Answers to Frequently Asked Questions
The following information gives answers to some of the most frequently asked questions concerning the REFPROP program (www.nist.gov/srd/nist23.htm). To download any of the files listed below, right-click on the file and select Save Target As... and then add the file to your Refprop or other appropriate directory. Sections with new information added during the last month or two are identified with a «. For questions not answered here, please see the user's guide at www.nist.gov/srd/PDFfiles/REFPROP8.PDF
«Please note:
There have been a number of users who have had problems with the Excel
link to Refprop. In some of these cases
and in other situations we have found that the refprop.dll file had also been
installed in either the c:\windows directory or the c:\windows\system32
directory. If you experience problems,
please do a full hard drive search for refprop.dll and delete all occurrences
except the one in your c:\program files\refprop directory.
1. Reference States (enthalpy and entropy differences)
2. REFPROP is a Program, not a Database Containing Measurements
3. Referencing the REFPROP Program in Publications
5.
Updates to Version 8.0
Questions about Installing the Program
7.
Help File
Fluids
8.
HFC-1234yf
9.
Humid Air
10. Transport
Properties for Nitrogen, Oxygen, Argon, and Air
11. Dimethyl Ether
12. Solids
Two-Phase States
13. Defining Two-Phase States for Pure Fluids
14. Multiple
and Metastable States
15. FLSH Routines and Metastable States
Error Messages
16.
Convergence Failures and Forcing Phase
Calculations
17. Mixture Error
Messages
Linking with other Applications
19.
Required Fluids for Distribution
20. Mixture Files
Reference
States. For differences in enthalpy and entropy
between the Refprop graphical interface and the Refprop Excel sample
spreadsheet, or for differences in enthalpy and entropy between Refprop and
tables of properties given in handbooks:
The absolute values of enthalpy, entropy, and energy at a single state point
are meaningless. It is only the difference between two different state
points that matter. Thus, the value for a single state point can be set
to any arbitrary value. Many handbooks set the arbitrary state point so
that the values of these properties are positive for most liquid or gas
states. The user can change the values of the arbitrary state points by
going to the Options/Reference State menu.
For mixtures, there are additional options that can be set to affect the manner
in which these properties are calculated. The preset values sent out with
Refprop 7.0 are different between the graphical interface and the Excel
file. In the Options/Reference State menu, there are two choices
at the bottom of the menu on the far right. By changing this option, the
two programs will then return the same values. This option can be
permanently saved in the graphical interface by selecting Options/Save
Options, and the saving the options under the file name defaults.prf.
To change the default in the Excel file, press Alt-F11 to bring up the VB
code. If the code does not appear, make sure the project explorer is
visible (View/Project Explorer), and then click on modules and then on
module1. Search for the call to SETREF, and
change the second input from 2& to 1&. More information on this
can be found in your Refprop\Fortran directory in the file SETUP.FOR
under the SETREF subroutine.
To permanently change the default setting for a pure fluid, edit the fluid file
and look at the 14th line of the file, which will appear similar to
this:
IIR
!default reference state
Remove this line and add two lines in its place, similar to the following:
OTH
!default reference state
300.0 1.0 10000.0
100.0 !tref, Pref, Href, Sref
These lines will set the enthalpy to 10000 J/mol and the entropy to 100 J/mol-K
at 300 K and 1 kPa. These values can be changed to any other desired
values.
REFPROP is a Program, not a Database Containing Measurements. The REFPROP "database" is actually a program and does not contain any experimental information, aside from the critical and triple points of the pure fluids. The program uses equations for the thermodynamic and transport properties to calculate the state points of the fluid or mixture. These equations are the most accurate equations available world wide. A link to one of these equations for R-125 is given below. Their high accuracy is obtained through many coefficients in the equations, and thus the calculation speed will be slower than other equations such as the Peng-Robinson cubic equations. The equations are generally valid over the entire vapor and liquid regions of the fluid, including supercritical states; the upper temperature limit is usually near the point of decomposition of the fluid, and the upper pressure (or density) limit is defined by the melting line of the substance.
