• Project Description
• Development of the Speciation Database for the University of Houston Project
• Summary and Progress to Date on the American Chemistry Council Project (Description and Objectives) (Profile Database) (Speciation Processing Programs)
• Description of the Speciation Assignment File
• Profile Database File
• Additional Work Needed
• References
• Update History
• Database files or Access
• Related links

Volatile organic compounds (VOCs) differ significantly in their effects on ozone formation, and these differences need to be represented appropriately in the airshed models used to predict the effects of changes of emissions on formation of ozone. This requires appropriate methods to specify the chemical compositions of the many types of VOCs that are emitted, and appropriate methods to represent the chemical differences of these compounds in the models. However, current models and emissions databases have significant problems in this regard:

• Most models use highly condensed mechanisms that lump the many different types of VOCs into a limited number of model species.
• It is difficult and expensive to re-process emissions if chemical mechanisms are changed or more model species are added, and do not permit use of the full capabilities of chemically detailed mechanisms such as SAPRC-99 (Carter, 2000a)
• The chemical speciation information used in current databases have many poorly defined and redundant chemical categorizations, and different databases use inconsistent categorizations. This makes it difficult to implement new chemical mechanisms or to process emissions for different mechanisms in a consistent manner.

More recently, we have been funded by Pecan and Associates to make model species classifications for species categories in the Speciate 4.0 database that is now under development. This has necessitated the addition of a number of additional categories, which are now included in the speciation database developed for the previous projects. As part of this effort, model speciation assignments were made for the new Carbon Bond '05 mechanism, so four mechanisms (SAPRC-99, RADM2, CB4, and CB05) are now supported by the current database.

The emissions speciation database assignments and profile data compiled for UH, ACC and Speciate 4.0 projects are given in Excel files that can be downloaded from the links below. However, ultimately we need the chemical assignments for the speciation profiles to be in a universally available, web accessible format that can be updated as needed using state-of-the-art database software.

Development of the Speciation Database for the University of Houston Project

The development of the speciation database carried out under contract with the University of Houston has involved the following components:

