties Calculated from
the AGA Report No. 8 Detail Characterization Method Using a Helmholtz Energy Formulation, proceedings
of the American Gas Association Conference and Biennial Exhibition, 2003.
9 Jaeschke, M. and P. Schley, Ideal-gas Thermodynamic Properties for Natural-Gas Applications, Int. J.
Thermophysics, 16, 1381-1392, 1995.
10 AGA Report No. 10, Speed of Sound in Natural Gas and Other Related Hydrocarbon Gases, Transmission
Measurement Committee.
11 Schouten, J. and J. Michels, Evaluation of the PVT Reference Data on Natural Gas Mixtures, Gas Research
Institute, 1992.
12 Jaeschke, M. and A.E. Humphreys, GERG Databank of High Accuracy Compressibility Factor
Measurements, GERG Technical Monograph 4, Fortschritt-Berichte VDI Reihe 6, No. 251, 1990.
13 Younglove, B.A., N.V. Frederick, and R.D. McCarty, Speed of Sound Data and Related Models for Mixtures
of Natural Gas Constituents, NIST Monograph 178, 1993.
32
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
(Informative)
A.1 Nomenclature
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Td Reference temperature for molar density (K)
Tgr Reference temperature for relative density (K)
Th Reference temperature for heating value (K)
xCH Mole fraction of the equivalent hydrocarbon
xN2 Mole fraction of nitrogen
xCO2 Mole fraction of carbon dioxide
xi Mole fraction of component i in the gas mixture
Z Compressibility factor
Mass density (kg/m3)
†The values given do not represent the most current scientific values, but are used here to conform with the 1994 edition of AGA 8 so that
calculated property values match the original equation.
When a body of information has been developed for a quantity such as the molar heating value at specific
conditions of temperature and pressure, it is common to refer to these as the reference conditions. In the gas
industry, reference conditions are also commonly referred to as base conditions, normal conditions, standard
conditions, etc.
A number of specific temperature-pressure conditions are used for reference conditions for natural gas
volumetric properties (density, specific volume, or compressibility factor) as outlined in AGA 5, Natural Gas
Energy Measurement, and GPA 2172, Calculation of Gross Heating Value, Relative Density, Compressibility and
Theoretical Hydrocarbon Liquid Content for Natural Gas Mixtures for Custody Transfer. In the U.S. gas industry,
the reference conditions that are the basis for a gas contract are commonly referred to as the base conditions. For
inter-state custody transfer, the reference conditions/base conditions are 60 °F and 14.73 psia. In much of Europe,
the reference conditions 15 °C and 0.101325 MPa are referred to as standard conditions, and the reference
conditions 0 °C and 0.101325 MPa are commonly referred to as normal conditions. The reference conditions for
tabulations of the supercompressibility factor Fpv are 60 °F and 14.73 psia. Because there is a plethora of choices
33
The reference conditions 25 °C and 0.101325 MPa are commonly used in the scientific literature for
experimental data tabulations of molar enthalpies of formation of compounds from the elements. The molar
enthalpies of formation of the compounds involved in combustion are used for calculations of component molar
heating values at the reference conditions. A molar heating value for a compound at specified conditions that is
different from the reference conditions can then be determined from the molar heating value at the reference
conditions plus the enthalpy difference between the two conditions for the stoichiometric amounts of the
compounds involved in the combustion reaction.
The volumetric heating value is the molar heating value multiplied by the molar density. Separate reference
conditions occur for the volumetric heating value when the reference conditions are different for the molar heating
value and the molar density. Therefore, the reference conditions for both the molar heating value and the molar
density must be specified to obtain the volumetric heating value.
When the complete compositional analysis of the natural gas is not known, heating values of reliable
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
accuracy can be measured with combustion calorimeters. Natural gas heating values can also be estimated from
limited characterization information such as relative density, nitrogen content, and carbon dioxide content,
although the resulting estimates of heating values generally have greater uncertainties than measurements from
combustion calorimeters or calculations with complete compositional analyses.