Referencing
the REFPROP Program in Publications. The following reference can be used to cite
the REFPROP program in publications:
Lemmon, E.W., Huber, M.L., McLinden,
M.O. NIST Standard Reference Database 23: Reference Fluid
Thermodynamic and Transport Properties-REFPROP, Version 8.0, National
Fitting Equations of State. The number of modelers developing new highly-accurate equations of state is decreasing rapidly throughout the world. Equations of state include equations for the thermodynamic properties for pure fluids, for the thermodynamic properties of mixtures, and for the transport properties of pure fluids. Many have expressed interest in acquiring new equations of state, but we do not have the time here at NIST to fulfill all these requests. The amount of work required to develop these equations can be seen in the paper for R-125 (see the link below). If you are interested in learning the procedure for developing new equations, please contact Eric Lemmon for further information (contact information is given at the bottom of this document).
Updates
to Version 8.0. Version 8.0 (from April, 2007) is still the
current version of the Refprop program.
We are currently working on version 8.1 which will contain several
important updates. Some of these include
the addition of the heating values for the natural gas components, better
convergence of VLE states, and the extended GERG-2004 equation of state
for natural gas mixtures. The release
date of this new version is unknown, but will not be within the next six
months.
(Feb. 18, 2009) Those that wish to play with the latest beta version of the DLL
can do so by downloading the file below and placing it in your Refprop
directory. Before you do this you should rename your old DLL file in case this
one produces severe mistakes (in which case you should revert back to the
original one that came with 8.0.) This
latest DLL contains updates to the saturation routines in SATT
and SATP. A
new algorithm has been added that is much more stable than the one in 8 for
finding the saturation states for binary mixtures. The change has not yet been made for ternary
or higher order mixtures, but that should soon become available. These new routines are especially helpful
with the ammonia/water system. This file
will be updated from day to day as additional pieces are added.
(Mar. 9, 2009) Several minor problems have been removed and the routine TPFLSH now uses the new algorithms resulting in much better
convergence in the 2-phase region.
(Nov. 18, 2009) A new algorithm written by Diego Ortiz has now been added to
the code. The program speed is still
about the same, but the convergence is now better for both binaries and multicomponent
mixtures. Keep an eye on this web page for
new updates during the next few weeks.
Installation
Problems. In some applications where calculations fail
after installing version 8.0, the old DLL from version 7.0 may be hiding in
your Windows\System32 directory. Open up
this directory and search for Refprop.dll.
If you find it, delete it (it should never be stored in the Windows
directory). The new one belongs only the
Refprop directory along with the executable.
If the following error message appears when the graphical interface is
launched:
run-time error 339
component MSHFLXGD.OCX
or one of its dependencies not correctly registered:
a file is missing or invalid
then check to see if you have a firewall or antivirus software that can
be turned off. Deactivating these will often allow successful installs,
especially on
MSHFLXGD.OCX
In some cases where you are not allowed to be the administrator on a machine, this
message will appear because it cannot access the System32 directory. Try placing the file in your Windows
directory to get the program to work.
The comdlg32.ocx file may also sometimes be needed:
COMDLG32.OCX
If an error message occurs referencing the file REFPROP.MSI
(which can occur after installing Google Chrome), try the following: Delete everything in the registry and on the
hard drive that has "refprop" in the title. Search for "NIST" in the registry
and you should find a folder that contains something similar to the format
"{79C12B2A-D29F-45A3-BEDD-9003DCFB723C}". Write down the sequence of numbers and
letters that you find in your registry and then delete this entry. Search the registry for the sequence of numbers
and letters that you just deleted. You
should find an instance of something like "HKEY_CURRENT_USER\
Software\ Microsoft\ Installer\ Products\
6398724E8B370524FA88122E26945D8F".
Delete that folder and then the installation should work.
Help
File. The most recent help file from Refprop is available
below (has not changed since version 8.0 came out).
REFPROP.HLP
HFC-1234yf. A very preliminary equation of state is now available (May 8) for this new fluid developed by DuPont and Honeywell. The fluid file is located below and should be placed in your Refprop\Fluids directory. This new equation has been developed with limited data only and should not be referenced in any publication or other use. It is for use only in preliminary calculations involving systems with R-1234yf and the user of this fluid file should check this website from time to time for updates to this fluid file.