• A list of all the chemical categories in the emissions speciation databases used by the EPA for processing emissions for Models-3 (Gipson, 2001), the EPA’s Speciate 3.2 database, (Speciate, 2001) the Texas Natural Resource Conservation Commission (TNRCC) (Yarwood, 2002), and the California Air Resources Board (CARB) (Allen, 2001) was compiled. The Speciate and California profile databases have since been updated as part of the ongoing ACC project.
• Duplicate categories were identified using various methods, a master list containing unique categories was compiled, and these unique speciation categories were assigned to the categories used in the various emissions databases. The resulting database contained 1855 speciation categories to represent the total of 3639 emissions categories from the 4 databases. Note that the database included some new categories that were added to define the compositions of the mixtures, as discussed below.
• CAS identification numbers were determined for all the speciation categories where these are well defined and uniquely identify a specific chemical or mixture, and were used to identify these categories using a "C"+CAS number designation (e.g., "C74-84-0" for ethane). The Chemfinder.com and ChemExper.com web sites were used as the primary means for this purpose, with the SciFinder Scholar program, which directly accesses the CAS database, being used for compounds not found on the other databases until that resource became unavailable as a free resource for academics in early 2002. Chemicals that do not have CAS numbers that we could determine or mixtures for which CAS numbers are not assigned or poorly defined were given separate identification codes based on SAROAD numbers used in EPA or other database (e.g., "S1-43155" for "isomers of heptadecane", based on the EPA SAROAD number, or "S2-97000" for "C4-naphthalenes", based on the CARB chemical codes.)
• Each of the new speciation categories were classified as either individual compounds, mixtures of isomers ("simple mixtures"), mixtures of compounds that are not necessarily isomers ("complex mixtures"), which were treated separately in the new database as described below. The current database has 1304 individual compounds, 319 simple mixtures, 200 complex mixtures, and 32 poorly defined classes.
• For each speciation category identified as an individual compound, the chemical formula (numbers of each type of atom) was determined and used as the basis for computing the molecular weight, and attempts were made to assign these compounds to model species in the SAPRC-99 detailed and condensed mechanisms and the Carbon Bond 4 (CB4) mechanism. 74% of the compounds could be assigned to SAPRC-99 detailed model species, 95% to SAPRC-99 lumped species, and 96% to CB4 groups. The more complete set of SAPRC-99 lumped species assignments resulted from the fact that a number of compounds did not correspond to existing SAPRC-99 detailed species but could be identified as being appropriately lumped with existing SAPRC-99 lumped groups. The CB4 assignments were based primarily on those in the TNRCC database, with most other CB4 assignments from databases received previously, though a few new CB4 assignments had to be made for this project.
• No direct assignments of model species were made to categories representing mixtures. Instead as many of these as possible were assigned to mixtures of compounds, whose model species assignments can then be used to determine the appropriate model species for the mixtures. This approach thus separates out chemical mechanism assignments from emissions composition assignments, so each can be updated separately and consistently as improved information or new mechanisms become available.
• For each speciation category identified as a mixture of isomers, a distribution of up to 8 individual compounds was chosen to represent the mixture for modeling purposes. The mixture was chosen based on either existing assignments in the SAPRC-99 mechanism or using a representative set of compounds of this type in the database. Such assignments were made for each of these "simple mixture" categories.
• Eventually, most complex mixture categories should be removed from speciation profiles used in emissions inventories, with the inventory developers providing their best estimate of the compositions in terms of actual compounds or mixtures of isomers. However, such categories are important in current inventories, and at present they cannot be neglected. Therefore, composition assignments were made for as many of the complex mixture speciation categories as possible. In most cases, a separate set of "profiles" were used to represent the compositions of the mixtures, though in a few cases representation by a single compound or simple mixture was considered to be appropriate. The representative profiles were taken from existing inventories or estimated in this work, and their derivations are documented in the database.
• Some of the 32 poorly defined emissions categories were identified as polymers or other substances that had no volatility, so these were identified as such so that their mass would not be counted when making assignments for emissions. The rest were treated as unspeciated mixtures for emissions processing purposes.
• Quality assurance was a major component of this project, and numerous procedures were used to test for consistency between the various databases and assignments employed. In some cases corrections had to be made based on interactions with Environ and others working on existing emissions databases. Nevertheless, because of the size of the database it is almost certain that at least some errors or inconsistencies may be present, so further checks and reviews are advisable.
• Speciation profiles used by the EPA and the CARB were obtained at the same time as the speciation categorizations (Gipson, 2001; Allen, 2001), and the extent to which the current compound and mixture assignments are sufficient to characterize these profiles was assessed. It was found that at least 99% of the mass was assigned to SAPRC-99 lumped species in almost 94% of the EPA profiles, and only one profile, "Solvent Utilization: Naphthenic Acids," had less than 50% of the mass assigned to detailed or lumped SAPRC-99 species. A somewhat larger number (~2%) of CARB profiles were less than 50% assigned, but ~94% of the profiles had over 90% of the mass assigned to SAPRC-99 model species, with the average for all profiles being 96%. Information was not available concerning Texas speciation profiles, so the extent of assignments in that database could not be determined. However, the coverage is expected to be similar to the EPA data base.
• Procedures were developed to utilize these assignments to derive the model species associated with each of the emissions categories used in the various databases. This includes procedures to produce ASCII files containing the mechanism assignments for the EPA, CARB, TNRCC, or Speciate 3.1 categories that can be used by emissions processing programs. However, formatting these files for use in existing emissions processing programs, or modifying existing emissions processing programs to utilize these files, was beyond the scope of the present project.

Project Description and Objectives

• Create the speciation database organization and assignments that appropriately represent relationships between mixtures, actual chemical compounds, and the model species in SAPRC 99 and other mechanisms. Much of this work has already been completed for the University of Houston project described above, but additional work is needed as described above. This task is being carried out primarily by us at CE-CERT.
• Develop the software and essential data to update both emissions processing methods and SMOKE for SAPRC and other mechanisms using the reorganized speciation database (developed in Task 1) and existing emissions datasets. This task is being carried out both by CE-CERT and UNC.
• Test SMOKE with the redesigned speciation database and SAPRC and other mechanism lumped species as modified by reactivity packets using data available from one previous modeling application. This is being carried out primarily by UNC
• Demonstrate application of reactivity controls to emissions using SAPRC and the Community Multi-scale Air Quality (CMAQ) modeling system. This is being carried out primarily by UNC
• Develop interim report, final report, and peer reviewed journal manuscript.

• Update SMOKE Tool to support using the SAPRC chemical mechanism with CMAQ.
• Design and implement a Java interface to prepare SMOKE control packets (including reactivity packets).
• Integrate Java interface with Models-3 and SMOKE.

Profile Database

Speciation Processing Programs

Figure 1. Diagram showing relationships between SMOKE and the speciation database programs and files. Mechanism-dependent files are indicated by dashed borders. The programs completed thus far are indicated with bold ovals, and the program still to be completed for this project is shown with the dashed oval.