The methods for use of the GROSS equation of state have been changed in this edition. Because the gas
composition is often available for use in calculations of thermodynamic properties, the methods given below are
not required and one can proceed directly to the calculation of compressibility factor through the use of Method 0
outlined in Section 5 without the use of the gross volumetric heating value or relative density. In cases where a
gas analysis is not available, either Method 1 or 2 given below is required to obtain the HCH value needed in
Equations (5-10) and (5-11) of the GROSS equation of state. These methods can be implemented when two
particular sets of three of the following four characterization parameters are available:
* volumetric gross heating value
* relative density (specific gravity)
* mole fraction of nitrogen
* mole fraction of carbon dioxide
Method 1 is used for inputs of heating value, relative density, and carbon dioxide content. Method 2 is
used for inputs of relative density, nitrogen content, and carbon dioxide content. Iterative procedures are required
to obtain the necessary parameters in the SGERG model with Methods 1 and 2. These procedures are based on
satisfying rigorous mole, mass, and energy balance equations. General equations specific to both models are given
below, and the details of the methods are explained in Sections A.3.1 and A.3.2.
One additional method, Method 3, is available when only the molar mass of the mixture and the mole
fractions of nitrogen and carbon dioxide are available. This method is not iterative, and is mostly identical to
Method 0, except in the calculation of the molar heating value of the equivalent hydrocarbon.
34
The reference conditions Th for the molar gross heating value may differ from the reference conditions Td
and Pd for the molar density. Th, Td, and Pd must be specified to obtain the reference conditions for the volumetric
heating value.
o
The molar ideal gross heating value HN at 25 °C is
HN o 25 C ZRTd / Pd HV Th , Td , Pd 1 1.027 104 Th 298.15 , (A-2)
where T is given in kelvins and Z is the compressibility factor at the reference conditions Td and Pd. This is only
an approximate equation for natural gas. In this equation, the volumetric gross heating value HV is given in
MJ/m3 at the reference conditions Th and the real molar density at the mixture reference conditions Td and Pd.
AGA 5, Natural Gas Energy Measurement, should be consulted for additional information on heating values and
reference states.
The relative density (specific gravity) Gr at the reference conditions Tgr and Pgr is defined by the relation
Tgr , Pgr
Gr Tgr , Pgr
ref Tgr , Pgr
, (A-3)
where ref is the real mass density of the reference fluid (previously defined as that for air, see GPA 2145, Table
of Physical Properties for Hydrocarbons and Other Compounds of Interest to the Natural Gas Industry, for
additional information), Tgr is the reference temperature, and Pgr is the reference pressure. (The reference
conditions indicated with a subscript "gr" are generally the same as those labeled with the subscript "d", but used
here for situations where the two are not identical.) In this relation for Gr(Tgr, Pgr), both the gas mixture density
(Tgr, Pgr) and the reference density ref (Tgr, Pgr) must be at the same temperature-pressure conditions Tgr and Pgr.
From Equation (2-6) for the real mass density of both the gas mixture and reference fluid, the following relation
results
The relative density is not a constant but varies with both Tgr and Pgr. The molar mass of the reference
fluid Mref is 28.9625 g/mol. The molar mass (molecular weight) M of the mixture can be estimated from Equation
(A-4) as a function of the relative density with the relation
where the equation in parenthesis in the denominator represents the second virial coefficient of the reference fluid
(air in this situation).