Humid Air. Although version 8.0 allows mixtures of nitrogen, argon, oxygen, and water as a consequence from the addition of the new natural gas mixture model, calculations for moist air have not been tested yet. It is likely that calculated values are reasonable, however, Refprop may not return results because the saturation routines may fail. The program calls the saturation routines to determine if the state is vapor, liquid, or two-phase. If you know that your state point is in the vapor phase, you can avoid the call to the saturation routines by using TPRHO instead of TPFLSH. The Fortran file FLSH_SUB.FOR gives additional information concerning the inputs to these routines. A moist air mixture could be made up starting with the composition of dry air used in Refprop: 0.7812 nitrogen, 0.0092 argon, and 0.2096 oxygen (on a mole basis). A small amount of water could be added to this composition and then normalized.
Transport Properties. The
transport properties for nitrogen, argon, and oxygen in version 7.0 did not
include the thermal conductivity enhancement for the critical region and did
not represent the experimental data as well as possible. The transport
equations for these fluids have been redone and now represent the data to
within their experimental uncertainties. The publication below documents
the new equations and shows all of the comparisons to data.
N2-Ar-O2
Transport equations documentation
Dimethyl
Ether. The synonym "ethylene oxide" was
incorrectly included in the DME.FLD file for dimethyl
ether. The correct fluid file is available below:
DME.FLD
Solids. The
REFPROP program does not know the location of the solid-liquid
interface for a mixture. For many of the pure fluids, melting line
auxiliary equations are available and can be used to calculate liquid
properties at the point where solids begin to form and can be used to keep the
program from entering the solid phase. When melting lines are not
available, the program uses the liquid phase density at the triple point as the
maximum density, thus valid states between this density and the melting line
will not be available. The location of the solid-liquid boundary can be
calculated under the Calculate/Saturation Tables option. This option will
also print out the vapor phase properties along the sublimation line if
requested (and if an auxiliary equation is available).
Defining Two-Phase States for Pure Fluids.
Defining the state of a fluid normally requires two inputs, such as
pressure-temperature, temperature-enthalpy, pressure-enthalpy, and so
forth. This is true for the single-phase
states and for two-phase solutions with mixtures of fluids. (Some inputs may have two solutions,
this was described earlier.) For pure fluids,
using inputs of pressure and temperature is not sufficient to describe the
state of the fluid since both remain constant between the liquid and vapor
states. Some other property, such as
quality, enthalpy, or density, is required to specify the two-phase state point
for the pure fluid. Once the quality is
known, some of the other thermodynamic properties can be calculated with the
equation M=q*Mvap
+ (1-q)*Mliq,
where M is the property of interest
and q is the quality. There are several properties that cannot be
calculated this way, including the heat capacities, the speed of sound, and the
transport properties. These quantities
are undefined in the two-phase region for any fluid, except Cv, which is defined very differently
than one would expect. There are some
people who use a different formula to calculate the speed of sound in the two
phase, but it is applicable only in certain specific applications. In these situations, it is best to consider
the properties of the liquid and of the vapor separately, and how they interact
with the application being developed.
As an example, consider Cp, which could be calculated from any
equation of state using the quality, but thermodynamically is not defined for a
two-phase mixture. Cp is the
heat capacity along an isobaric process and is equal to dH/dT. Since pressure and temperature do not change
across the two phase for a pure fluid, then that means Cp would be equal to infinity because the heat capacity changes but
the temperature does not, thus the definition makes it thermodynamically
impossible to calculate it. The only place that Cp can be infinity is at the critical point.
Multiple and Metastable States. There are cases where an input state point can result in two separate valid states. The most common is temperature-enthalpy inputs. Viewing a P-H diagram (or a T-H diagram with very high pressures) will help show how there can be two valid states points for a given input. For example, nitrogen at 140 K and 1000 J/mol can exist at 6.85 MPa and at 60.87 MPa. When this situation occurs, Refprop returns the state with the higher density. For these state points , the character '<' or '>' can be added to the number (before or after) to find the lower or upper root, respectively. In the Excel sample file, the use of these characters was added after the release of version 8.0; the new example file REFPROP.XLS (given below in the Excel section) contains new code and an example of its use. These same characters can also be used to find metastable fluid states for temperature and pressure inputs. For example, the saturation pressure for nitrogen at 100 K is 0.778 MPa. Inputs of 100 K and 0.777 MPa will return a vapor state, but inputs of '100' for temperature and '0.777>' for pressure will return a metastable liquid state. See below for additional information about the use of the '>' and '<' symbols.