Description of the Speciation Assignment File

The current set of speciation assignments (which incorporates work on the ACC project that will be described later) is incorporated in an Excel 2000 file emitdb.xls, which can be downloaded as indicated below. The file includes a "documentation" sheet that describes in detail the format of all the other sheets, and also a number of Excel macros that can be used to process the data or output ASCII files with the mechanism assignments. Given below is a summary of all the sheets incorporated in the current database file.
 

Documentation	Summarizes the purpose of each sheet and describes the columns or data fields they contain. Also gives the color conventions used throughout the file and reference citations for additional information on the mechanisms.
Parameters and Commands	Gives the parameters or definitions used in the various sheets or macros, and also lists the macros used and gives controls that can be used to run them. See comments in the sheet for details.
Master List	List of all the emissions classifications in my current database. Each class should be a unique compound or mixture.
Compounds	Lists the emissions categories that refer to single compounds (or mixtures of isomers that are not distinguished, optical isomers), and gives their mechanistic assignments
Simple Mixes	Lists the emissions categories that refer to mixtures of isomers, and summarizes their assignments to compounds. (Note that the assignments used are in a separate "Simple Mixes Asst's" sheet, and the assignments on this sheet are produced by the "CompileSimpMixAsnts" macro.)
Simple Mixes Asst's	Used to give the assignments of compounds to the simple mixtures listed in the "Simple Mixes" sheet. One row is used for each assignment, so multiple rows will be used for mixtures assigned to more than one compound.
Complex Mixtures	Gives summaries, compositional assignments, and other information about complex mixtures. Details of assignments and additional information about complex mixtures are given in separate sheets as indicated below.
Mix Profile Summary	Gives summary information about the profiles used to specify the compositions of some of the complex mixtures. The profiles are associated with the complex mixtures in the "Assignment of Mwt" column in the "Complex Mixtures" sheet for the mixtures with assignment type given as "CMP". The composition of the profiles is given in the "Mix Profile Assignments" sheet, and additional documentation notes are given in the "Profile Documentation Notes" sheet.
Mix Profile Assignments	Gives composition assigned to the profiles used to specify the compositions of some of the complex mixtures. One row is given for each component assigned to a profile, so in general a profile will have more than one column.
Profile Documentation Notes	Footnotes corresponding to numbers given in the "notes" column of the mixture profile summary sheet, giving additional information about the derivation of the composition of the mixture profiles used.
Other Categories	Lists and gives summary and other information about the emissions classifications that are poorly defined or inappropriate or have not yet been categorized. These include (1) compounds or mixtures that are not appropriate for VOC profiles; (2) Polymers, salts, or extremely high molecular weight materials; (3) categories that are not compounds or mixtures, such as elements or CB4 model species; (4) categories that are designated incorrectly or are too ambiguous to assign; (5) categories that may be compounds or mixtures of compounds but the compounds' structures could not be determined or have not yet been determined. (6) categories that have not yet been classified because they are not present or are extremely minor constituents in current profiles.
Emissions Groups	Gives the table of SAPRC-99 and emissions groups and the SAPRC-99 and RADM-2 lumped model species associated with each.
DMS	Lists SAPRC-99 detailed model species and gives relevant summary information, including lumped emissions groups assigned to each.
Master Assignments	Gives the complete mechanism assignments for all of the emissions categories in the database. All data on the sheet except the headers are derived using the "MergeAllCmpds" macro. Multiple rows are given for categories represented by more than one compound, one per compound used.
SAROAD Assignments	Assigns master emissions database categories to the SAROAD or chemical codes used in the EPA, California ARB, TNRCC, and Speciate 3.2 and 4.0 emissions databases. Also gives the complete mechanism assignments for these categories. The latter are derived by running the "ProcessSARmodelSpec" macro.
Emit Worksheet	Can be used as an example to derive detailed and lumped model species assignments for an emissions profile given in terms of a given type of SAROAD or pseudo-SAROAD categories.

The file incorporates several macros to output the data in the spreadsheet into various ASCII files that can be read by the speciation database processing programs that are being developed for the ACC project discussed above.

The current version of the speciation assignment file can be downloaded from ../emitdb/emitdb.xls

Profile Database File

The profile database file contains the profiles in the Speciate 3.2 database (Speciate, 2002), the profiles used by the EPA for processing emissions for Models-3 (Gipson, 2001), the Texas Natural Resource Conservation Commission (TNRCC) (Yarwood, 2002), and the profiles provided by the California Air Resources Board (CARB, 2003). The file contains a "documentation" sheet that describes the data in the file. Given below is a summary of the sheets incorporated in the current file.
 