35
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
The procedures for calculating the ideal molar gross heating value of the equivalent hydrocarbon HCH and
the nitrogen mole fraction xN2 based on inputs of volumetric gross heating value, relative density, and mole fraction
of carbon dioxide is outlined below.
o
The compressibility Z is initially set to 1, and the molar ideal gross heating value HN at 25 °C is calculated
from Equation (A-2). The molar mass M is calculated from Equation (A-5). The mole fraction of the equivalent
hydrocarbon is
xCH
M xCO2 1 M N 2 xCO2 M CO2 G2 HN o
, (A-6)
G1 M N 2
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
where G1 = –2.709328 g/mol and G2 = 0.021062199 g/kJ. The mole fraction of nitrogen xN2 is
HN o
H CH . (A-8)
xCH
BCH-CH is calculated with this value from Equation (5-10), and Bmix is calculated with Equation (5-2). The
compressibility factor is then updated as
Bmix Pgr
Z new 1 . (A-9)
RTgr
The calculation procedure is iterative. These equations are repeated, replacing Z with Znew iteratively until the
absolute value of (Z/Znew – 1) is less than the convergence criterion (5×10-11 in double precision or 5×10-7 in single
precision). This is not a rigorous calculation for the compressibility factor because it does not include the third
virial coefficient Cmix; however, this contribution is mostly negligible near atmospheric pressure.
The procedure for calculating the heating value of the equivalent hydrocarbon HCH based on inputs of
relative density at the reference conditions Tgr and Pgr and mole fractions of N2 and CO2 is outlined below. In this
method the mole fraction of the equivalent hydrocarbon xCH is determined by
where the molar mass of the mixture M is calculated from Equation (A-5) with an initial value for the
compressibility factor Z set to 1.
36
M CH G1
H CH , (A-12)
G2
where G1 = –2.709328 g/mol and G2 = 0.021062199 g/kJ. BCH-CH is calculated with Equation (5-10) and Bmix is
calculated with Equation (5-2). The compressibility factor is then updated as
Bmix Pgr
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Z new 1 . (A-13)
RTgr
The calculation procedure is iterative. These equations are repeated, replacing Z with Znew iteratively until the
absolute value of (Z/Znew – 1) is less than the convergence criterion (5×10-11 in double precision or 5×10-7 in single
precision). This is not a rigorous calculation for the compressibility factor because it does not include the third
virial coefficient Cmix; however, this contribution is mostly negligible near atmospheric pressure.
The procedure for calculating the heating value of the equivalent hydrocarbon HCH based on inputs of
molar mass M and mole fractions of N2 and CO2 is outlined below. In this method the molar mass of the equivalent
hydrocarbon MCH is
M xN 2 M N 2 xCO2 M CO2
M CH . (A-15)
1 xN 2 xCO2
M CH G1
H CH , (A-16)
G2
37
(Informative)
B.1 Nomenclature
This appendix provides example calculations that can be used to verify computer programs. Table B.1
provides five reference natural gas compositions. Relative densities and volumetric gross heating values at 60 °F
and 14.73 psia calculated from the compositions given in Table B.1 are given in Table B.2. Tables B.3 and B.4
provide computed compressibility factors from the DETAIL and GROSS equations of state for the compositions
in Table B.1 at various temperatures and pressures that can be used to verify calculations, and Table B.5 provides
values for the speed of sound. The Excel worksheet used to generate the outputs is given in the supplementary
material.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Component mole percent
Component Gulf coast Amarillo Ekofisk High N2 High CO2-N2
Methane 96.5222 90.6724 85.9063 81.441 81.212
Nitrogen 0.2595 3.1284 1.0068 13.465 5.702
Carbon dioxide 0.5956 0.4676 1.4954 0.985 7.585
Ethane 1.8186 4.5279 8.4919 3.3 4.303
Propane 0.4596 0.828 2.3015 0.605 0.895
i-Butane 0.0977 0.1037 0.3486 0.1 0.151
n-Butane 0.1007 0.1563 0.3506 0.104 0.152
i-Pentane 0.0473 0.0321 0.0509 0.0 0.0
n-Pentane 0.0324 0.0443 0.048 0.0 0.0
n-Hexane 0.0664 0.0393 0.0 0.0 0.0
38
SI units†
Inputs for Method 1
HV (MJ/m3) 40.7259 40.6797 43.5627 35.6381 36.6798
Gr 0.581202 0.608804 0.649717 0.645006 0.686144
CO2 (mole %) 0.5956 0.4676 1.4954 0.985 7.585
Intermediate values for Method 1
HCH (kJ/mol) at 25 °C 917.200 942.049 996.951 930.348 944.256
o