FLSH Routines and
Metastable States. For subroutines such as ENTHAL, ENTRO, CVCP, etc., where temperature and density are the input
properties, the output property data will differ from those returned when using
routines such as PHFLSH or TPFLSH
when in the two-phase region. Any routine
(except TDFLSH) that uses temperature and density as
inputs will return what appears to be erroneous results in the two-phase
region. These results actually show calculations directly from the
equation of state without taking into account the fact that the mixture has
split into two phases. The results would be valid for such situations
where a substance is heated beyond its boiling point (a metastable state), but
without boiling (such as water in a glass container being heated in a
microwave). The results eventually end up at the spinodal,
beyond which it is no longer possible to increase the temperature without
boiling the liquid. Thus, for typical results these routines should never
be used for two-phase calculations, rather the FLSH
routines should be called. The flash routines return the properties in
the liquid and vapor states, and these can be used to call routines such as THERM, CVCP, TRNPRP,
etc., with the associated liquid or vapor density at the saturated temperature.
Convergence Failures and Forcing Phase Calculations. (Please also see the updates section above with improved convergences.) When using mixtures, error messages will sometimes be reported when Refprop fails to converge during an iteration, usually during the calculation of a VLE state close to the critical point. In some situations you may know that the point is in the single phase and wish to force the calculation in spite of the error message. This can be done with the '>' and '<' symbols. When one of these is attached to a number in the Specified State Points table, Refprop will assume that the point is in the liquid phase when '>' is attached and in the vapor phase when '<' is attached. Care must be taken because metastable states will be returned in the point is in the two phase. When a state point is above the critical point but Refprop reports a nonconvergence, you can use either of these symbols to force the calculation. For example, a 0.5/0.5 mixture of methane/propane will report a nonconvergence at 300 K and 10 MPa. A P-rho diagram shows that the point is above the critical pressure. Entering '300' for temperature and '10>' for pressure will return the valid state point at 10.826 mol/l.
Mixture Error Messages. The error message "No mixture data are available for one or more binary pairs in the specified mixture. The mixing parameters have been estimated." (message #117) occurs when an interaction parameter for two different fluids is not available in the program. The most common reason for this occurrence is a lack of experimental data to describe VLE, densities, and so forth for the binary pair. In other situations, such as many mixtures with water, interaction parameters have not been fitted to the REFPROP model even though data exist. An estimation scheme is available in the program to approximate one of the interaction parameters that helps improve bubble and dew point pressures. This error message indicates that you should be aware that calculated properties are estimates only. For similar fluids, especially among the refrigerants, the scheme works fairly well. It breaks down with dissimilar fluids and eventually the scheme will produce large interaction parameters and will report the error message "No mixture data are available for one or more binary pairs in the specified mixture. The mixture is outside the range of the model and calculations will not be made." The most common occurrence of this message comes from mixing methane with higher hydrocarbons that are not in the GERG-2004 mixture model. For example, mixing natural gas components with propylene will generate this message. With time we will add additional interaction parameters for those binary mixtures which are currently not described. At that time, further information will be made available about the status of the new mixing file.
.NET
Applications. A wrapper for .NET
and C applications has been written by Craig Fennessy
(of Cryogenic Industries) and he has given us permission to distribute his code
to others. The applications are included below:
RefProp_NET.vb
RefProp_NET.cs
Mixture Files. The Fortran code and Excel sample file in version 7.0 are not able to read the *.mix files, but that has changed in version 8.0. However, the .mix files are ASCII files and thus can be read from any editor and are quite simple to understand. The first two lines are just information and can be ignored.
Required Fluids for Distribution. There are several fluid files that Refprop
accesses in order to run properly. These are nitrogen.fld,
propane.fld, R134a.fld, and C12.fld (dodecane).
These fluids are used as reference fluids in extended corresponding states
methods employed in Refprop to predict transport properties for some instances.
These fluids (and the hmx.bnc file) should be distributed in addition to those
required in a particular application if the Refprop routines have been
incorporated into a software package.