Documentation	Summarizes the purpose of each sheet and describes the columns or data fields they contain.
Parameters	Gives the parameters and controls for the macros in this file and also lists and documents the profile types incorporated in the database. Gives the locations of the profile data (.EMI) files to be output by the macro in this file and the control to run the macro.
Profile Descriptions	Lists all the profiles and gives descriptive and documentation information available for them. Documentation consists of reference and note numbers that are associated with documentation text in the "Profile Reference" and "Profile Notes" sheets.
Profile Compositions	Gives the mass fractions of the chemical categories in the profiles, and also the descriptions of the categories used in the profile databases as obtained. Note that the categorizations used are those of the database from which the profiles were obtained, and differ for each database. The assignments of compounds to these categories is made in the speciation database incorporated in emitdb.xls.
Profile References	Gives the text associated with the reference numbers used in the "Profile Descriptions" sheet.
Profile Notes	Gives the text associated with the note numbers used in the "Profile Descriptions" sheet.

This Excel file also incorporates a macro that outputs the profile composition data and available documentation information into the ASCII profile data (.EMI) files that are used as input to the speciation processing programs discussed above

The current version of the profile database file,can be downloaded from ../emitdb/profdb.xls

Additional Work Needed

Although portions of this new database have been provided to others, the complete database has not been externally reviewed by other groups or checked for general utility or other problems. However, the available funding for the project for the University of Houston has been spent, so the remaining work for this project must be restricted to correcting any problems or inconsistencies found in the database and completing the final documentation and report. Additional work to incorporate this into the SMOKE emissions processing system is underway as part  an ACC project for the RRWG as discussed above.

Even after the current ACC project is completed, this speciation database must be considered to be a work in progress. Ongoing work that is needed is summarized below:

• The assignments of compounds to mixtures need to be reviewed by experts in emissions speciation. Arbitrary and undocumented assignments need to be improved, and assignments need to be made for currently unassigned classes. Note that this is best done by an emissions speciation expert rather than a chemical mechanism developer.
• CAS numbers need to be determined or assigned for compounds for which they could not be determined in current databases. This will probably require a public agency contracting with CAS to do this.
• The model species assignments to the individual compounds need to be independently reviewed and assignments need to be made to more of the chemical classes that are currently unassigned. This may require some chemical mechanism development to represent compounds that are not currently well represented in any mechanism.
• New compounds will need to be added as needed when new profiles using these compounds are developed, and mechanism assignments for these compounds will need to be made.
• Poorly defined or complex mixture categories will need to be removed as use of these categories are discontinued as part of ongoing projects to improve existing emissions profiles. Ultimately, the only mixture categories that should be retained are those necessitated by analytical uncertainties.
• The current database needs to be made available on the web in such a manner that emissions profile developers can readily utilize it to determine how to categorize the species in their profiles without adding duplicate or incompatible entries.

References

CARB (2003): Organic Gas Speciation Profiles provided by the California Air Resources Board at a web site that has been moved. Downloaded file named ORGPROF_03_19_03.xls and dated 3/19/2003.

Carter, W. P. L. (2000a): "Documentation of the SAPRC-99 Chemical Mechanism for VOC Reactivity Assessment," Report to the California Air Resources Board, Contracts 92-329 and 95-308, May 8. Available at ../reactdat.htm.

Carter, W. P. L. (2000b): "Implementation of the SAPRC-99 Chemical Mechanism into the Models-3 Framework," Report to the United States Environmental Protection Agency, January 29. Available at ../absts.htm#s99mod3.

Carter, W. P. L. (2007): "Development of the SAPRC-07 Chemical Mechanism and Updated Ozone Reactivity Scales," Final report to the California Air Resources Board Contract No. 03-318. August. Available at ../SAPRC.

Gipson (2001). PROFILE.VOC.DAT file, used for processing VOC emissions data for Models-3. Received from Gerald L. Gipson, EPA, Research Triangle Park, NC,. November, 2001

MCNC (2000): "Sparse Matrix Operator Kernel Emissions (SMOKE) Modeling System,"

SPECIATE (2001): Speciate 2.1 database was available at http://www.epa.gov/ttn/chief/software/speciate/ as of late 2001. Data in spreadsheet format provided by Ronald Ryan, EPA, Research Triangle Park, NC, November 2001. It has now been superseded by the Speciate 3.2 database, which is now what is available at that website.

Speciate (2002). Speciate 3.2 database available at http://www.epa.gov/ttn/chief/software/speciate/ as of early 2004. Dated November 3, 2002.

Yarwood (2002). Texas Natural Resource Conservation Commission emissions species data file "compound.database.eps2x.17May02", provided by Greg Yarwood, Environ corporation, Novato, CA, Latest version provided June 18, 2002

Allen (2001). California Air Resources Board chemical species data file provided by Paul Allen, CARB, Sacramento, CA, November 26, 2001. Superseded by the CARB (2003) profiles.