HN (kJ/mol) at 25 °C 909.539 908.418 972.297 796.129 819.024
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
39
40
41
SI units
°C MPa m/s
0 0.101325 419.588 409.328 393.904 399.395 385.819
0 5 399.337 388.304 366.146 384.553 364.887
0 10 402.828 391.991 369.013 390.693 367.458
20 0.101325 433.516 422.883 406.859 412.701 398.592
20 5 419.292 407.997 386.433 402.896 383.711
20 10 423.185 411.795 387.916 409.351 386.542
40 0.101325 446.689 435.714 419.137 425.307 410.701
40 5 437.013 425.457 404.202 419.322 400.369
40 10 441.898 430.145 406.370 426.488 404.188
60 0.101325 459.204 447.914 430.831 437.297 422.230
60 5 453.052 441.245 420.134 434.281 415.420
60 10 458.970 446.925 423.389 442.197 420.294
42
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Compressibility factors and densities calculated with both equations of state were compared with
experimentally measured values on five simulated natural gas mixtures (Reference [2]). These mixtures were
gravimetrically prepared to represent key natural gas samples occurring in North America and Europe. Table B.1
provides the compositions of these samples. The uncertainty of these reference data for temperatures from –10 °F
to 170 °F (250 K to 350 K) and pressures to 1750 psia (12 MPa) at a 95 % confidence level is 0.07 % in density.
The uncertainty between –10 °F to 170 °F (250 K to 350 K) and pressures from 1750 psia to 5200 psia (12 MPa to
36 MPa) is less than 0.2 % in density (References [11,12]).
Computed compressibility factors and densities from the DETAIL equation of state agree on average
within 0.025 % of a best fit of the experimental data in the region from 35 °F to 170 °F and pressures up to 1750
psia (275 K to 350 K and pressures up to 12 MPa). At conditions up to 5200 psia (36 MPa), the agreement between
calculated values and a best fit of the experimental data above 35 °F (275 K) is 0.1 %. For high pressure reservoir
applications, the DETAIL equation of state predicts compressibility factors often better than 0.2 % when compared
to the PVT reference database.
Computed compressibility factors and densities from the GROSS equation of state agree on average within
0.04 % of a best fit of the experimental data in the region from 35 °F to 170 °F and pressures up to 1750 psia
(275 K to 350 K and pressures up to 12 MPa) (References [2] and [3]).
Both equations of state were used for compressibility factor and density calculations for 4850 natural gas
data points in the GERG data bank (Reference [12]). These samples were not gravimetrically prepared. The
compositions were determined by gas chromatography. The uncertainty in the chromatographic analysis and
component identification is not insignificant when considering calculation uncertainties of 0.05%. Statistical
analyses of the differences between calculated and experimental compressibility factors were performed to evaluate
the uncertainties in the calculated compressibility factors and densities. Statistics were calculated with the
following equations where N is the number of data points:
Z Z exp
Z diff calc 100% (B-1)
Z
exp
N
1
AAD
N
Z
i 1
diff,i
(B-2)
N
1
BIAS
N
Z
i 1
diff,i
(B-3)
1
1 N
Z diff,i BIAS 2
2
Std. Dev.
N 1 i 1
(B-4)
1
1
Z
N 2
RMS
2
diff, i (B-5)
N i 1
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
Calculations were performed for 96 natural gas compositions in the range 17 °F to 143 °F (265 K to 333 K)
with pressures up to 1750 psia (12 MPa). These natural gas compositions were classified into seven groups
according to composition:
43
For the DETAIL equation of state, the uncertainties in calculated compressibility factors are smallest for
the natural gas compositions with more than 94 mole percent methane (Group 6), for which the standard deviation
is 0.034 % and the maximum deviation for 530 data points is 0.148 %. The uncertainties in calculated
compressibility factors are largest for the natural gas compositions with more than 9.5 mole percent nitrogen
(Group 1), for which the standard deviation is 0.061 % for 660 data points. The maximum deviation point in the
total of 4850 data points occurs for the natural gas compositions with more than 8 mole percent ethane (Group 3);
this maximum deviation is 0.256 %. Overall, for the total of 4850 data points, the standard deviation is 0.049 %,
the root mean square deviation is 0.050 %, the bias is –0.010 %, and the average absolute deviation is 0.036 %.