Python Applications. An example file for the Python
programming language has been written by Bruce Wernick and he has given us permission to distribute his
code to others. The application is included below:
Refprop.py
MatLab Applications. The files that came with Refprop 8.0 for MatLab
will no longer work with version 2009a.
To first test the specified configuration, unzip the file, go to the resulting
directory, and run refprop_example in MatLab. This will use the g95 version of
repropm.mexa64. For 2009a, MatLab now
needs to specify types for pointers to values and arrays being passed to and
from MatLab. The types are mwPointer and mwSize. These are in the refpropm.f90 file. In order to use them you must include a
c-style header file in your Fortran code with the line #include "fintrf.h". The compiler must be able to find this file
with a -I path specified. In order for
g95 to preprocess the file you will need to use the -cpp
switch. Because refprop is coded in
fixed format Fortran you will also need -ffixed-format. In order to recompile, several files are
included to help with rebuilding the mexa64 file; these are:
#to rebuild refpropm.mexa64 from within matlab using
the g95 compiler use
make_refprop_mex_g95.m
#to rebuild refpropm.mexa64 from a shell prompt using the g95 compiler use
make_refprop_mex_g95.sh
#to rebuild refpropm.mexa64 from within matlab using
the gFortran compiler use
make_refprop_mex_gFortran.m
#to rebuild refpropm.mexa64 from a shell prompt using the gFortran
compiler use
make_refprop_mex_gFortran.sh
MatLab may not be able to recognize and compile Fortran files with the .FOR
extension, rather the extensions .for should work. refpropm.f90 was changed so it will work on
Windows 32 and 64 bit machines and Linux.
You can look there to see what paths were defined for the fluid files in
each case. If recompiling works then you
can reset the path however you would like.
64-bit Linux is the GLNXA64 case.
MatLab2009a.zip
MatLab2009a64bit.zip
2009a 64bit linux.zip
Excel
Applications. When using the methods described in the help
file (REFPROP.HLP) to make the Excel routines
available globally, the Step 1 Instructions will only work if Excel is
restarted after the environment variables have been modified. The
environment variables RPprefix and Path should look
similar to the following examples:
In the case of the path statement, only the highlighted text in the last
picture above should be added, do NOT delete anything else in the Path
variable. The information in the
examples worksheet of the Refprop Excel file explains where to find the
environment variables.
Another problem that occurs on some systems is that an old refprop.dll file had
been copied to the Windows or Windows\System32 directories. You should do a full search of your machine,
making sure that system directories are included, for the refprop.dll
file. If you find one in the Windows,
System, or System32 directories, then that file should be deleted.
In some cases where Excel cannot find the Refprop files, you can copy the Refprop
files into the root directory; thus if you work on C:, you can copy the Refprop
files into a new C:\REFPROP directory.
If you work on D:, then create a D:\REFPROP directory. When the program fails to find the fluid
files, it will look to see if a C:\REFPROP\FLUIDS (or D:) is available, and if
so it will use the files from that source.
In Excel 2003, the following may apply: If a user has called REFPROP from within
their own spreadsheet and has enabled the refprop Add-In as well as the macros
but it still fails to calculate data, the error can be overcome by first
loading the example spreadsheet and re-enabling macros when requested. The calculations invoked within the user's
spreadsheet should then work.
In Excel 2007, you
can add "C:\Program Files\REFPROP" to the
Excel 2007 appears to have a "feature" that is not in the 2003
version. In the older version, the following is allowed:
=Pressure("water", "TVAP", "SI", 298)
This results in a #VALUE! in the newer version. Excel 2007 may require the comma for the
missing parameter:
=Pressure("water", "TVAP", "SI", 298,)
The property quality in version 7 and 8
was assumed to always be given on a molar basis, even though it was labeled as
on a mass basis when mass units were in effect (this only affects mixture
calculations). The file has been fixed
and the latest is available below. This
file also contains the new subroutine FluidString
which replaces the Concatenate command making it easier to set up mixtures.
Refprop.xls
(last updated October 6, 2009)
Please send questions to Eric Lemmon: Eric.Lemmon@NIST.gov
Last modified: Nov. 18, 2009