For the GROSS equation of state (Method 0), the uncertainties in calculated compressibility factors are
smallest for the natural gas compositions with more than 94 mole percent methane (Group 6), for which the
standard deviation is 0.036 % and the maximum deviation for 530 data points is –0.137 %. Aside from the group
with high hydrogen content (Group 4), the uncertainties in calculated compressibility factors are largest for the
natural gas compositions with more than 8 mole percent ethane (Group 3), for which the standard deviation is
0.061 % for 1403 data points. The maximum deviation point in the total of 4850 data points occurs for the natural
gas compositions with more than 8 mole percent ethane (Group 3); this maximum deviation is 0.579 %. Overall,
for the total of 4850 data points, the standard deviation is 0.057 %, the root mean square deviation is 0.058 %, the
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
bias is –0.012 %, and the average absolute deviation is 0.038 %. Table B.6 also shows the uncertainties for the
GERG-2008 equation of state, which is described in Part 2 of this document.
The uncertainty in computed supercompressibility factors Fpv is related to the uncertainties in the computed
values of both the compressibility factor Z at any given temperature and pressure and the compressibility factor
Z(60 °F, 14.73 psia). If there were no uncertainty in Z(60 °F, 14.73 psia), the uncertainty in Fpv would be one-half
the uncertainty in Z. If the uncertainty in Z(60 °F, 14.73 psia) were equal to the uncertainty in Z , the uncertainty
in Fpv would be 0.707 times the uncertainty in Z. Because the uncertainty in Z(60 °F, 14.73 psia) cannot exceed the
uncertainty in Z, the uncertainty in Fpv generally will be between 0.5 and 0.7 times the uncertainty in Z. For the
4850 data points for the 96 natural gas compositions in the custody transfer region of Table B.6, the expected
uncertainty in computed values of Fpv is 0.024 % to 0.035 % for both equations of state. With the DETAIL equation
of state, if there were no uncertainty in the natural gas composition, then 95 % of the time the uncertainty in a
calculated supercompressibility factor would not exceed 0.035 %, and would not be expected to exceed 0.13 % in
Range A of Table 1 in the main document. For the GROSS equation of state, if there were no uncertainty in the
natural gas characterization (heating value, relative density, and nonhydrocarbon content), then 95 % of the time
the uncertainty in a calculated supercompressibility factor would not exceed 0.034 %, and would not be expected
to exceed 0.16 % in Range A of Table 1 in the main document.
44
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
All 96 4850 0.034 –0.005 0.050 0.051 0.606
The uncertainties of calculated speeds of sound depend on temperature, pressure, and composition. The
uncertainties were evaluated by comparing calculated values to experimentally measured speed of sound from
NIST Monograph 178 (Reference [13]). Calculations were compared with experimentally measured values for 17
gravimetrically prepared natural gas mixtures over the range of –10 °F to 165 °F (250 K to 350 K) and pressures
up to 2500 psia (17 MPa). Some of the gas mixtures included in the uncertainty analyses are outside the range of
Table 1. The measurements conducted demonstrate that the uncertainty in the speed of sound is within 0.1 % for
Gulf Coast, Amarillo, and Ekofisk gas for pressures up to 1750 psia (12 MPa) and temperatures between –10 °F
and 165 °F (250 K and 350 K). The uncertainty in the speed of sound is also within 0.1 % for other gas mixtures
whose characteristics fall within the ranges of Table 1. Higher levels of uncertainty exist for gas mixtures outside
of these ranges.
45
(Informative)
1. Property Calculation & Verification
AGA8.xls – This Excel spreadsheet can be used to calculate thermodynamic properties of natural gas at
any composition, temperature, and pressure, or to determine the differences between calculated values
from the DETAIL, GROSS, and GERG-2008 equations of state.
To access this file, use the link: http://www.techstreet.com/direct/aga/AGA8.xls
2. Directory: TestData
NG Compositions.xls – Contains 200 compositions collected from industry that represent the full spectrum
from nearly 0 to 100 % methane.
Test Data.xls – Contains calculated data points at multiple temperatures and pressures. These data can be
used for validation of applications incorporating the AGA 8 Parts 1 and 2 source code or for testing
modifications of other code. Compositions of the data include both binary mixtures of every component,
and the data from the NG Compositions.xls spreadsheet. The data are given along the dew and bubble-
point curves. Many of the data points are outside the ranges of the AGA 8 Part 1 equations, and the values
should only be used to confirm that a separate application is working correctly.
Deviation Calculator.xls – Calculates the deviation at any temperature and pressure between the DETAIL
and GERG-2008 equations of state. The calculations will identify if the point is within the 2-phase or liquid
phase. This file requires that the Refprop program from NIST be installed on the machine performing the
calculations (which is used to determine the phase of each point). The spreadsheet also contains the data
from the NG Compositions.xls file for comparisons of a large range of gas mixtures. The AGA8.xls file in
the AGA8CODE directory can also calculate deviations between the different models, including the
GROSS model that is not available in this Excel file, but it does not locate the phase boundaries. For states
known to be in the gas phase, the AGA8.xls file is the best method for calculating deviations, and does not
require the Refprop program.
Tables.xls – Contains the calculations used to generate the tables in AGA 8 Parts 1 and 2 plus calculations
to verify that the tables in the 1994 edition are duplicated correctly in this revision.
3. Directory: AGA8Code
This directory contains three subdirectories, each containing identical code originally developed in VB,
and then translated to Fortran and C. Instructions are available at the top of each file that help explain the
use of the code. The Property Calculation & Verification file, AGA8.xls, described above, is also included
in this directory. The VBA code in the Excel file is identical to the code given in the VB subdirectory.
AGA8.xls – Contains the VB code in AGA8.bas and allows for calculations within Excel to obtain
properties from the AGA 8 Parts 1 and 2 equations.
C directory – Contains C code implementing the AGA 8 Parts 1 and 2 equations.
Fortran directory – Contains Fortran code implementing the AGA 8 Parts 1 and 2 equations. The *.cmn
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
files are ‘include’ files that contain all of the common blocks.
VB directory – VB code implementing the AGA 8 Parts 1 and 2 equations. The VB code is identical to
that in the AGA8.xls file.
To access the files described in items 2 and 3 above, use the link: https://pages.nist.gov/AGA8/
46
47
48
Name: ________________________________________________________________
Company:_____________________________________________________________
Address_______________________________________________________________
Statement of problem and substantiation for proposal, (use separate sheet): (State the
problem that will be resolved by your recommendation; give the specific reason for your proposal including copies of
tests, research papers, etc.)
This proposal is original material. (Note: Original material is considered to be the submitter’s own
idea based on or as a result of his/her own experience, thought or research and, to the best of his/her knowledge, is not
copied from another source.)
This proposal is not original material; its source (if known) is as follows:
______________________________________________________________________
Type or print legibly. If supplementary material (photographs, diagrams, reports, etc.) is included, you may be
required to submit sufficient copies for all members of reviewing committees or task forces.
I hereby grant the American Gas Association the non-exclusive, royalty-free rights, including non-exclusive, royalty-free
rights in copyright, in this proposal and I understand that I acquire no rights in any publication of the American Gas
Association in which this proposal in this or another similar or analogous form is used.
--```,,````,,,`,``,,``,`,,`,,-`-`,,`,,`,`,,`---
49
50